V. A.

Etkin

THERMOKINETICS

(SYNTHESIS OF HEAT
ENGINEERING
THEORETICAL GROUNDS)

Haifa, 2010
 V.A. Etkin . THERMOKINETICS (Synthesis of Heat Engineering
Theoretical Grounds).

The book calls attention to a method of describing and investigating various

physicochemical processes in their inseparable link with the heat form of energy.
The method is based on the law of energy conservation and free of hypotheses and
postulates.
All the basic principles, laws and equations of equilibrium and non-
equilibrium thermodynamics, heat- and mass-exchange theory, thermo-economics
and thermodynamics of finite-time processes are here derived from this method as
particular cases.
The book considers also with much attention phenomena at the interfaces
between heat engineering and other engineering disciplines, elaborates new
applications of the energy transfer and conversion theories, as well as analyzes
paralogisms arising in thermodynamics due to its inconsistent extrapolation.
The book is intended for researches, engineers and university students keen-set
for updating, extension and integration of knowledge in heat engineering
disciplines. It may be useful also for a wide audience interested in issues relating to
perfection of the modern natural science conceptual frameworks.

343 p. Fig. 36. Ref. 265.

Translation into English of the monograph "Thermokinetics", 2nd edition, has been
provided by N.V. Abashkin.

ISBN 978-0-557-25040-0
© V.A. Etkin
 INTRODUCTION

No sooner had the heat theory appeared, it immediately bifurcated into

two branches. In 1822 a known J. Fourier’s work appeared, which laid the
foundation of the heat transfer theory; in 1824 not less famous S Carnot’s
work laid the foundation of thermodynamics. Both works were based on the
afterward-rejected theory of thermogen as the indestructible fluid, both
considered temperature as some potential which gradient conditioned the
heat transfer direction or conversion of heat into ordered forms of .
However, both branches mentioned developed quite independently without
any points of contacts. Their difference showed up not only in terminology,
but it rather rooted in a basic methodological nature. Carnot-Clausius’
thermodynamics maintained aloofness from the transfer ideas and the heat
exchange rate concept. The heat exchange theory, on the contrary, had
nothing to do with the conversion of heat into other forms of and
considered entropy as an extraneous concept. The so “fancy separation of
two branches within the same area in macroscopic physics” (to K.
Denbigh’s locution) was too hard to be overcome by the TIP as well. Even
today the definition of heat concept remains different in thermodynamics
and the heat exchange theory. In thermodynamics this is the part of
exchange caused by exclusively the temperature difference between bodies
and not associated with substance exchange between them
(Thermodynamics. Terms, 1973). The heat exchange theory, on the
contrary, considers heat as the part of internal associated with random
motion (because a system can exchange just what it has) and studies, along
with heat conductivity, the heat transfer carried out by substance and
enabled by heterogeneity of the fields of other physical values (Heat
Transfer. Terms, 1980). Such a situation demands searching more cardinal
means to unify the two said fundamental disciplines.
Further, the growing comprehension of the fundamental role the rate
and productivity play in real processes as one of the basic indices of their
efficiency gave rise to two new branches in thermodynamics of the XX
century called thermodynamics of irreversible processes (TIP) (Onsager, L.,
1931; Prigogine I., 1947, 1955; Casimir H., 1945; Denbigh K., 1951; De

3
Groot S., 1952, 1962; Meixner I., 1954; Gyarmati I., 1960, 1970; Haase R.,
1962, et al) and finite-time thermodynamics (Curson F., Ahlborn B., 1975;
Rubin M., 1979, 1980, 1983; Andresen B., Salamon P., Berry R., 1977,
1980, 1982, 1984; Band Y.B., Kafri O., 1981; Rozonoer L., Tsirlin A.,
1983; Linden C., 1992, et al), respectively. The first of the branches
involves time as physical parameter introduced into the thermodynamic
equations and a new macrophysical method developed on this basis for
investigating kinetics of interrelated transfer phenomena. This branch has
enriched the theoretical conception of the XX century with a number of new
principles of general-physics meaning (including those of linearity,
reciprocity, minimal entropy generation) and evidently contributed into
cognition of the in-depth interrelations between dissimilar phenomena.
However, TIP basing on the entropy rise law is restricted to investigating
dissipative processes such as thermal conduction, electrical conduction,
diffusion, as well as effects of their superposition, but has nothing to do
with the issues of productivity of useful conversion processes which are the
principal object of thermodynamics. As a result, the broadest spectrum of
irreversible processes with a relative efficiency above zero appeared to have
fallen beyond cognizance of this theory. The second branch, on the contrary,
sets as a top-priority task the definition of the conditions to achieve
maximum useful of cyclic heat engines with due consideration given to the
irreversibility of heat exchange processes and finite-time contact between a
working media and heat wells and sinks. Within the frames of this theory
the problem has been posed for the first time and in the most general form
about the relation between the capacity (productivity) of engineering
systems and their thermodynamic efficiency, as well as about the critical
capabilities of irreversible processes. However, this theory in its existing
configuration is restricted to units wherein the operation at maximum power
conditions is economically most advantageous. Though the spectrum of
such units is broad enough (includes nuclear power plants, renewable
plants, space vehicle power plants, etc), it does not include a number of
power and process units wherein maximum of their efficiency does not
correspond to maximum power. In this context a necessity arises to
synthesize the said branches and create on this basis thermokinetics as a
unitary theory of rate and productivity pertaining to transfer and conversion
processes, which would cover the entire spectrum of real processes and
occupy the same position relative to the classic theory of heat engines as
dynamics to statics.
The necessity to create a division of thermodynamics supplementing the
classic theory of heat engines with the analysis of interrelation between
thermodynamic efficiency and productivity (net capacity) of various kinds of
converters (cyclic and non-cyclic, heat and non-heat) is dictated by logic of
the development in many fields of knowledge. Kinetics of useful
conversion processes is a concern to not only power engineering and power

4
processing wherein these processes are principle. The thermodynamic
investigation of biological systems is also impossible without catering for
work supporting the non-equilibrium state of such systems and providing
their vital activity. The application of thermodynamics to cosmological
objects that develop, according to current ideas, bypassing equilibrium
would also be incomplete without work considered as ordered form of
exchange. This refers to also antirelaxation phenomena observed under
regular conditions at superposition of dissimilar irreversible processes
(superposed when running simultaneously in the same spatial areas) and
studied by synergetics.
The practicability to revise the grounds of thermokinetics in this case is
dictated by the well explainable wish to avoid application of whatever
hypotheses and postulates preventing TIP from achieving the completeness
and rigor incident to the classic thermodynamic method. The acuteness of
the problem intensifies with the necessity to overcome the certain restriction
of the theory of irreversible transfer processes to linear systems and states
close to equilibrium.
The thermodynamic theory of real processes of transfer and conversion
in spatially heterogeneous media offered to the reader features a system
approach to the theory construction (from the whole to the part), the
exclusion of whatever hypotheses or postulates from theory grounds (like
the “local equilibrium principle” or the “beginnings” of the classic theory of
heat engines) and the negation of the traditional idealization of processes
(their “quasi-static character” and “linearity”, ideal character of cycles and
their working media, etc). This theory is based on exclusively the empirical
facts that underlie the law of conservation and the representation in terms
of state parameters of systems under investigation. All other information
about an object under investigation thermokinetics attributes to uniqueness
conditions. The phenomenological and deductive theory of such a kind
keeps undisturbed the key advantage of the classic thermodynamic method,
viz. the immutable validity of its consequences. This is confirmed in the
book by the fact that all the basic principles, laws and equations pertaining
to equilibrium and non-equilibrium thermodynamics, as well as to the heat-
mass exchange theory, which have gained multiple experimental
verifications – all have been therein obtained as particular cases of
thermokinetics under adequate uniqueness conditions. This allows, on its
base, not only to correctly interpret various phenomena at the interfaces
between thermodynamics and mechanics, physical chemistry,
hydrodynamics, electrodynamics, biophysics, astrophysics, economics, etc,
but also to predict new effects arising at the superposition of processes these
disciplines study. It is significant that such an application of
thermodynamics does not require a preliminary study of the above
disciplines, neither of statistical physics on which base the TIP key
principles have been earlier obtained. This allows a pathbreaking study of

5
phenomena at the interfaces of fundamental disciplines as a unitary
academic course of thermodynamics.
Much attention is given in the book to analyze and eliminate the
paralogisms arising in thermodynamics from its unreasonable extrapolation
beyond the primary concepts of equilibrium and reversibility, as well as to
obtain a great number of other nontrivial consequences following from the
consideration of spatially heterogeneous media as a unitary non-equilibrium
whole. Besides, the book lays the foundations of two new applications of
thermodynamics, viz. the theory of similarity and the theory of productivity
of conversion systems, wherein a wide step is made on the way
approximating the results of their thermodynamic analysis to reality.
The book offered to the reader consists of five parts. Part 1 –
Methodological Principles of Thermokinetics (Chapters 1, 2) – describes the
features of thermokinetics as a single to date scientific discipline free of
hypotheses based on exclusively the law of conservation of and its
representation in terms of measurable (or calculable) parameters of non-
equilibrium systems being investigated.
While keeping the advantages of thermodynamics as a consistently
phenomenological (i.e. exclusively empirical) and strictly deductive (i.e.
developing from the general to the special) scientific discipline
thermokinetics at the same time excludes from its grounds whatever
postulates (like the “beginnings” of classic thermodynamics) and negates
the idealization of processes and systems when constructing theory grounds
and substantiating theory laws. Whatever simplifying assumptions are
admitted here at only uniqueness conditions thermokinetics imports from
outside at the final investigation stage when solving particular problems.
Unlike thermodynamics which is substantially thermostatics,
thermokinetics as a generalized theory of heat transfer and conversion
process rate is built on an extremely general notional and conceptual base
not alien to the transfer and real process productivity concepts. The
irreversibility of such processes is here catered for by introducing specific
spatial heterogeneity parameters varying arbitrarily with relaxation of a
system like entropy, but, unlike entropy, varying with also reversible work
the system does. A rather simple way to find these parameters if offered,
which pioneers the thermodynamic description of spatially heterogeneous
systems as a whole.
The major feature of the thermokinetic notional system is that the is
brought back to its simple and clear primordial meaning as a capacity of a
system to do work. This is obtained by generalizing the notion of work and
defining it as a quantitative measure of a process associated with
overcoming any (long-range and short-range, ordered and unordered)
forces. The further division of into its ordered and unordered parts and
introducing for those characteristic functions as expressed by various groups
of parameters for systems under investigation enabled the quantitative and

6
qualitative description of system order and convertibility of in various
forms. Such a generalization of the thermodynamic method of potentials
allowed suggesting criteria of evolution and partial equilibrium much more
informative than entropy and having a body of mathematics for
thermokinetics equally applicable for study of processes with any
irreversibility degree.
Part 2 – Heat Engineering Fundamentals (Chapters 3-5) – describes the
application of mathematical and notional tools of thermokinetics in order to
substantiate all basic principles, laws and equations of the heat engineering
disciplines – classic and non-equilibrium thermodynamics, as well as the
heat- and mass-exchange theory. The methodologically unitary exposition
of all mentioned disciplines as consequences from thermokinetics is the
most important result of this part. This approach eliminates the historical
delimitation of thermodynamics and the heat-exchange theory as expressed
in not only terminology and different interpretation of the transfer substrate,
but in the methodology of these disciplines as itself. At the same time the
book is a pioneer work to provide a consistently thermodynamic (with no
recourse to hypotheses, postulates and statistical-mechanical considerations)
substantiation of all statements pertaining to the theory of irreversible
processes generalizing the heat-exchange theory to the associated
phenomena of substance, charge, momentum, etc transfer. Thus a real
possibility opens to include this subdivision of thermodynamics – fresh and
implying multiple applications – into academic curriculums, which is a
telling contribution to solution to the problem of scientific knowledge
integration.
Part 3 – Elimination of Negative Consequences of Thermodynamics
Extrapolation (Chapters 6-10) – exposes the contradictions arisen in
thermodynamics because of its ungrounded extrapolation beyond the
validity of its primary concepts pertaining to equilibrium and reversibility.
Here come, in particular, the conclusion of inevitable “jump” of entropy
when mixing non-interacting and however poorly distinguishable gases
(Gibbs’ paradox); the appearance of thermodynamic inequalities when
coming on to irreversible processes; the conclusion of violated law of
excluded perpetual motion of the second kind in open, spin, relativistic, etc
systems; the extension of the ban on environmental heat usage to non-heat
forms conversion; the extrapolation of the entropy rise principle to the
Universe resulting in the theory of “heat death” of the Universe; the
application of the relativistic transformation to absolute values; the
introduction of negative absolute temperature; the substitution of statistical-
informational entropy for thermodynamic one; the statement of the heat
conversion laws as exclusive; the attempts to apply the theory of dissipative
processes to anti-relaxation structuring in biological processes, etc. It is
shown here that such deductions are actually paralogisms, i.e. unintended
logical errors caused by inconsistent extrapolation of classic

7
thermodynamics. In each case the sources of the difficulties arisen are
exposed and the way is pointed out how to overcome them from the
positions of thermokinetics.
Part 4 – Further Generalization of Heat Engine Theory (Chapters 11-13)
– presents new results obtained from application of thermokinetics to
conversion processes in heat engines. The results include the theory of
conversion processes similarity generalizing the theory of heat engines to
non-heat and non-cyclic ones (including the muscular movers of bio-
organisms); the theory of engineering systems productivity combining
thermodynamics with two new fields of its application, viz. “thermo-
economics” and “finite-time thermodynamics”. The major result of such an
approach is the substantiation of the possibility to use field forms of
differing from its material sources. This opens up new vistas in creating
machines on renewable sources being now erroneously attributed now to
the category of “perpetuum mobile”.
The last part (Part 5) of the monograph – Extension of Transfer Theory
Applicability (Chapters 14-16) – generalizes the existing theory of
irreversible processes to non-linear systems and states standing far from
equilibrium. A new method is offered for investigating multiple thermo-
mechanical, thermo-chemical, thermo-electrical, etc phenomena caused by
“superposition” of dissimilar transfer processes in poly-variant systems and
being of great practical interest. The method neither cumbersome equations
of entropy balance nor application of reciprocal relations (violated in non-
linear systems) while resulting in simplified transfer laws and further
reduction of empirical coefficients they contain. This part also describes the
application of thermokinetics to bio-systems exposing a special role of
useful conversion processes in functioning of bio-systems – contrary to the
existing ideas of their anti-entropy character.
All the said along with the methodologically unitary exposition of heat
engineering fundamentals confirms the uniqueness and heuristic value of
thermokinetics as a pro-thermodynamic method of investigating multi-
functional systems completely based on system approach.

8
 Part 1
---------------------------------------
THERMODYNAMICS AS THEORETICAL
BASIS OF HEAT ENGINEERING

Each independent scientific discipline features its own subject and

method. Thermokinetics may be considered as a basically thermodynamic
method of investigating real processes of energy transfer and conversion in
any forms being in inseparable link with the heat form of motion.Being
consistently deductive, this method meets the system approach requirements
with studying a part through the whole as one of them. That allowed classic
thermodynamics to obtain, based on few primary principles (“the
beginnings”), a great number of consequences covering various phenomena
and having the status of oracles within the applicability of its primary
concepts. That imparted an exclusive position to thermodynamics among a
number of other scientific disciplines. However, when changing to studying
non-equilibrium systems with irreversible processes running therein such an
approach gave place to the inductive investigation method adopted in classic
mechanics and a number of other fundamental disciplines. Continuum was
divided there into elementary parts which may be considered locally
equilibrium. The method assumes the possibility to study the whole through
its elementary parts by summarizing their extensive properties. However,
system-forming properties of real objects are basically non-additive.
Therefore the theory of irreversible processes as a substitution for classic
thermodynamics has essentially denied the system approach and has not
gained the completeness and rigor inherent to the classic thermodynamic
method.
The intention to keep the advantages of the thermodynamic method
when generalizing it to non-thermal forms of energy and non-static
processes impelled us to base thermokinetics on the same methodological
principles of the consistently phenomenological and deductive theory as
classic thermodynamics. However, we have based thermokinetics not on the
laws of excluded perpetual motion running now the gauntlet of all-round
criticism, but on only those truisms that do not need an additional
experimental verification.

9
 The thermokinetic approach to the body-of-mathematics construction is
also different. It features using the whole arsenal of mathematical theorems
covering a particular math model structure for the object under investigation
(rather than “contriving” one’s own means required for its investigation).
Such an approach after N. Burbaki (1959) is called “structural-
mathematical”. It is based on the classic method of thermodynamic
potentials, i.e. characteristic functions, the math operations on which define
all properties of investigator’s interest with regard to the object under
investigation. This method is generalized to spatially heterogeneous systems
by introducing, along with usual (thermodynamic) parameters such as
temperature, pressure, volume, etc, also specific parameters of spatial
heterogeneity. It rests on the properties of the exact differential of a number
of non-equilibrium state functions and thus features quite general character.
As for the rest of data describing the object under investigation,
thermokinetics “imports” it “from outside” as uniqueness conditions of a
kind when solving particular problems. Due to this all assumptions the
investigator has made on the uniqueness conditions (hypotheses on
substance structure, on molecular mechanism of processes, considerations
on their properties correlation character, etc) do not touch the crux of the
theory itself, i.e. the relations following from the math properties of
characteristic functions. Such an approach makes it possible to keep the
main advantage of the classic thermodynamic method and to lay it into
foundation of a number of natural-science disciplines.

Chapter 1

METHODOLOGICAL PRINCIPLES OF THE THERMODYNAMICS

There are periods occurring now and then in the development of any
natural-science theory when new ideas and experimental facts can not be
crammed into “Procrustean bed” of its obsolete notional and conceptual
system. Then the theory itself – its presuppositions, logical structure and
body of mathematics – becomes the object of investigation.
Thermodynamics went through such periods more than once (Gelfer, 1981).
So was yet in the mid-XIX century when under the pressure of new
experimental facts the concept of heat as an indestructible fluid collapsed
and “entrained” (as seemed then) the S. Carnot’s theory of heat engines
(1824) based on it. A few decades later the threatening clouds piled up over
the R. Clausius’ mechanical theory of heat (1876) because of the “heat
death of the Universe” – a conclusion deemed then as inevitable.

10
 In the late XIX century great difficulties arose from attempts to conduct
a thermodynamic analysis of composition variation in heterogeneous
systems (at diffusion, chemical reactions, phase transitions, etc). J. Gibbs
(1875) overcame the majority of those difficulties by representing closed
system as a set of open subsystems (phases and components), which
allowed him to reduce the internal processes of system composition
variation to the external mass transfer processes. However, some of those
difficulties have remained as yet and are showing, in particular, in the
unsuccessful attempts to thermodynamically resolve the “Gibbs’ paradox” –
a conclusion of stepwise entropy rise when mixing non-interacting gases
and independence of these steps on the nature the gases feature and the
degree they differ in (Chambadal, 1967; Kedrov, 1969; Gelfer, 1981;
Bazarov, 1991).
During the XIX century thermodynamics also more than once
encountered paradoxical situations that arose around it with the human
experience outstepped. One of such situations arose with thermodynamics
applied to the relativistic heat engines (contain fast moving heat wells) and
showed in the statement that those could reach efficiency higher than in the
Carnot’s reversible engine within the same temperature range (Ott, 1963;
Arzelies, 1965; Meller, 1970; Krichevsky, 1970), as well as in the
recognized ambiguity of relativistic transformations for a number of
thermodynamic values (Bazarov, 1991). A little bit later a situation, not any
less paradoxical, arose as connected with attempts to thermodynamically
describe the systems of nuclear magnets (spin systems) with inverted
population of energy levels. The negative absolute temperature concept
introduced for such states led investigators to a conclusion of possibility for
heat to completely convert into work in such systems and, on the contrary,
impossibility for work to completely convert into heat, i.e. to the
“inversion” of the principle fundamental for thermodynamics – excluded
perpetuum mobile of the second kind (Ramsey, 1956; Abragam, 1958;
Krichevsky, 1970, Bazarov, 1991).
That fate became common for also the theory of irreversible processes
(TIP) created by extrapolating classic thermodynamics to non-equilibrium
systems with irreversible (non-static) processes running therein. Problems
arose primarily from the necessity to introduce into thermodynamics the
transfer concepts inherently extraneous for it, from the incorrectness to
apply the equations of equilibrium thermodynamics to irreversible processes
in view of their inevitable change to inequalities, from the inapplicability of
the classic notions of entropy and absolute temperature to thermally
heterogeneous media, etc, which demanded to introduce a number of
complimentary hypotheses and to attract from outside balance equations for
mass, charge, momentum, energy and entropy with time involved as a
physical parameter. Even heavier obstructions arise with attempts to
generalize TIP to non-linear systems and states far away from equilibrium

11
where the Onsager-Casimir reciprocal relations appear to be violated (S.
Grot, 1956; R. Mason et al, 1972) and the law of entropy minimal
production becomes invalid (I. Prigogine, 1960; I. Gyarmati, 1974).
Attempts to overcome these difficulties without whatever correction on the
conceptual fundamentals and body of mathematics of classic
thermodynamics failed.
A remedy can be found in building of modern thermodynamics on its
own more general notional and conceptual foundation with maximal care
for the classic thermodynamic heritage.

1.1. Exclusion of Hypotheses and Postulates from Theory Grounds

One of the most attractive features of the thermodynamic method has

always been the possibility to obtain a great number of consequences of
various phenomena as based on few primary principles (“the beginnings”),
which are empirical laws in their character for the thermo-mechanical
systems. Being consistently phenomenological (i.e. empirical), that method
enabled to reveal general behavior of various processes without intrusion
into their molecular mechanism and resort to simulation of structure and
composition of a system under investigation. Therefore, it is not by pure
accident that all the greatest physicists and many mathematicians of the last
century (Lorenz, Poincaré, Planck, Nernst, Caratheodory, Sommerfeld,
Einstein, Born, Fermi, Neiman, Landau, Zeldovich, Feynman, etc) in their
investigations placed high emphasis on thermodynamics and, based on it,
have obtained many significant results.
However, thermodynamics have presently lost its peculiar position
among other scientific disciplines. It sounds now in increasing frequency
that thermodynamics relates to real processes to the same degree as
Euclidean geometry to the Egyptian land surveyors’ work. Such a
standpoint is not groundless. Classic thermodynamics is known to have
always done with two primary postulates taken for its “beginnings” – the
laws of excluded “perpetuum mobile” of the first and second kinds. Those
principles have had the exclusion character and empirical status. However,
classic thermodynamics restricted to those two laws appeared to have been
unable to solve the problems that arose with its extension to phenomena of
another nature. So in consideration of open systems exchanging substance
with the environment, the entropy absolute value and the substance internal
energy had to be known. To know the values, the third “beginning” would
be needed as stating their becoming zero at the absolute zero of temperature.
In-depth analysis of the thermodynamic logic structure in works of C.
Caratheodory (1907), T.A. Afanasjeva-Erenfest (1926), A.A. Guhman
(1947, 1986) and their followers later led to the comprehension that the
second law of thermodynamics would need to be split in two independent

12
laws (existence and rise of entropy), as well as to realizing the important
role of the equilibrium transitivity principle named the zeroth law of
thermodynamics (Gelfer, 1981). Starting to study non-equilibrium systems
with irreversible processes running therein additionally required the L.
Onsager’s reciprocity principle sometimes named the fourth law of
thermodynamics from the phenomenological positions. Further
investigations have revealed the fundamental difference between statistical
thermodynamics and phenomenological thermodynamics and the
fundamental role that plays for the latter the equilibrium self-non-
disturbance principle, which has been assigned a part of its “general
beginning” (I. Bazarov, 1991). Thus present day thermodynamics appears to
be arisen from not two, but even seven beginnings! Meantime, the
disputable consequences of thermodynamics are growing in number thus
causing doubts in its impeccability as a theory. As R. Feynman wittily noted
about this, “we have so many beautiful beginnings…but can’t make ends
meet nonetheless”.
The law of excluded perpetuum mobile of the second kind being denied
in open system thermodynamics (M. Mamontov, 1970), relativistic
thermodynamics (H. Ott, 1963), spin system thermodynamics (M.
Vukalovich, I. Novikov, 1968) excludes the possibility for thermokinetics to
be based on the postulates of such a kind adopted for “the beginnings”. The
grave dissatisfaction investigators feel with such state of affairs has resulted
in multiple attempts to build thermodynamics as based on other fundamental
disciplines. This tendency has been most highlighted by A. Veinik (1968) in
his thermodynamics of real processes based on a number of postulates of
quantum-mechanical character, by M. Tribus (1970) in his informational
thermodynamics based on the information theory formalism, and by C.
Truesdall (1975) in his rational thermodynamics topology-based. All these
theories feature a denial of the consistently phenomenological (i.e. based on
only empirical facts) approach to the theory of irreversible processes, which
deprives them of the basic advantage intrinsic for the classic thermodynamic
method – the indisputable validity of its consequences.
In our opinion, one of the reasons of such a situation is that
thermodynamics has lost its phenomenological nature with considerations of
statistical-mechanical character gaining influence in its conceptual basis.
Whereas the founders of statistical mechanics strived to lay the
thermodynamic laws into the foundation of statistical theories, a statement
has become now common that phenomenological thermodynamics itself
needs a statistical-mechanical substantiation (despite “there are much
ambiguity” in the grounds of the statistical theories (R. Cubo, 1970)). In
particular, L. Onsager, the founder of the theory of irreversible processes
(TIP), in order to substantiate the most fundamental concept of his theory –
reciprocal relations, appealed to the principle of microscopic reversibility,
the theory of fluctuations with a complementary postulate for linear

13
character of their attenuation. All these statements evidently outspread
beyond the thermodynamic applicability, therefore L. Onsager, not without
reason, termed his theory quasi-thermodynamics.
Adoption of the local equilibrium hypothesis (I. Prigogine, 1947) for a
basis of TIP construction became even “further-reaching” assumption. This
hypothesis assumes (a) equilibrium in the elements of heterogeneous
systems (despite the absence of the necessary equilibrium criterion therein –
termination of whatever macro-processes); (b) possibility to describe their
status with the same set of parameters as for equilibrium (despite the actual
use of additional variables – thermodynamic forces) and (c) applicability of
the basic equation of thermodynamics to these elements (despite its
inevitable transformation into inequality in case of irreversible processes).
As a result, the existing theory of irreversible processes does not reach the
rigor and completeness intrinsic for the classic thermodynamic method.
Striving for excluding postulates from the grounds of thermokinetics
dictates the necessity to base thermokinetics on only those statements that
are beyond any doubt and construed as axioms. These statements include, in
particular, the equilibrium self-non-disturbance axiom reading that a
thermodynamic system once having reached equilibrium cannot
spontaneously leave it. Unlike the equilibrium self-non-disturbance
principle (general law of thermodynamics), this axiom does not claim that a
thermodynamic system, being isolated, reaches equilibrium for a finite time.
The axiom just reflects the evident fact that processes in a system that has
reached equilibrium may be generated by only impact applied to it from
outside and are, therefore, never observed in isolated systems. Being a result
of the experience accrued, this axiom excludes the possibility the macro-
physical state of a system will vary as a result of short-term spontaneous
deviations from equilibrium (fluctuations) caused by the micro-motion of
the constituent particles. Indeed, if fluctuations do not cause any variation in
the microscopic (statistical in their nature) parameters of the system, they
can not be considered as an energy-involving process since the energy of the
system remains invariable. Here lies the fundamental difference between
thermokinetics and statistical physics – the latter does consider fluctuations
as the object of investigation. At the same time the equilibrium self-non-
disturbance axiom allows for existence of systems that omit the equilibrium
state in their development since this axiom does not claim for relaxation
time finiteness, which is hardly provable.
The process distinguishability axiom is another primary statement
thermokinetics appeals to. It states there are processes existing and
definable (by all experimental means) which cause system state variations
as specific, qualitatively distinguishable and irreducible to any other ones.
In classic thermodynamics these are isothermal, isobaric, adiabatic, etc
processes. It will be shown hereinafter that these two axioms, in conjunction
with experimental data underlying the energy conservation, are enough to

14
construct a theory both internally and externally consistent and generalizing
thermodynamics to transfer processes and conversion of energy in any
forms.

1.2. System Approach to Objects of Investigation

The deductive interpretation of classic thermodynamics (thermostatics)

and the theory of non-static (irreversible) transfer processes as
consequences from thermokinetics demands considering systems of a more
general class as the object of its investigation. Classic thermodynamics is
known to have been restricted to the investigation of intrinsically
equilibrium (spatially homogeneous) systems wherein the intensive
parameters such as temperature T, pressure p, chemical, electrical etc
potentials were similar for all points of the system. It was dictated by not so
much the simplicity of system description as by the necessity to keep the
equations of thermodynamics from passing into inequalities with the
spontaneous variations of system parameters. Thermodynamics of
irreversible processes resting upon the local equilibrium hypothesis divided,
with the same purpose, a system in whole heterogeneous into a number of
homogeneous subsystems (down to elementary, supposedly equilibrium,
volumes). The same is typical for the field theories such as continuum
mechanics, hydrodynamics and electrodynamics, which also assume the
continuum properties to be at that identical in whole to the properties of
these elements and may be described in terms of relevant integrals.
However, the extensive properties of heterogeneous systems are far from
being always additive ones, i.e. the sum of properties of constituent
elements. First of all, non-additive is the property of a heterogeneous
system to do useful work as none of its local parts possesses it. It was
S. Carnot who awoke to that statement in application to heat engines
(1824) and put it into historically the first wording of the second law
of thermodynamics. According to it, only thermally heterogeneous
media possess a “vis viva (living force)”, i.e. are able to do useful
work. In itself the notion of perpetual motion of the second kind as a
system with no heat well and heat sink in its structure evidences the
importance of considering such media as a single whole (but not as a
set of thermally homogeneous elements). This is just the reason, why,
at study of heat engines, the so-called “extended” systems have to be
considered, which include, along with heat wells (sources), also heat
sinks (receivers) (the environment).
Another non-additive property of heterogeneous media is the
internal relaxation processes progressing and resulting, in absence of
external constraint, in the equalization of densities, concentrations,
15
electric charges, etc. among various parts of such a system. These
processes are, however, absent in any element of the continuum
considered as a locally equilibrium part of the system.
More non-additive properties are the self-organization ability of a
number of systems, which is absent in any of their homogeneous part
(S. Keplen, E. Essig, 1968; I. Prigogine, 1973,1986), as well as the
synergism (collective action) phenomena appearing at only a definite
hierarchic level of the system organization. The said refers in general
to any structured systems, which specific features are determined by
the inter-location and inter-orientation of the functionally detached
elements and disappear with decomposition of the object of
investigation into these elements (G. Gladyshev, 1988). Many of such
elements (e.g., macromolecules and cells) being detached remain,
however, spatially heterogeneous (locally non-equilibrium) despite
their microscopic size (constituting microcosmos of a kind). This
demands them being approached in the same way as the “extended”
macro-systems.
For the further reasoning it is very important to show in the most
general form that the main differential calculus technique – discretization of
an object of investigation into infinity of homogeneous elementary volumes
– results in the loss of backbone (system-forming) links. With this in mind
let us consider the most general case of a system comprising the entire set of
interacting material objects. Such a system is isolated by definition, while
its internal energy U remains invariable with time t, i.e. dU/dt = 0. Let us
represent this energy as the space integral U = ∫ρudV of the energy density
ρu. This gives for the system as a whole:

dU/dt = ∫ (∂ρε/∂t)dV = 0. (1.2.1)

Integral (1.2.1) may be equal to zero with some processes available in a

system if only the sign of the derivative (∂ρu/∂t) is opposite in different
areas of this system. This conclusion regards not only energy, but any other
parameter obeying the law of conservation (mass, charge, momentum and
angular momentum of the system), as well as the internal forces acting in
such a system (as a closed one). From this it follows the dichotomy law as
the major assertion for natural science in whole: processes running in
isolated (closed) systems cause opposite variations of the properties in the
different parts (areas) of such systems. That is why the set of interacting
systems acquires new properties. From this it follows the main feature of the
system approach, viz. the requirement to keep all backbone links
undisturbed when investigating some part of a system, i.e. investigating the
part through the whole (but not vice versa).

16
 The system approach involves the binding consideration of spatial
heterogeneity of the object under investigation. This becomes a necessary
requirement for any theory pretending to be the one adequately explaining
the reasons of processes running within whatever system. It is hardly
necessary to prove the invalidity of the fundamental sciences to meet this
requirement when they fraction systems into infinity of “elementary” areas,
material points or particles supposed to be intrinsically equilibrium
(homogeneous).
The such approach to objects of investigation dictates in itself the
necessity of changing to consideration of spatially heterogeneous
systems, which dichotomy (existing subsystems with opposite
properties) is the reason of whatever processes arising therein. In line
with this requirement the object of investigation by thermokinetics
comprises spatially heterogeneous media considered as a single non-
equilibrium whole. The size of such a system depends on the rate of
its heterogeneity; therefore thermokinetics covers the broadest range
of material objects – from nanoworld to megaworld – providing their
properties can be successfully described with a finite number of state
parameters. At the same time thermokinetics does not either exclude
from consideration such a set of interacting bodies which may be
considered, with acceptable accuracy, as a closed or isolated system.
Thus, what was considered in thermodynamics as an “extended”
system (including the environment besides the system itself), in
thermokinetics becomes just a part of the system (subsystem) existing
along with the similar other subsystems or with the object of work. In
this case thermokinetics, from the very outset, has been consistent
with the general scientific paradigm stating that any material object
may not be deprived of its key property – extent, while any extended
object – structure determined by its spatial heterogeneity.

1.3. Negation of Process and System Idealization outside

the Framework of Uniqueness Conditions

Thermokinetics as a successive science needs to be extremely delicate

to the classic thermodynamic inheritance. First of all this applies to the
scope of the correctives introduced at that into the primary notions of
thermodynamics. Let us dwell on those absolutely necessary in view of
changing to consideration of systems of a broader class. Such a correction
relates, in the first turn, to the notion of process as itself because of existing
in heterogeneous systems a specific class of stationary irreversible
processes wherein local parameters of a system as the object under

17
investigation remain invariable despite the flows of heat, substance, charge,
etc available in this system. Striving to keep the primary notion of “process”
as a succession of state variations makes it necessary to define this notion as
any space-time variation of macro-physical properties pertaining to an
object of investigation. Thereby the state variations associated with the
spatial transfer of various energy forms are included in the notion of
process.
Changing to consideration of real processes also demands to negate the
process idealization as implied in such notions as the quasi-static,
reversible, equilibrium, etc process. The notion of process as a sequence of
state variations of an object under investigation and the notion of
equilibrium as a state featuring the termination of whatever processes are
mutually exclusive. To eliminate this contradiction is to recognize that any
non-static (running with a finite rate) process means equilibrium disturbance
and is, therefore, irreversible. The acknowledgment of the fact that any non-
static (running with finite rate) process involves the equilibrium disturbance
and thus is irreversible was a turning point in the logical structure of
thermodynamics. That demanded, as will be shown hereinafter, to negate
the first law of thermodynamics as based on the energy balance equation
and to seek for other ways to substantiate the law of energy conservation.
Being though somewhat previous, we can note that the solution to that
problem was found by construing energy as the function of state for a
spatially heterogeneous system and through its representation in terms of the
parameters of that state without respect to the character of the processes in
the system. As a result, all the remainder information about an object under
investigation including the equation of its state and the kinetic equations of
the processes running therein may be successfully attributed to the
uniqueness conditions that thermodynamics imports “from outside” when
applied to solving particular problems. In thermodynamics so constructed
all the assumptions an investigator imposes on the uniqueness conditions
(including the hypotheses on matter structure and process molecular
mechanism, which simplify the preconditions for the equations of state and
laws of transfer) do not affect the core of the theory itself, viz. those
relations which follow from the mathematical properties of energy and other
characteristic functions of system state.
Such a construction of thermodynamics is advisably to be started off
with a notion of action introduced into thermodynamics long before the law
of energy conservation was discovered. The action in mechanics is
construed as something that causes the momentum variation Мdvo, where М
– mass of the system, vo – velocity of the mass center. According to the
laws of mechanics the action value is expressed by the product of the force
F and the duration of its action dt. This value is also called the impulse of
force, N·s. A mechanical action is always associated with state variation, i.e.
with process. Generalizing this notion to non-mechanical forms of motion

18
the action will be construed as a quantitative measure of a process
associated with overcoming some forces. The product of the action and the
moving velocity v = dR/dt of the object the force is applied to characterizes
the amount of work W, J. The notion of work came to thermodynamics from
mechanics (L. Carnot, 1783; J. Poncelet, 1826) where it was measured by
the scalar product of the resultant force vector F and the induced
displacement dR of the object it was applied to (radius vector R of the force
application center)

đW = F·dR (1.3.1)

Thus work was considered as a quantitative measure of action from one

body on another1. Later on forces were called mechanical, electrical,
magnetic, chemical, nuclear, etc depending on their nature. We will denote
the forces of the ith kind by Fi – according to the nature of this particular
interaction form carrier. Forces are additive values, i.e. summable over the
mass elements dM, bulk elements dV, surface elements df, etc. This means
that in the simplest case they are proportional to some factor of their
additivity Θi (mass М, volume V, surface f, etc) and accordingly called
mass, bulk, surface, etc forces. Forces are also subdivided into internal and
external depending on whether they act between parts (particles) of the
system or between the system and surrounding bodies (the environment).
However, when considering non-equilibrium and, in particular, spatially
heterogeneous media, another property of forces takes on special
significance, viz. availability or absence of their resultant F. To clarify what
conditions this availability or absence, it should be taken into consideration
that from the positions of mechanics the work some force does is the only
measure of action from one body (particle) on another. The forces of the ith
kind generally act on the particles of different (the kth) sort and hierarchical
level of matter (nuclei, atoms, molecules, cells, their combinations, bodies,
etc) possessing this form of interaction. Denoting the radius vectors of these
elementary objects of force application by rik and the “elementary” force
acting on them by Fik gives that any ith action on a system as a whole is
added of elementary works

đWik = Fik ·drik (1.3.2)

done on each of them (đWi = Σk Fik · drik ≠ 0).

1
Note that according to the dominating scientific paradigm only the interaction
(mutual action) of material objects exists so that work is the most universal measure
of their action on each other.

19
 The result of such action will evidently be different depending on the
direction of the elementary forces Fik and the displacements drik they cause.
Let us first consider the case when the elementary forces Fik cause the like-
sign displacement drik of the objects of force application (particles of the kth
sort), i.e. change the position of the radius vector Ri for the entire set of the
kth objects the elementary forces Fik are applied to. In such a case dRi = Σk
drik ≠ 0 and the forces Fik acquire the resultant Fi = ΣkFik. This is the work
done by mechanical systems and technical devices (machines) intended
for the purposeful energy conversion from one kind into another.
Therefore in technical thermodynamics such a work is usually called
useful external or technical (A.I. Andryushchenko, 1975, et al). However,
since in the general case such a work is done by not only technical
devices, but biological, astrophysical, etc systems as well, we will call it
just the ordered work and denote by Wе. The work of the ith kind is defined
as the product of the resultant Fi and the displacement dRi it causes on the
object of its application:

đWiе = Fi⋅dRi . (1.3.3)

In Chapter 3 we will make certain that this work definition is universal.

The ordered work process features its vector character.
The work done at the uniform compression or expansion of a gas with
no pressure gradients ∇р therein is another kind of work. Considering the
local pressure p as a mechanical force acting on the vector element of the
closed surface df in the direction of the normal and applying the gradient
theorem to the pressure forces resultant Fр, gives:

Fр = ∫ pdf = ∫ ∇p dV = 0. (1.3.4)

Thus the uniform compression work on an equilibrium (spatially

homogeneous) system is not associated with the pressure forces resultant
to be overcome, while the compression or expansion process itself is not
associated with changing the position of the body as a whole. From the
standpoint of mechanics where work has always been understood as an
exclusively quantitative measure of energy conversion from one form into
another (e.g., kinetic energy into potential one) this means that at the
uniform compression the energy conversion process itself is absent. Due
to the absence of the ordered motion of the ith object (its displacements
dRi = 0) the work of such a kind will be hereinafter called unordered and
denoted by Wн. This category should also include many other kinds of
work not having a resultant, in particular, the work of uniformly
introducing the kth substances (particles) or charge into the system,
imparting relative motion momentum to the system components, etc. This

20
category should further include heat exchange that is nothing but “micro-
work” against chaotic intermolecular forces. As will be made certain
hereinafter, the absolute value of the specific unordered forces Fi/Θi is
construed as the generalized potential Ψi (absolute temperature T, pressure
p, electrical φ, chemical μk potential of the kth substances, etc). Thus the
unordered work is done against forces not having a resultant. Therefore
the unordered work process features the scalar character characterizing the
transfer of energy in the same form (without energy conversion). This is
the situation we encounter at the equilibrium heat or mass exchange and
uniform cubic strain.
The work of dissipative character Wd constitutes a special work
category. This work is done by the ordered forces Fi against the so-called
dissipative forces not having a resultant because of their chaotic
directivity. Therefore the dissipative work features a mixed (scalar-vector)
character, i.e. is associated with changing from ordered forms of energy to
unordered ones.
Fig. 1.1 illustrated a work classification based on the force difference.
energy conversion available in the ordered work processes is here indicated
by superseding the subscript i by the subscript j = 1, 2, …, n according to
the nature of the forces being overcome. External work done (against
environmental forces) is denoted by the superscript env. This work involves
transferring a part of energy in a modified form to other bodies
(environment). Internal work keeps the energy of the system invariable and
involves its conversion from one form into another (as it occurs in
oscillation processes or cyclic chemical reactions of Belousov-
Zhabotinsky’s type). Hereinafter this classification will underlie the
classification of energy by its forms.

Work
of the ith kind

Ordered Unordered Dissipative

env
dRi ↔ dRj drik ↔ drjk dRi → drjk

External Internal External Internal

env env
dRi ↔ dRj dRi ↔ dRj dRi → dRj dRi → dRj

Fig.1.1. Work Classification for Non-Equilibrium Systems

The term heat in the present-day technical literature is used in two

meanings: as a state function (called briefly the body heat) and as a process
function serving as a quantitative measure of heat exchange (and called

21
briefly the process heat)1). This duality in construing heat appeared
historically with considering heat as a chaotic form of motion (amongst
such phenomena as light, sound, electricity, magnetism) and has remained
notwithstanding multiple discussions. The conception of heat as a form of
energy has been reflected in the notion of heat capacity of system. It has as
well strengthened its position in the heat exchange theory (to the principle: a
system can exchange only what it has). In non-equilibrium systems such an
understanding of heat is dictated by a number of thermal effects caused by
dissipation (friction, diathermic or induction heating, chemical
transformations, etc). These heats are not supplied from outside either,
though relate to process. However, in equilibrium systems of such a kind
thermal effects are absent and heat becomes just a quantitative measure of
heat exchange process. Therefore in equilibrium thermodynamics heat is
often interpreted as the energy being transited from one body to another, i.e.
something that is supplied from outside across the system borders, but not
contained in the system itself.
Accepting the said duality for objectivity we will take into consideration
both the body heat and the process heat denoting the former by UB (to avoid
mishmash), while the latter – by Q and applying the exact differential sign
d for infinitesimal increments of any state function (including UB), while
the inexact differential đ – for the elementary heat amounts đQ as process
functions (C. Neuman, 1875). The same inexact differential sign will be
applied for also the work đW when it becomes dependent on the process
path (i.e. becomes process function).
A specific kind of the energy exchange existing in the general case of
open systems and associated with the substance (mass) exchange compels
us to completely refuse the classic division of the energy exchange in such
systems into heat exchange and work. The point is that putting substance
into material medium always involves the so-called “input work” and the
interchange of internal heat energy (body heat) between bodies. Therefore
the notions of heat and work lose their sense “on the border where substance
diffusion takes place” (M. Tribus, 1970).
The impossibility to reduce the process heat to only “one of the forms of
energy exchange” (K. Putilov, 1971), as well as the existence of only one
kind of energy exchange (mass exchange) in open systems forces absolutely
rejecting the classic division of energy exchange into heat and work in non-
equilibrium systems. The fundamental difference between the ordered and
unordered work in non-equilibrium systems being perceived, heat exchange
needs to be ascribed to the category of unordered work (against forces not
having a resultant like the work of dissipative character). Interpreting the

1)
Thermodynamics. Terminology // under the editorship of I.I. Novikov. M.: AN
USSR, 1973. Edition 85

22
heat exchange as some micro-work against intermolecular forces directed in
random way means the above notions are realized as different in their scale,
while work is realized as a unitary quantitative measure of action from some
material objects on other ones.
The abovementioned order of concepts clarifies the meaning and
position of the notion of energy. The term energy (from Greek activity) was
introduced into mechanics in the early XIX century by T. Young, an
authoritative physicist, as a substitution for the notion of living force and
meant the work which a system of bodies could do when decelerating or
going over from a particular configuration into the “zeroth” one (adopted
for the base). The energy was accordingly divided into kinetic Ек и
потенциальную Еp. The term potential meant that the energy could be
realized in the form of work only with appearing the relative motion of
interacting bodies, i.e. with changing their mutual position. The sum of
kinetic and potential energies in an isolated (closed) system did not remain
constant because of a known phenomenon of energy dissipation caused by
unordered work done against the dissipation (friction) forces. Because of
dissipation the real systems (with friction) spontaneously lost their
capacity for external work. That meant the only thing, viz. the transition of
energy as a microscopically ordered form of motion into the latent
(microscopic) form of motion (interaction). Later, with thermodynamics
appeared, that standpoint was supported by proving the internal energy U
as inherent to bodies. That allowed stating the law of conservation of total
energy that was construed as the sum of kinetic Ек, potential Еp and
internal U energies of an isolated system:

(Ек + Еp + U)isol = const. (1.3.5)

However, in that case the notion of energy lost its primary sense of the
capacity for external work as ensued from the word-group etymology of
Greek εν (en) for external and εργον (ergon) for action. Indeed, according
to the second law of thermodynamics the internal energy U can not be
entirely converted into work. For this reason the energy of real systems
ceased to be determined by the amount of useful work done. And the work
itself ceased to be the exact differential since became dependant on the
process path and rate (dissipation conditioning), but not on exclusively the
initial and final states of the system. To simplify the situation, mechanics
was supplemented with a provisional notion of conservative system, where
the sum of kinetic and potential energies could be considered as a value kept
and dependant on exclusively the initial and final states of the system.
However, in that case all the consequences of mechanics as having ensued
from the energy conservation law were naturally restricted to only the
conservative systems.
That engendered some ambiguity in the notion of energy, which has not
yet been resolved. A reader is usually very surprised with not finding in

23
physical guides and encyclopedia a definition of this notion more substantial
than the philosophic category of general quantitative measure of all kinds of
matter. As H. Poincaré bitterly noted, “we can say of energy nothing more
but that something exists remaining invariable”. Regarding the value that
brings together all phenomena of the surrounding world such a definition is
absolutely insufficient, the more so because not only energy remains
invariable in isolated systems, but also mass, momentum, charge and
angular momentum!
The definition thermokinetics offers for work through action and work
interpreted as the only quantitative measure of action from some bodies on
other ones allows returning to energy its simple and clear meaning as the
capacity of a system to do work. However, now energy becomes a
quantitative measure of all (ordered and unordered, external and internal,
useful and dissipative) works a system can do. This approximates to the J.
Maxwell’s definition of energy as the “sum of all effects a system can have
on the surrounding bodies”. Next chapters will be dedicated to the
substantiation of formal consistency and advisability of such an approach.

1.4. Number-of-Degrees-of-Freedom Theorem as a Basis

of Real Processes Classification

Changing over to non-equilibrium systems with spontaneous processes

running therein needs to generalize the thermodynamic principle of process
classification itself. The point is that the same state variations (e.g., heating
of a body) in spatially heterogeneous systems may be caused by both the
external heat exchange and appearing internal friction heat sources,
chemical reactions, diathermic heating, magnetization reversal, etc. In the
same way the cubic strain of a system can be induced by not only the
compression work, but a spontaneous expansion into void as well. Hence
processes in thermokinetics should be classified regardless of what causes
whatever state variations – the external heat exchange or internal (including
relaxation) processes.
In this respect thermokinetics differs from both physical kinetics that
classifies processes by reasons causing them (distinguishing, in particular,
concentration diffusion, thermal diffusion and pressure diffusion) and the
heat exchange theory that distinguishes processes by the mechanism of
energy transfer (conductive, convective and radiant). Processes in
thermokinetics will be classified by their consequences, i.e. by special state
variations they cause as phenomenologically distinguishable and
irreducible to others. We will call such processes, for short, independent.
These include, in particular, isochoric, isobaric, isothermal and adiabatic
processes thermal physics considers. Here comes the heat process as well
(K. Putilov, 1971), which we will construe as a variation of the body

24
internal thermal energy UB regardless of what causes it – either heat
exchange or internal heat sources. Other processes are also included, e.g.,
the system composition variation process that may be caused both by
substance diffusion across the system borders and chemical reactions
inside the system.
With the principle of process classification adopted as based on
distinguishability of processes specific demands are made on choosing their
“coordinate”, i.e. a physical value which variation is the necessary and
sufficient criterion of running a particular process. These demands consist
in choosing only such a parameter as the process coordinate that does not
vary, when other, also independent, processes are simultaneously running in
the same space points. It is that approach wherefrom the requirement in
classic thermodynamics follows for the invariability of entropy as the heat
exchange coordinate in adiabatic processes as well as the requirement itself
for the process reversibility, i.e. the absence of spontaneous entropy
variations not connected with the external heat exchange.
The principle of classification of real processes by their consequences
and the axiom of their distinguishability enable substantiating a quite
evident though fundamental statement stipulating that the number of
independent coordinates conditioning the state of any (either equilibrium or
non-equilibrium) thermokinetic system equals the number of degrees of its
freedom, i.e. the number of independent processes running in the system.
This statement (or theorem) is easily provable “by contradiction”. Since
a thermodynamic process is construed as variation of the properties of a
system expressed in terms of state parameters, at least one of such
parameters necessarily varies when processes are running. Let’s assume that
several state parameters necessarily vary when some independent process is
running. Then these parameters will not evidently be independent, which
violates the primary premise. Now let’s assume that some coordinate of the
system necessarily varies when several processes are running. Then these
processes will not evidently be independent since they cause the same
variations of the system properties – the fact that also violates the primary
premise. We have nothing to do, but to conclude that only one independent
state coordinate corresponds to any (equilibrium or non-equilibrium, quasi-
static or non-static) independent process. Such coordinates are generally
extensive variables since each of them defines, in the absence of other
degrees of freedom, the energy of a system, which is an extensive value as
well.
The proven statement defines the necessary and sufficient conditions for
unique (deterministic) definition of state for whatever system. Therefore, it
may be, for ease of reference, reasonably called the state determinacy
principle. This principle makes it possible to avoid both the “under-

25
determination” and “over-determination” of a system1) as the main cause of
the methodological errors and paradoxes of present-day thermodynamics
(V. Etkin, 1979, 1991). The continuum state “under-determination” as
resulting from the local equilibrium hypothesis adopted is, e.g., far from
evidence. This hypothesis excludes the necessity of the gradients of
temperature, pressure and other generalized potentials (i.e. thermodynamic
forces) in the fundamental equation of non-equilibrium thermodynamics on
the ground that the bulk elements are assumed to be equilibrium. On the
other hand, the continuum “over-determination” due to the infinite number
of degrees of freedom ascribed to it despite the finite number of macro-
processes running therein is either not evident.
The theorem proven allows, in its turn, to concretize the notion of
system thermokinetic state, which is construed as a set of only such
properties that are characterized by the set of state coordinates strictly
defined in their number. This means that such system properties as color,
taste, smell, etc, which are not characterized by state parameters
quantitatively and qualitatively may not be considered as thermodynamic.
This relates, in particular, to also the “rhinal”, “haptic”, etc number of
freedom A. Veinik arbitrarily introduced (1968) for a system.
One of the consequences of the determinacy principle consists in the
necessity to introduce additional state coordinates for systems where, along
with external heat exchange processes, internal (relaxation) processes are
observed as tending to approximate the system to the equilibrium state.
Without such variables introduced it is impossible to construct a theory
covering the entire spectrum of real processes – from quasi-reversible up to
critically irreversible.

1.5. Extension of Variables Space with Introduction of Spatial

Heterogeneity Parameters for System as a Whole

The fact that relaxation vector processes (temperature, pressure,

concentration, etc equalization) run in non-equilibrium systems requires
introducing specific parameters of spatial heterogeneity characterizing the
state of continuums as a whole. To do so, it is necessary, however, to find a
way how to change over from the density (fields) distribution functions

ρi = dΘi/dV

of some extensive physical values Θi to the parameters of the system as a

whole, which thermodynamics operates with. This change may be

1)
I.e. the attempts to describe the system state by a deficient or excessive number
of coordinates.

26
conducted in the same way as used in mechanics to change over from
motion of separate points to system center-of-mass motion. To better
understand such a change, let us consider an arbitrary continuum featuring
non-uniform density distribution ρi = ρi(r,t) of energy carriers1) over the
system volume V. Fig.1.2 illustrates the arbitrary density distribution ρi(r,t)
as a function of spatial coordinates (the radius vector of a field point r) and
time t. As may be seen from the figure, when the distribution Θi deviates
from that uniform (horizontal line), some amount of this value (asterisked)
migrates from one part of the system to other, which displaces the center of
this value from the initial Ri0 to a current position Ri.
Position of the center of a particular extensive value Θi defined by the
radius vector Ri is given by a known expression:

Ri = Θi-1 ∫ ρi(r,t) rdV , (i = 1,2,…,n) (1.5.1)

For the same system, but in a homogeneous state, the Θi center position
Ri0 may be derived if factoring ρi = ρi0(t) in equation (2.1.1) outside the
integral sign:

R i 0 = Θi−1 ∫ ρi 0 ( t ) rdV = V −1 ∫ rdV . (1.5.2)

V V
Thus the state of a heterogeneous system features the emergence of
specific “distribution moments” Zi of the energy carriers Θi:

ρi Zi = Θi(Ri – Ri0) =
Ψi = ∫ [ρ i (r, t ) − ρ i 0 (t )]rdV . (1.5.3)
Xi V
The electrical displacement
θ* ρi(t) vector D = Θе∆Rе is one of such
ΔRi moments with Θе as electrical
Ri ρi(r,t)
Riо charge and ∆Rе as displacement
of its center.
r
Expression (1.5.3) most
evidently manifests that the
Fig.1.2. To Generation of Distribution parameters Zi of spatial
Moment heterogeneity are additive values
and summed up providing the
ρi0(t) value remains the same in various parts of a heterogeneous system.

1)
The Energy carrier is construed as a material carrier of the ith Energy component, which
quantitative measure is the physical value Θi.. So the mass Mk of the kth substance is a carrier
of the rest Energy; the charge Θe – a carrier of the electrostatic Energy of the system; the
component momentum Mkvk – a carrier of its kinetic Energy, etc.

27
This follows from the conservation of integral (1.5.3) at its partition into
parts with a volume V’ < V1). However, these parameters become zero at
“contraction” of the system to a material point, when ρi(r,t) → ρi0(t). This
stands in absolute conformity with the degrees-of-freedom theorem because
the processes of density redistribution ρi(r,t) are absent in material points.
And once again this confirms the fact that an entity of continuum elements
considered as a system, non-equilibrium in whole, possesses additional
degrees of freedom.
For any part of a homogeneous isolated system the Ri0 value remains
unvaried since running of any processes is herein impossible. Therefore the
Ri0 may be accepted for such systems as a reference point r or ri and set
equal zero (Ri0 = 0). In this case the vector Ri will define a displacement of
the Θi center from its position for the system being in internal equilibrium
state, and the moment of distribution of a particular value Θi in it will
become:

Zi = ΘiRi (1.5.4)

Herein the moment Zi becomes an absolute extensive measure of the

system heterogeneity with respect to one of the system properties – like
such absolute parameters of classic thermodynamics as mass, volume,
entropy, etc.
As follows from expressions (1.5.2) and (1.5.3), the distribution
moment Zi emerges due to exclusively the displacement of the Θi center and
has nothing to do with the variation of this value itself. Thus the expression
for the exact differential of the function Zi = Zi(Θi,Ri) becomes:

dZi = (∂Zi/∂Θi)dΘi + (∂Zi/∂Ri)dRi , (1.5.5)

resulting in:

Θi = ∇⋅Zi and ρi = ∇⋅ZiV , (1.5.6)

where ZiV = ∂Zi/∂V – distribution moment in the system unit volume.

In case of discrete systems the integration over system volume will be

replaced by the summation with respect to elements dΘi of the Θi value:

Zi = ΘiRi = Σ i ri dΘi , (1.5.7)

1)
With symmetrical density ρi(r,t) distributions for whatever parameter, e.g., fluid-velocity
profiles in tubes, expression (1.5.3) should be integrated with respect to annular, spherical,
etc layers with V′ > 0, wherein the function ρi(r,t) is monotone increasing or decreasing.

28
where ri – radius vector of the element dΘi center. Therefore expressions
(1.5.4) through (1.5.6) remain valid for also the systems with discrete
distribution of charges, poles, elementary particles, etc. Only the
geometrical meaning of the ∆Ri vector changes; for symmetrical
distributions the vector is defined by the sum of the displacements ∆ri of all
elements dΘi. This may be instantiated by the centrifugal “shrinkage” of the
particles’ momentum flow in moving liquid when forming turbulent or
laminar fluid-velocity profiles in channels (“boundary layer” formation and
build-up).
Explicitly taking into account the spatial heterogeneity of systems under
investigation is decisive in further generalization of the thermodynamic
investigation method to non-equilibrium systems. As a matter of fact, this is
the spatial heterogeneity (heterogeneity of properties) of natural objects that
causes various processes running in them. This implies the exclusive role
the distribution moments Zi play as a measure for deviation of a system in
whole from internal equilibrium of the ith kind. Introducing such parameters
allows precluding the major drawback of non-equilibrium thermodynamics,
viz. lack of extensive variables relating to the gradients of temperature,
pressure, etc. Classic thermodynamics is known to have crystallized into an
independent discipline after R. Clausius succeeded in finding a coordinate
(entropy) related to temperature in the same way as pressure to volume and
thus determinately described the simplest thermo-mechanical systems. The
distribution moments Zi play the same part in thermokinetics coming into
being. As will be shown later, these relate to the main parameters introduced
by non-equilibrium thermodynamics – thermodynamic forces, in the same
way as the generalized potentials to the coordinates in equilibrium
thermodynamics. These are the distribution moments which make the
description of heterogeneous media a deterministic one thus enabling
introducing in natural way the concept of generalized velocity of some
process (flow) as their time derivatives. They visualize such parameters as
the electrical displacement vectors in electrodynamics and generalize them
to phenomena of other physical nature. In mechanics the Zi parameters have
the dimension of action (Θi – momentum of a body, Ri – its displacement
from equilibrium position), imparting physical meaning to this notion.
These are the parameters which allow giving the analytic expression to the
system working capacity having thus defined the notion of system energy.
Using such parameters provides a clear view of the degree of system energy
order, enables proposing a universal criterion of the non-equilibrium system
evolution, etc. Paraphrasing a M. Planck’s statement regarding entropy one
may positively say that the distribution moments are exactly the parameters
entire non-equilibrium thermodynamics is “standing and falling” with.

29
 1.6. Coordinates of Non-Equilibrium Redistribution
and Reorientation Processes

The moments of distribution (1.5.5) contain vectors of displacement Ri,

each of which can be expressed product of a basic (individual) vector еi,
characterising its direction, on module Ri = |Ri | this vector. Therefore the
complete variation of the displacement vector Ri may be expressed as the
sum of two summands:

dRi = еidRi + Ridеi , (1.6.1)

where the augend еidRi = dS characterizes elongation of the vector Ri, while
the addend Ridеi – its turn.
Let us express now the dеi value characterizing the variation of the
distribution moment direction in terms of an angular displacement vector φ
normal to the plane of rotation formed by the vectors еi and dеi. Then the dеi
will be defined by the external product dφi×еi of vectors dφi and еi, so the
addend in (1.6.1) will be ΘiRidеi = dφi×Zi. Hence, expression of full
differential of the distribution moments looks like:

dZi = (∂Zi/∂Θi)dΘi + (∂Zi/∂Si)dSi + (∂Zi/∂φi)dφi. (1.6.2)

According to the degrees-of-freedom theorem this means that any state

function describing a heterogeneous system in whole are generally defined
by also the full set of variables Θi, Si and φi. Since further resolution of the
vector Zi is impossible, expression (1.5.5) indicates there are three
categories of processes running in heterogeneous media, each having its
own group of independent variables. The first-category processes running at
Ri = const involve the uniform variation of the physical value Θi in all parts
of the system. Such processes resemble the uniform rainfall onto an
irregular (in the general case) surface. Here comes, in particular, the
pressure field altered in liquid column with variation of free-surface
pressure. These processes also cover phase transitions in emulsions,
homogeneous chemical reactions, nuclear transformations and the similar
scalar processes providing the composition variations they induce are the
same in all parts of the system. We will call them hereinafter the uniform
processes regardless of what causes the increase or decrease in amount of
whatever energy carrier Θi (and the momentum associated) – either the
external energy exchange or internal relaxation phenomena. These
processes comprise, as a particular case, the reversible (equilibrium)
processes of heat exchange, mass exchange, cubic strain, etc, which, due to
their quasi-static nature, practically do not disturb the system spatial
homogeneity.

30
 Processes described by the addend in (1.5.5) run with the Θi parameters
being constant and consist in their redistribution among the parts (zones) of
a heterogeneous system. They involve decreasing, e.g., the entropy S΄, mass
М΄, its momentum Р΄, its volume V΄, etc, in some parts of the system and by
increasing the same in other parts. Such processes are associated with the Θi
value center position variation Ri within the system and resemble the
migration of fluids from one part of a vessel into another. Therefore we will
call them the redistribution processes. Such processes are always non-
equilibrium even if they run infinitely slowly (quasi-statically) since the
system remains spatially heterogeneous in this case. State modifications of
such a kind are caused by, e.g., the useful external work of external forces,
the non-equilibrium energy exchange processes that induce non-uniform
variation of the Θi coordinates inside the system, and the vector relaxation
processes involving equalization of temperature, pressure, chemical and
other system potentials. All processes of such a kind feature a directional
(ordered) character, which distinguishes the useful work from the work of
uniform (quasi-static) introduction of substance, charge, etc, or the
expansion work. According to (1.3.2) the coordinates of the processes
pertaining to this category are understood as the displacement vectors Si.
These coordinates should be attributed to the external parameters of the
system since they characterize the position of the energy carrier Θi center in
whole relative to external bodies (the environment) just as the center of
mass Sm of the system or its center of inertia Sw.
There are also the processes of reorientation of magnetic domains,
electrical and magnetic dipoles, axes of rotation of bodies, etc., running in a
number of systems, e.g., in ferromagnetic materials. The micro-world
manifests them in, e.g., the unified spin-orientation arrangement’ the macro-
systems – in the spontaneous magnetization of ferromagnetic materials,
while the mega-world – in the close-to-equatorial plane alignment of the
galaxies’ spirals, asteroidal belts, orbits of the primary planets and their
satellites, etc. The systems with processes of such a kind will hereafter be
called, for short, oriented. These include also the bodies with shape
anisotropy. The reorientation processes are not reducible to the transfer and
redistribution processes either. This means that the coordinate of such kind a
process is a variation of the angle φi characterizing the orientation of
distribution moment Zi of the system as a whole.
Thus, all processes running in heterogeneous systems may be broken
down into three groups: uniform, redistribution and reorientation processes,
which coordinates are, respectively, variables Θi, Si and φi. This
fundamentally distinguishes thermokinetics from classic thermodynamics
and the theory of irreversible processes, where the state of a system is
defined by exclusively a set of thermostatic variables Θi.
The undertaken expansion of the space of variables by introducing the
vectors of displacement Ri makes it possible to cover not only quantitative,

31
but as well qualitative variations of energy in various forms. The fact that
vector processes run in systems along with scalar processes means that both
the ordered We and unordered Wun works are generally done in such
systems. It becomes clear that the irreversibility of real processes associated
with the energy dissipation (i.e. with losing the capacity for ordered work)
becomes apparent in the process scalarization, i.e. in losing vector character
of the process. Furthermore, a possibility appears to further distinguish
between the energy transfer processes (i.e. the energy transfer between
bodies in the same form) and the energy conversion processes (i.e. the
energy conversion from one form into another)1).
This is enough in principle to construct a unitary theory of real
processes enabling investigation of any systems (simplex and complex,
closed and open, homogeneous and heterogeneous, isolated and non-
isolated, tending to and omitting equilibrium) not outstepping the strict
applicability of its primary concepts.

Chapter 2
Specificity of Thermokinetics in Application to Spatially
Heterogeneous Systems

As the practice shows, studying a scientific discipline, especially

acquiring it inasmuch as necessary for an independent activity is most
successful if a reader is from the very first introduced to a critical
presentation of the existing situation, to fundamentals of the discipline and
general methods of solving particular problems. Therefore in this chapter
we will review, along with a historical synopsis, specific features of
thermokinetics allowing it to apply its body of mathematics to any systems
– simplex and complex, closed and open, homogeneous and heterogeneous,
tending to and omitting equilibrium in their development.

1)
As will be shown hereinafter, the Energy transfer is associated with unordered work done,
whereas the Energy conversion – with ordered work.

32
 2.1. Substantiation of Total energy Conservation Law

Classic thermodynamics is known to be based on the principle of heat Q

and work W equivalence. R. Clausius, the founder of classic
thermodynamics, formulated this principle as follows, ”In all cases, when
heat becomes work in a cyclic process, the amount of the heat expended is
proportional to the work done and vice versa, work done is converted into
an equivalent amount of heat” (Clausius, 1876). If heat and work are
measured in the same units of the international system of units, SI, the
equivalence principle may be written as a simple relationship:

Wc / Qc = ∫ đW/ ∫ đQ = 1, (2.1.1)

где Wc, Qc – work done and heat supplied for cycle; đW, đQ – their
elementary amounts for particular parts of the cyclic process under
consideration.
Taking into consideration the rule of signs accepted in thermodynamics
(the work done by a system and the heat supplied to it are positive) equation
(2.1.1) becomes:

∫ (đQ – đW) = 0. (2.1.2)

Clausius was the first who noticed that the above result did not depend
on the path of the process under consideration. That allowed him to use a
known mathematical theorem of curvilinear integrals. It states that if a
curvilinear integral of an arbitrary differential form (in our case đQ – đW)
becomes zero along any closed path within some space of variables, the
integrand represents the exact differential of function of these variables U:

dU = đQ – đW or đQ = dU + đW. (2.1.3)

R. Clausius did not concretize the space of variables wherein he

considered the curvilinear integral (2.1.2) since he had not yet found the
heat exchange (entropy) coordinate. Therefore he initially called the
function U the total heat of a body having construed it as the sum of the heat
the body received from outside and the heat released as a result of the
disgregation work (of dissipative character). That caused some confusion in
notions since imparted the status of state function to heat and disgregation
work. Therefore a rather heated discussion combusted about the U function.
In particular, W. Thomson recommended the term mechanical energy of a
body in particular state for the U value. From that time on this function has

33
been referred to as the internal energy. Being the state function of a system,
that function did not depend on the motion or position of the system relative
to the environment. In such a case the isolation of the system from the
environment (Q, W = 0) left that function invariable. Based on that fact,
expression (2.1.3) started to be considered in classic thermodynamics as a
particular case of the energy conservation law called the first law of
thermodynamics.
Since classic thermodynamics from its origin has always been restricted
to describing the behavior of internally equilibrium (spatially homogeneous)
systems with parameters the same for all of the system parts, all kinds of
work such a system could do had the unordered character (see Chapter 1).
Here comes, in particular, the uniform compression work đWр expressed
through the product of the absolute pressure p and the volume variation dV. In
such a case the U function could actually be construed as the dissipated part
of energy, which corresponded to the law of energy conservation in the form
of (1.3.4). However, in the more general case of non-equilibrium and
especially spatially heterogeneous systems this is far from being so. In
particular, the chemical and nuclear energy of homogeneous systems is also
partly convertible into other forms despite they do not either depend on the
position of the system relative to surrounding bodies, i.e. relate to internal
energy. The situation became even more complicated with changing to the
spatially heterogeneous systems to be studied, in particular, to the so-called
extended systems with the environment included. Such systems can do some
useful work before internal equilibrium has set in there. These systems may
comprise also polarized and magnetized bodies located in external force
fields. In all those cases the internal energy U ceased being that “dissipated”
part of energy meant in (1.3.4).
All this impels to search for a more general substantiation of the law of its
conservation. To this end let us consider the results of those experiments on
definition of the heat and work equivalence principle, which related to non-
equivalent systems with relaxation processes running therein. Their specific
character was such that heat was obtained there by friction (dissipation).
These include classic experiments by Joule, in particular, the experiment with
calorimeter and agitator driven by dropping weight; also his experiment with
the Proney brake that brakes the drum calorimeter (1843-1878); Girn’s
experiments with lead flattening on anvil with drop hammer (1859); Lenz’s
experiments with solenoid discharging to active resistance in vessel
calorimeter (1972) and many other experiments involving battery charging,
gas transfer between vessels, electrolyte decomposition, etc (J. Gelfer,
1969,1973). Those experiments had such a result that a system disturbed from
equilibrium by a mechanical (ordered) work Wiе= Wiе(Ri) done on it returned
to the initial equilibrium state after a heat amount Q strictly equivalent to the
work had been removed from the system. Taking the work of both ordered
and unordered character (according to the above classification) into

34
consideration means the necessity to extend the space of variables wherein the
above considered cyclic processes take place. It is easy to reveal in this case
that integrand (2.1.2) constitutes a state function in the space of variables (Θi,
Ri), i.e. a more general one than the system internal energy. This function
depends on both the internal Θi and external Ri system coordinates, i.e.
constitutes the sum of the external and internal system energies. Such energy
is usually called the total energy of the system. The decrease of the function
U=U(Θi, Ri) defines the sum of all (ordered and unordered) works the system
do:

– dU(Θi, Ri) = Σi đWiн + Σi đWiе . (2.1.4)

According to this expression the energy of a system, in the absence of

external impacts on the system (ΣiđWiн = 0, ΣiđWiе = 0), remains invariable
at any variations of its state. In other words, the energy of an isolated
system is constant. Thus the generalization of the heat and work
equivalence principle to non-equilibrium systems directly leads to the law
of conservation of “total” energy as a state function for the entire set of
interacting (mutually moving) bodies. However, for such a system
(isolated) all its energy is internal. This fact reveals the imperfection of
dividing the energy into external and internal. From the position of
thermokinetics considering the entire set of interacting bodies as a single
non-equilibrium whole it is more important that its energy be measured in
an own (absolute) reference frame not connected with the state of any of
the bodies within the environment1). Since the term system energy with
regard to the function U(Θi,Ri) unambiguously tells the energy belongs to
the system itself, the terms “total”, “external”, “internal”, etc energies
become superfluous. This allows focusing on other properties of energy
and its other components which characterize its conversion capacity.

2.2. Cancellation of Dissipative Terms in energy Balance Equations

Let us consider the consequences ensuing from the fact itself of

existing the system energy U(Θi, Ri) as a function of the quite certain set of
arguments (state coordinates) As shown above, the energy of a
heterogeneous system as a function of its state is generally expressed as U =
U [Zi(Θi,Si,φi)], where i = 1, 2, …, n – number of energy components equal

1)
Otherwise, should the Energy conservation be violated, the Energy of a system would vary
with the state variation of these bodies despite the absence of Energy exchange with the
system.

35
to the maximal number of independent processes for some of their
categories (uniform processes, redistribution and reorientation processes).
This means that the exact differential of energy may be expressed by the
following relationship:

dU = Σi(∂U/∂Θi)dΘi + Σi (∂U/∂Si)dSi + Σi (∂U/∂φi)dφi. (2.2.1)

Derivatives of some system parameters (U) with respect to other ones

(Θi, Si, φi) are also system parameters. Therefore denoting them as:

Ψi ≡ (∂U/∂Θi); (2.2.2)
Fi ≡ – (∂U/∂Si); (2.2.3)
Мi ≡ – (∂U/∂φi), (2.2.4)
gives the fundamental identity of thermokinetics in the form:

dU ≡ ΣiΨi dΘi – Σi Fi·dSi – Σi Мi·dφi. (2.2.5)

For isolated systems the right-hand member of identity (2.2.5)

becomes zero. For systems not changing its spatial orientation the two last
terms in (2.2.1) and (2.2.6) may be combined (ΘidSi + Zidеi = ΘidRi), then
the fundamental identity of thermokinetics becomes:

dU ≡ ΣiΨi dΘi – Σi Fi·dRi. (2.2.6)

Identities (2.2.5) and (2.2.6) are nothing else but a result of the joint
definition of the related parameters Ψi and Θi, Fi and Si or Zi, Мi and φi. To
clarify the physical meaning of the parameters thus introduced, let us first
consider the particular case of internally equilibrium (spatially
homogeneous) and stationary thermo-mechanical systems. Such simplest
systems may be instantiated as the working media of heat engines in the
vaporous or gaseous state. They have two degrees of freedom – thermal and
mechanical, i.e. the capacity for the heat exchange Q and the uniform
expansion work Wex. Due to the absence of redistribution and reorientation
processes in homogeneous systems (dSi, dφi = 0) the parameters Ψi are the
same for all points of such a system and equal to their local values ψi, so
that identity (2.2.5) goes over into a joint equation of the first and second
laws of thermodynamics for closed systems:

dU = Σi ψi dΘi = TdS – рdV . (2.2.7)

Since the variation of the coordinates Θi in an equilibrium system is

caused by exclusively the external heat exchange (their internal sources are
absent), the terms of this relationship characterize, respectively, the

36
elementary heat exchange in the system đQ = TdS and the elementary
expansion work đWex = рdV. In this case the parameters Ψi acquire the
meaning of the absolute temperature T and absolute pressure p. In the
more general case of spatially heterogeneous systems the parameters Ψi
are, as will be shown hereinafter, the generalized local potentials ψi
averaged by mass in all elements of the system.
To clarify the meaning of the terms of the second sum in (2.2.5), we
must take into account that they correspond to the redistribution processes
running at constant parameters Θi and φi, i.e. with invariable direction of the
unit vector e. Then (∂U/∂Si) = e·(∂U/∂Ri) = Fi·e, i.e. represents a projection
of the resultant Fi on the direction of movement of the object Θi the force is
applied to. Thus the terms of the second sum in (2.2.5) characterize the
elementary useful work đWiе = Fi·dRi that has the same form as in
mechanics. This work may be expressed in terms of the so-called
thermodynamic forces Хi defined as:

Хi ≡ – (∂U/∂Zi). (2.2.8)

Since at Θi, φi = const, (∂U/∂Zi) = Θi–1(∂U/∂Ri) = Fi/Θi, then the Хi

thermodynamic forces thus introduced are actually the specific forces in
their usual (Newtonian) meaning, i.e. the forces Fi per unit of the value Θi
they transfer. These are, in particular, the specific mass, bulk and surface
forces, for which the Θi value is construed as, respectively, mass M, volume
V and surface f of the body. This category also includes the Lorenz force Fе
related to the electric charge Θе transferred. In the theory of irreversible
processes such forces are termed the thermodynamic forces in their energy
representation (Gyarmati, 1974). Using them enables representation of
work by two equivalent expressions:

đWiе = Fi·dRi = Хi·dZi. (2.2.9)

The work described by expression (2.2.9) may be mechanical, thermal,

electrical, chemical, etc (depending on nature of the forces to overcome);
external or internal (depending on where the forces arise – either in the
system itself or outside); useful or dissipative (depending on what the work
involves – either purposeful conversion of energy or its dissipation).
Lastly, the terms of the third sum in (2.2.5) correspond to the
reorientation processes running with constant Θi and Ri. In this case Fi·dRi =
Fi·[dφi,Ri], and the parameter Мi acquires the meaning of a torque from the
force Fi:

Мi = Fi×Ri (2.2.10)

37
 This “torsion” torque is advisable to be called the “orientation” torque
in the case it becomes zero when the direction of the force Fi coincides with
the direction of the displacement vector Ri.
The fundamental identity of thermokinetics thus obtained is valid
regardless of what causes the variation of the parameters Θi, Si and φi –
either the external heat exchange or the internal (including relaxation)
processes. Therefore it is applicable to any processes (both reversible and
irreversible). At the same time it is most detailed of all the relationships
connecting the parameters of spatially heterogeneous systems since it allows
for any possible categories of processes running in such systems.

2.3. Distinction between Convertible and Inconvertible Parts

of energy (inergy and anergy)

Due to the fact that thermokinetics rejects in its grounds the process
idealization expressed in such notions as “quasi-static” (infinitely slow),
“equilibrium” and “reversible” a possibility appears to introduce time as a
logically consistent physical parameter into the equations of thermokinetics.
For that it is enough to rewrite identity (2.2.5) in the form containing total
derivatives of the state parameters earlier introduced with respect to time t:

dU/dt ≡ ΣiΨi dΘi/dt – ΣiFi·vi – Σi Мi⋅ωi . (2.3.1)

Here vi ≡ еidRi/dt – translation velocity of the energy carrier Θi; ωi ≡ dφi

/dt – angular velocity of its reorientation (or rotation). In the particular case,
when the parameter Θi means mass of a system, the values v and ω
characterize its linear and angular velocity as a whole. For the future it is
quite important to obtain the local statement of this identity true for any
element of the continuum. For this purpose let us apply equation (2.2.5) first
to the system where redistribution processes are absent. Then the second
and the third sums in (2.2.5) disappear, and the identity becomes:

dU/dt. ≡ Σi Ψi dΘi /dt. (2.3.2)

In the systems this equation represents the variation of the Θi1)

parameters is caused by exclusively the transfer of some amount of energy
carrier across the system borders. This allows representing the behavior of
these parameters in the time domain by a known expression:

1)
From the physical standpoint the value Θi that is actually the extensive measure of
particular kind interaction (Energy) carrier is advisable to be called for short the Energy
carrier. This will facilitate the understanding of many processes under investigation.

38
 dΘi/dt = – ∫jie·df , (2.3.3)

where jie =ρivi – local density of flow of the energy carrier Θi through a
vector element df of the closed surface f in the direction of external normal
n; vi – velocity of energy carrier transfer through the system surface
element df in stationary reference frame (Fig. 2.1).
Substituting (2.3.3) into (2.3.2) gives:

dU/dt = – Σi Ψi ∫ jie·df. (2.3.4)

This equation is evidently a particular case of the more general

expression

dU/dt = –Σi ∫ψijie·df, (2.3.5)

when the local value ψi of generalized potential Ψi is the same for all system
points and may therefore be factored outside the integral sign. The product
ψiji is the ith component of the energy flow density je = Σi ψiji through an
element df of the system surface f. Therefore changing in (2.3.4) to the
integral taken over system volume as based on the Gauss-Ostrogradsky’s
theorem we come to the expression for the law of energy conservation for
an arbitrary continuum area, which was proposed by N. Umov in 1873:

dU/dt = – ∫∇⋅jedV. (2.3.6)

z According to this expression the

je system energy variation equals the
amount of energy having passed
n across the system borders for that
df
V
particular time. Or according to
Umov himself, “energy flow…is
r
caused by energy admission or
x
release a medium provides across its
borders”. It should be noted that the
validity of this statement is by no
y
means restricted to the mechanical
Fig. 2.1. Energy Flow across System Borders
energy N. Umov meant.
This equation may be developed
by representing the energy flow divergence ∇⋅je = Σi∇⋅(ψijie) as a sum of two
summands Σi ψi ∇⋅jie + Σijie·∇ψi:

dU/dt = – Σi ∫ψi ∇⋅jie dV + Σi∫ xi·jie dV, (2.3.7)

where

39
 xi ≡ –∇ψi (2.3.8)

is a local motive force of the ith process expressed as negative gradient of

generalized potential and named in the theory of irreversible processes as
the “thermodynamic force in its energetic representation” (I. Gyarmati,
1974).
Equation (2.3.7) enables clarifying the meaning of the “global”
variables Ψi and Xi introduced earlier for a system in whole. Taking into
account that volume or mass elements in continuums do not change their
spatial orientation (dφi = 0) identity (2.2.6) may be expressed in the form:

dU/dt ≡ ΣiΨi dΘi/dt – Σi Xi·Ji , (2.3.9)

where

Ji ≡ ΘiеidRi/dt = Θivi , (2.3.10)

i.e. are total flows of displacement (transfer) of the ith energy carrier Θi.

These flows at dφi = 0 may be expressed in terms of their densities ji ≡

ρidri /dt through the evident relationship:

Ji ≡ ΘidRi /dt = ∫(dri /dt)dΘi = ∫ ji dV . (2.3.11)

It is easy to see that the flows Ji differ in their dimensions from the
more usual notion of flow rate and in their meaning as per (2.3.10) are
closer to the “generalized momentum” Рi = Θivi of the ith energy carrier Θi
for a system in whole. Such flows play an important role in many
phenomena. These are, e.g., the vector flows of electric displacement in a
system with the volume V defined by the product of the system free charge
Θе and the velocity of its center displacement in the free charge
redistribution processes. This is the value, to which the following
parameters are proportional: magnetic field induction vector (Biot-Savart
law), Thomson–Joule heats in conductors and thermo-elements,
electromagnetic force driving a conductor with current (Ampere’s law), etc.
We will hereinafter be referring to them time and again when dealing with
the transfer and conversion of energy in any forms, which will confirm the
necessity and usefulness of generalizing the Maxwell’s displacement current
concept to phenomena of other nature.
To find the relation between the “global” (pertaining to a system in
whole) and the local thermodynamic forces xi, let us take into account that
the parameters Ψi in identity (2.3.5) are defined for the coordinates Zi being
constant, i.e. for the difference ρi(r,t) – ρi0(t) invariable in all points of the
system volume V. From this it follows that in the expression

40
 dΘi/dt = ∫ (dθi/dt) ρdV , (2.3.12)

the specific parameters θi vary uniformly in all parts of the system, so that
dθi/dt may be factored outside the integral sign. Hence,

Ψi = М–1 ∫ ψidМ , (2.3.13)

being the system mass-averaged value of the local potential ψi. Similarly
proceeding from the invariance of the process power Ni = Xi·Ji when
representing it in terms of the local and global parameters

Xi = Ji–1 ∫ xi·ji dV, (2.3.14)

gives that the “global” thermodynamic force Xi is some averaged value of

the local thermodynamic force xi ≡ –∇ψi.
The relationship thus obtained between the local variables the field
theories operate with and the thermodynamic parameters characterizing the
state of a continuum in whole opens the possibility of describing their
properties from the positions of thermokinetics. In this case particular
importance is attached to introducing in thermodynamic equations the most
significant for natural science in whole concepts of flows Ji,ji as generalized
rates of the transfer processes and the concept of power (capacity) of the
energy conversion process in a whole system Ni = Xi·Ji and in its unit
volume xi·ji. It should be noted that the notion of capacity refers to only the
useful energy conversion processes and, therefore, could not appear in the
depths of the theory of irreversible processes restricted to consideration of
exclusively dissipative phenomena. On the contrary, all basic relationships
of this theory will hereinafter be obtained as a consequence from
thermokinetics (Chapter 4).

2.4. Elimination of Dissipative Terms from energy Balance Equations

In non-equilibrium systems the majority of the state parameters Θi and

Rj vary due to not only energy interchange between the systems and the
environment, but also in relaxation processes running in them, i.e.
spontaneously. In particular, the system volume may increase with no work
done, but due to expansion into void, the mass Мk of the kth substances –
due to chemical reactions running in the system, the entropy S – due to
conversion of ordered forms of energy into chaotic (thermal) form, the
displacement vectors Ri – due to equalization of distributing the above

41
parameters over the system. This fact is usually taken into account by the
so-called balance equations, which integral form looks like:
dΘi = deΘi + duΘi , (2.4.1)
dRi = deRi + duRi . (2.4.2)

Here deΘi – two-sided variations of the Θi coordinates caused by

external energy exchange (heat transfer, expansion work, mass transfer,
etc.); deRj – similar variations of the Rj coordinates caused by useful work
of the ith kind; duΘi, duRj – one-sided spontaneous variations of the same
parameters caused by running internal (relaxation) processes and, depending
on sign, named usually either “wells” or “sinks” of corresponding
parameters.
Expression (2.4.2) may be more conveniently represented as the flow
balance equation:

Ji = – Jiе + Jip , (2.4.3)

where Jiе = –deRi/dt ; Jip = duRi/dt – useful and relaxation components of the
Ji flow, respectively. Here the positive direction of the flow, just as the
forces Fi and Хi, is construed as the direction toward equilibrium, which
corresponds to the rule of signs accepted in thermodynamics, viz. the
external work Fi·dеRi is positive if done by a system.
To derive the differential form of balance equation (2.4.1), let us
represent the equation in the form:

dΘi/dt = – ∫ ∇⋅jiе dV + ∫ σi dV, (2.4.4)

where the augend corresponds to the term deΘi/dt and the addend – to
duΘi/dt; σi – density of internal wells or sinks Θi.
To change over from the integral form of (2.4.4) to the differential one,
let us represent the Θi as the integral ∫ρidV, where ρi = (∂Θi/∂V) = ρi(V,t) –
density of the Θi value. We are going to consider a continuum, which mass
elements dМ = ρdV remain invariable during its deformation (such a
description is named spatial or Lagrangian). Then dΘi/dt = ∫ (dθi/dt)ρdV,
and one can write instead of (2.4.3):
ρdθi/dt + ∇⋅jiе = σi . (2.4.5)

where jiе = ρiwi – density of the energy carrier flow across the borders of the
system elements; wi = vi – v – velocity of the value Θi transfer relative to the
moving border of a system element. Equation (2.4.5) is called the
substantial equation describing balance of some field value Θi.

42
 In the similar way one can change over to the differential form of
balance equation (2.4.3) allowing for relationship (2.3.11):

ji = –jiе + jiр . (2.4.6)

Substituting balance equations (2.4.5) and (2.4.6) into (2.3.7) and

considering that, as in (2.3.12), dθi/dt is constant all over the bulk of the
system and factoring the mean-mass potential Ψi and the force Xi out of the
integral sign gives according to (2.3.14):

dU/dt = Σi∫ψi (dθi/dt)ρdV – Σi∫xi·jidV – Σi∫ψiσidV + Σi∫xi·jiрdV =

= ΣiΨi deΘi/dt – ΣiXi·Ji – Σi∫ψiσidV + Σi∫xi·jiрdV. (2.4.7)

Comparing this expression with (2.3.9) gives that the last two of its
terms containing wells and sinks of various coordinates mutually balance.
From here the relation ensues between the wells and the sinks of various
state coordinates:

Σi ψiσi = Σi xi·jip , (2.4.8)

which will hereinafter (Chapter 5) directly lead to the dissipative function

expression as fundamental for the theory of irreversible processes.
Due to the mutual balance between the wells and the sinks of various
coordinates thermokinetic identity (2.2.9) remains valid even when the
variation of its terms is caused by internal (irreversible) processes. In other
words, the thermodynamic identity remains valid within the entire spectrum
of real processes – from “quasi-reversible” to critically irreversible.

2.5. Ordered and Unordered energy (inergy and anergy)

as Characteristic Functions

The fact energy lost its primary meaning as a measure of capacity for
work after the notions of internal and dissipation energy have been
introduced engendered the problem to find a measure of motion or
interaction, which would characterize such an important property of energy
as its capacity for ordered (useful) work. The notion of external energy
should have seemingly got the goals. The external energy was construed as
that part of the energy of a system which did not depend on its internal state
and was defined by exclusively the motion or position of the system as a
whole relative to other bodies. As in mechanics, the external energy was
defined by the useful external work a system could do when transiting from
one configuration into another (adopted for a reference frame). The internal
energy was accordingly construed as another part of the energy which did

43
not depend on the position or motion of a system relative to other bodies
and was defined by exclusively the internal (latent) motion and interaction
of particle comprising the system. Since in equation (1.3.4) the internal
energy played the role of dissipation energy, i.e. the energy that lost its
capacity for work, it was equivalent to its division into workable and
unworkable parts. However, later it became clear that a part of external
energy, nevertheless, depends on the internal state of the system. E.g., in
dielectric and magnetic materials the resulting electrical and magnetic field
depends on the temperature of these bodies. In this case the polarization and
magnetization work of these bodies involves a work done against external
fields, i.e. changes both the external and internal energies. In some
particular cases the situation could be simplified by introducing an
additional notion of “internal self-energy of such bodies without field
energy in vacuum” (I. Bazarov, 1991 et al). However, this energy, strictly
speaking, is not anymore a “self-energy” since the external field has already
been changed by the polarized or magnetized bodies available therein.
Furthermore, with the advent of the specific theory of relativity it was
revealed moving bodies did not feature at all a whatever part of energy
which would not depend on the rate of their motion. (R. Tolman, 1974).
As a result, the energy irreversibility and dissipation phenomena – most
fundamental for thermodynamics and natural science – have become non-
interpretable in whole as resulting from the conversion of external energy
into internal. The way out of the situation was seen for a long time as the
idea offered by H. Helmholtz and consisted in the division of energy into a
“free” part” Г = U – TS and a “bound” (to heat motion) part TS. As will be
shown in Chapter 4, the decrease of the Helmholtz free energy F under
certain conditions actually defines the capacity of systems for reversible
work. However, with the arbitrariness classic thermodynamics tolerates in
choosing the internal energy reference frame the bound energy value TS
appears in many cases to be in excess of the internal energy itself (K.
Putilov, 1974). Under these conditions this may not be considered as a part
of internal energy. The same may be also said of the Gibbs’ “free enthalpy”
G = Н–TS (where Н = U + рV is enthalpy). Furthermore, in open systems
these state functions do not define anymore the work a particular system
does (G. Gladyshev, 1988; I. Bazarov, 1991). In this connection the division
of energy into “free” and “bound” loses its heuristic value in many respects.
The situation was not saved with the notion of exergy subsequently
introduced (Z. Rant, 1955) as the maximal work a system can do before
complete thermodynamic equilibrium with the environment has been
obtained. The remainder unworkable part of the system energy Rant termed
as anergy. The exergy and the anergy of a system define its technically
usable and unusable parts, respectively. However, their values depend on
not only the parameters of a system itself, but on the parameters of the
environment as well so that the exergy is not the function of state of a

44
system itself. Furthermore, the exergy does not at all define the capacity of a
system itself to do work since depends on the amount of heat supplied to the
system from outside in the process of its work. And what is more, the
exergy may even be a negative value (unlike the energy) if the intensive
parameters of the system (e.g., its temperature) are below the same of the
environment. We say nothing of the difficulties in calculating the exergy
under inconstant environmental conditions, as well as of the arbitrariness in
choosing its reference point under the conditions when environmental
parameters substantially differ for different points of the planet. As a result,
the division of energy into exergy and anergy has not either gained a wide
spread.
A solution to the said problem becomes possible when changing over to
studying non-equilibrium systems and introducing additional non-
equilibrium state parameters of spatially heterogeneous media Xi and Zi. Let
us assume that we deal with a spatially homogeneous medium (Xi = 0)
where the generalized potential values Ψi are the same for all points, i.e.
equal to their equilibrium values Ψi . For such a system in the form of (2.2.9)
instead of identity (2.2.6) the following equation is available:

dŪ = Σi Ψi dΘi, (2.5.1)

where Ū – equilibrium part of the energy in non-equilibrium system.

The right-hand terms of this equation characterize the equilibrium heat
exchange between the system and the environment. The term “equilibrium”
means that at Xi → 0 processes are running infinitely slowly (quasi-
statically) and, therefore, do not disturb the spatial homogeneity of the
system (Zi → 0). The work of these processes refers to the unordered
category due to the absence of a resultant of the forces to be overcome in
homogeneous systems. The work of such a kind involves introducing the kth
substance across the system borders (when not only the cubic strain occurs,
but the interparticle interaction forces are also overcome), electrical charge
introduced, uniform compression, mass transfer, momentum convective
transfer along with the mass introduced, and heat exchange we attribute to
the “micro-work” category on the grounds that heating of a body accelerates
the chaotic motion of particles.
Applying the Legendre transformation to the above summands Ψi dΘi =
dΨi Θi – Θi dΨi in the extreme case of Xi = 0 and Ψi = Ψi gives:

d(Ū – Σi Ψi Θi) = – Σi Θi dΨi . (2.5.2)

This expression refers to a system with no work done on, which could
disturb it from equilibrium. Hence, according to the self-non-disturbance
axiom both parts of (2.5.2) become zero. From this it follows that the

45
equilibrium part of system energy Ū up to a constant (which could be
assumed equal to zero) is defined as:

Ū = Σi Ψi Θi = ΣiŪi. (2.5.3)

According to (2.5.3) the unordered energy of a system is the sum of the

system component energies Ūi = Ψi Θi (by the number of degrees of
freedom), each being defined as the product of the momentum
(interaction) of a particular kind Θi by the potential Ψi as a measure of its
intensity. The value Ūi characterizes the part of the non-equilibrium system
energy left after the completion of all relaxation processes in the system
(equalization of all potentials and completion of all chemical and the like
internal transformations) and unable to do useful (ordered) work.
Therefore we will call it the unordered energy or (for short) the anergy.
The remainder part of the system energy Е = (U – Ū) is ordered, i.e.
capable of doing ordered work. This part of the system self-energy will be
called (for short) the inergy1). The fact of its order shows in the energy
dissipation phenomenon meaning transformation of motion into unordered
form. To express the ordered and unordered energies in terms of state
parameters, let us consider that the spatial heterogeneity parameters Fi or Ri
appear in the expression for the system energy U as a consequence of the
uniform distribution of the usual “thermostatic” parameters Θi all over the
system volume V (Chapter 1). When a system “contracts” to a material point
(V→ 0), the parameters ρi(r,t) → ρi0(t) and Ri → 0. This means that in
themselves the bulk elements of a heterogeneous system feature the same
set of the variables Θi as in the homogeneous state. Hence, the unit volume
energy as a function of the local variables ψi and Θi = (∂Θi/∂М) has the same
form of (2.5.4) as for an equilibrium system, while the energy of the system
as a whole is expressed through the integral:

U = Σi∫ψi θi ρdV. (2.5.4)

This part of the system energy may be represented in the even shorter
form:

U = ΣiΨiΘi, (2.5.5)

1)
In contrast to the Anergy, this term reflects the capacity of a system for internal work, i.e.
for transmutations of Energy (including the dissipation character). It is evident that a system
not having this ability can not either participate in the Energy exchange involved in these
processes, i.e. can not do useful internal work.

46
if factoring outside the integral sign in (2.5.5) some “mean-energy” value of
the generalized potential Ψi = Ui/Θi (not coinciding, in the general case, with
Ψi. Then the system ordered energy as a state function becomes:

Е = Σi (Ψi – Ψi)Θi . (2.5.6)

The ordered energy of non-equilibrium systems should also be

complemented with chemical, atomic, nuclear, etc types of system energy.
These types as well have the capacity for work and dissipation just as the
energy of motion and interaction between macroscopic parts of the system.
Chemically reacting media, spontaneously fissionable nuclear fuels,
optically excited and ionized gases, etc are non-equilibrium even if spatially
homogeneous. The non-equilibrium of such a kind existing in each of the
system bulk elements and not associated with the spatial heterogeneity of
the system as a whole will hereinafter be called local. To describe the state
of locally non-equilibrium systems, it is also necessary to introduce special
parameters of non-equilibrium. Here comes, in particular, the completeness
degree (intensity, range, etc) of the rth chemical reaction ξr. The similar
coordinates may be introduced for phase transitions (V. Etkin, 1991). As
will be clearly shown hereinafter, the scalar nature of such state coordinates
will impose certain constraints on the possible conversion of such forms of
energy into ordered work.
As seen, the thermokinetic approach makes superfluous such notions as
external and internal, free and bound energies, Gibbs’ energy and exergy
superseding them by the intuitively comprehensible terms of ordered and
unordered energies. Another nontrivial consequence of such an approach
consists in the conclusion there is an ordered and unordered component in
any of the ith forms of matter motion including the internal thermal energy
Uт that in thermal physics has always been entirely attributed to the
chaotic form of motion and, therefore, considered as an entirely
depreciated (degraded) energy. Meanwhile, thermal energy has also an
ordered part that depends on the temperature gradients or drops in a
system. In itself the fact a thermally heterogeneous system is able to do
useful work has been known yet since S. Carnot (1824) who expressed it
as, “all over where a temperature difference is available, the “living force”
(i.e. the thermal “motive force” (author’s note)) may also appear”. In
thermal relaxation processes this ordered part of thermal energy dissipates
(degrades) just as other ordered forms of energy. This phenomenon may
be experimentally confirmed by, e.g., the so-called measurable heat
content rise effect (L. Brovkin, 1960, 1964).
In series of his experiments Brovkin discovered a phenomenon
unexplainable from the classic thermodynamic standpoint and consisting
in increasing mean-integral temperature in the process of thermal

47
relaxation of an isolated thermally heterogeneous system. The experiments
were conducted as follows. A resistance temperature detector sensing
element was inserted into the gap of a closely packed roll of paper,
cardboard, rubber tape or other sheet materials all over their length. Then
the roll was subjected to irregular heating from an external heat source
followed by the spontaneous process of thermal relaxation. In that process
the resistance variation for such a “distributed” thermometer was
registered. Surprising was the fact that for all of the materials tested a
considerable (up to 17%) rise of the mean-integral temperature was
observed. That rise had continued for tens of minutes until the temperature
fall due to cooling of the specimen not quite thermally protected became
dominating. The experimental results were attempted to have been
explained by the dependence of thermal capacity of the materials on the
temperature gradients therein and were construed as a rise of “measurable”
(i.e. different from actual) heat content (enthalpy) of the system.
Meanwhile, the experimental results directly demonstrated the rise of
unordered part of internal thermal energy (analog of the Helmholtz’s
“bound” energy) due to the decrease in its ordered part (analog of his
“free” energy).
The fact the ordered E and unordered Ū energies belong to the non-
equilibrium state functions with the exact differential properties incident to
them in this connection makes it possible to generalize the classic method of
characteristic functions to spatially heterogeneous (in their properties)
systems with an arbitrary number of degrees of freedom. The characteristic
functions are construed in thermodynamics as such state functions of a
system, mathematical operations on which allow direct definition of
fundamental properties of this system. In itself the term characteristic
function was introduced by Massier (1969), who first realized those functions
as convenient to be used in finding the system parameters as their derivatives.
One of such characteristic functions is the system energy U itself, which
complete variation in any processes is represented by (2.2.5). In accordance
with (2.2.5) and (2.2.3) the reason why any of the ith non-equilibrium
processes arises – the generalized force Fi – may be found as the derivative of
system energy with respect to the displacement vector Ri under the constancy
of other similar vectors Rj and all “thermostatic” coordinates Θi.
In the similar way the derivatives of the system anergy U with respect to
the “thermostatic” coordinates Θi under the constancy of other similar
coordinates Θi and all displacement vectors Ri allow finding the generalized
potentials for the non-equilibrium state of the system:

Ψi = (∂U/∂Θi). (2.5.7)

48
 In the particular case when Θi constitutes the heat exchange coordinate
ΘТ (entropy) expression (2.5.6) defines the absolute temperature T of the
system:

Т = (∂U/∂ΘТ). (2.5.8)

In the other particular case when Θi = Θр = Vо–V (where Vо means the

theoretical compression limit) this derivative allows definition of the
absolute pressure p in the system:

р = (∂U/∂Θр). (2.5.9)

The second derivatives of the energy U with respect to the coordinates

Θi allow finding other thermodynamic properties of thermodynamic
systems. In particular, for a thermo-mechanical system, where U = U(S,V),
the second derivative of U with respect to entropy under the constancy of
volume and other equilibrium coordinates allows finding the isochoric heat
capacity СV of the system:

(∂2U/∂S2)V = (∂Т/∂S)V = Т/СV. (2.5.10)

Due to the breakdown of energy into the ordered and unordered

components we receive in addition to it two more characteristic functions Е
= ΣiЕi and Ū = ΣiŪi, where Еi, Ūi – components of Е and Ū, which we will
hereinafter call the partial energies (of the ith kind). These functions known
allow to solve the same problems as the energy under less strict constraints.
In particular, the forces Fi can be found as the derivatives of ordered energy

Fi = – (∂Е/∂Ri) = – (∂Еi /∂Ri), (2.5.11)

omitting in the first case the constancy requirement for all thermostatic
variables Θi and in the second case – the constancy requirement for all other
displacement vectors Rj ≠ Ri. Furthermore, U replaced by Ei allows finding
the internal forces Fi acting within isolated systems (where dU = 0). In
mechanical systems, where the ordered energy E is identical to the potential
energy of the system, the above expression goes over into the standard force
expression in mechanics (L. Landau, E. Livshits, 1973).
Another valuable property of the characteristic functions Е and Ū is that
they may serve as thermodynamic potentials, i.e. as the values, which
decrease defines the maximal work a system can do. In particular, the
decrease of the ordered energy E defines the sum of all kinds of the useful
external Wie and the dissipative external WiD works the system do:

–dЕ = Σi Xi·dZi = ΣiđWie + ΣiđWiD. (2.5.12)

49
 From this it follows that the ordered energy of a system is measured by
the value of the maximal useful external work Wie the system can do subject
to the reversible (dissipativeless) character of this process (WD = 0). This
makes the notion of ordered energy related to the external energy Еk + Еp in
mechanics. However, firstly, unlike mechanical systems, where the external
energy does not depend on the internal state of the system, the ordered
energy in the generally polarized, magnetized, irregularly deformed, etc
media vary with the variation of temperature, composition and other internal
parameters of the system. Secondly, from the positions of thermokinetics
considering as a system the whole set of investigator’s interest comprising
interacting (mutually moving) bodies or body’s parts the notion of external
energy is devoid of sense. Indeed, the external energy by definition depends
on the motion of the system or its position relative to the environment, i.e. is
“mutual” and may be ascribed to one of them in only the simplest case.
Therefore, when the environment is included in the “extended” non-
equilibrium system the external energy becomes an ordered part of its “self-
energy” U not depending on the environment. Accordingly we will refrain
from using the term “external energy” where it might cause
misunderstanding. Similarly the decrease of the unordered energy Ū defines
the maximal external unordered work a system can do on the environment
subject to the reversible character of the process.
The advantage of representing thermodynamic potentials in terms of
ordered and unordered energy becomes apparent when allows finding the
ordered Wie and unordered Wiн work for each of its ith forms. In this respect
they stand out from the Gibbs and Helmholtz free energies in classic
thermodynamics, which decrease characterizes just their sum of We and Wн.
It is significant that in this case the energy components introduced herein
can be found for any current state of the system by known fields or
distribution function for the energy carrier Θi and known equations of
system state in the general form as ψi = ψi (θ1, θ2, …, θn). In particular, the
potentials Ψi can be found as mean-mass values of the local potential ψi,
while the potentials Ψi – as their “mean-energy” values. To find them, the
equations are used interconnecting the variables ψi and Θi similar to the
equations of ideal and real gas state.
The further advantage of representing characteristic functions in terms
of the non-equilibrium parameters becomes apparent when revealing a
whole class of internal conversion processes involved in the external energy
exchange. As expressions (2.5.1) and (2.5.2) show, the partial energy of any
degree of freedom Ūi can vary both as a result of the external action dŪiе
= Ψi dΘi and due to the thus caused internal conversions of this energy into
other forms (dŪiu = Θi dΨi ). In this case

50
 –Σi Θi dΨi = 0. (2.5.13)

This expression generalizes a known Gibbs-Duhem’s relationship (I.

Bazarov, 1991) earlier obtained for the particular case of open thermo-
mechanical systems to the polyvariant equilibrium systems with any finite
number of degrees of freedom. However, the sense of this relationship from
the energy standpoint as the law of energy conservation at internal energy
conversions from one form into another becomes clear with only
introducing the notion of partial energy. The existence of such internal
conversions by no means follows from classic thermodynamics since classic
thermodynamics proceeds from the assumption that quasi-static impacts
such as Ψi dΘi do not disturb equilibrium in a system and, therefore, can not
engender whatever internal processes therein. Meanwhile, for the isolated
systems thermokinetics considers studying internal energy conversion
processes of such a kind is brought to the forefront because of no other
processes existing in such systems. Therefore, it seems to be important to
demonstrate that expression (2.5.13) represents nothing else but the energy
conversion processes that acquire the quasi-static character in equilibrium
systems
According to (2.5.4) the rate of system energy variation dU/dt is derived
as:

dU dθ dψ i
= ∑ i ∫ ψ i i ρdV + ∑ i ∫ θi ρdV . (2.5.14)
dt V
dt V
dt

Since in non-equilibrium systems the potentials ψi are a function of the

field point radius vector r and time t, then:

dψ i ⎛ ∂ψ i ⎞ ⎛ ∂ψ i ⎞ dr
=⎜ ⎟ +⎜ ⎟ (2.5.15)
dt ⎝ ∂t ⎠ r ⎝ ∂r ⎠ t dt

Substituting (2.5.15) into (2.5.14) and allowing for (∂ψi/∂r) = – xi ; dr/dt

= vi gives:

dU d θi ∂ψi
= ∑ i ∫ ψi ρdV − ∑ i ∫ xi ⋅ ji dV + ∑ i ∫ θi ρdV . (2.5.16)
dt dt ∂t

Comparing this expression with (2.4.7) gives that the last sum in
(2.5.16) becomes zero owing to equality (2.4.8), i.e. due to the mutual
balance of the terms containing wells and sinks of various coordinates:

51
 Σi ∫ θi(∂ψi /∂t) ρdV = – Σi ∫ ψiσi dV + Σi ∫ xi·jipdV = 0. (2.5.17)

This expression is the further generalization of the Gibbs-Duhem’s

relationship to spatially heterogeneous systems when the potentials become
a function of the point ψi = ψi(r, t). It reflects the fact that during the internal
conversions of the ordered energy forms (xi·jip) into the unordered ones
(ψiσi) the energy of the system remains invariable (in strict compliance with
the law of its conservation)1). At the same time this relationship once again
confirms that such conversions are possible in only spatially heterogeneous
systems (where xi ≠ 0). This means that the Gibbs-Duhem’s classic
relationship refers to the case of quasi-static processes when jip → 0. It is
this internal conversion of energy that makes sense for the Gibbs-Duhem’s
generalized relationship from the energy standpoint.

2.6. Definition of General Form for Equations of State and Transfer

The above-developed body of mathematics is in itself insufficient

for the investigation of particular processes. For that uniqueness conditions
are required enabling to complete the set of the thermokinetic equations.
Although thermokinetics imports these uniqueness conditions from outside
it, nevertheless, dictates certain requirements for the number, general form
and character of such uniqueness conditions.
The equations of state describing systems under investigation are
the fundamental component of the uniqueness conditions. They reflect the
relationship between the variables entering into thermodynamic identity
(2.2.5) or (2.2.6). The total number of such equations is defined by the
number of the independent variables entering into this identity, i.e. by the
number of the independent processes really running in a system being
described. Particular form of these equations is found by experimental or
theoretically experimental way based on the simulation procedures
including those of statistically mechanical character.
To go on with, let us restrict the consideration to an ample spectrum of
systems lacking the reorientation processes. Let us assume the experimental
investigation of such a system has revealed n transfer or redistribution
processes running in it (i = 1, 2, …, n). Then according to identity (2.2.5) or
(2.2.9) the total number of variables characterizing the state of the system
and its energy E is generally equal to 2n. However, due to the constraints
imposed some number of the processes of one of these categories may be

1)
Hereinafter (Chapter 10) this will facilitate the understanding that thermal Energy is
inevitably converted into its ordered forms when bodies accelerate up to the relativistic
velocities.

52
absent. Let us assume the number of the transfer processes is ni (i = 1, 2,...,
ni), while the same of the redistribution processes – nj (j = 1, 2,..., nj). Then
the total number of the variables entering into identity (2.2.9) is equal
2(ni+nj). Since only (ni+nj) of them are independent, there must exist for
such a system ni equations interconnecting the variables Ψi and Θi and
having the general form as:

Ψi = Ψi (Θi, Rj) , (2.6.1)

and nj equations interconnecting the variables Хi and Zi and having the

general form as:

Хi = Хi (Θi, Rj) . (2.6.2)

Relations (2.6.1) may be instantiated by Clapeyron-Mendeleev’s law

and Van der Waals’ law for gases, while relations (2.6.2) – by Hooke’s law
for elastic bodies, Curie-Weiss’ law for dielectrics, etc. The fact itself such
equations exist enables definition of a number of relations between
physicochemical properties of the systems under investigation. These are,
e.g., the relations between thermal coefficients of expansion for gases or
isothermal compression coefficients and coefficients of elasticity in
thermodynamics of solid bodies (Bazarov I.P., 1991).
When studying non-static processes, together with the equations of state
it is necessary to know those relating velocity of some process with its
motive forces. These equations include the laws of transfer of heat, charge,
the kth substance, momentum, etc. (Fourier’s, Ohm’s, Fick’s, Newton’s
laws, etc.). They relate the generalized velocity of vector processes of heat
conduction, electric conduction, diffusion and viscous friction. Arrenius’
law relating the velocity of some rth homogeneous chemical reaction wr with
its affinity Аr is an example of the scalar process equation. Fundamental
equations of thermokinetics (2.2.5) and (2.2.6) without such empirical laws
are equally as barren as the unified equation of the first and second laws of
thermodynamics in the absence of the equations of state of the systems
under investigation. The particular form of the transfer equations is derived
a posteriori (from experience). However, thermokinetics makes certain
demands on the number and general form of such equations imposing some
constraints on the character of the interrelation between the state parameters
these equations contain. In fact, having equations of state (2.6.1) and (2.6.2)
at our disposal, let us represent the displacement vectors Rj as a function of
ni coordinates Θi and nj motive forces Fj:

Ri = Ri (Θi, Fj) (i = 1,2,..., ni; j = 1,2,..., nj). (2.6.3)

53
Taking total derivative of both parts of this expression with respect to time
and keeping in mind that in stationary processes (where Ψi, Fj =const) these
derivatives refer to the value Ri only and characterize velocity of the ith
transfer process wi = dRi/dt gives:

wi = wi (Θi, Fj) (i = 1,2,..., ni; j = 1,2,..., nj). (2.6.4)

Respectively, the displacement flow Ji ≡ ΘidRi /dt may be defined as:

Ji = Ji (Θi, Fj) . (2.6.5)

This relationship reflects the general form of the transfer laws ensuing
from the equations of state of the system under investigation. The particular
form of these equations is derived a posteriori. However, the fact itself of
existing equations for the flows Ji as a function of the state parameters
describing a non-equilibrium system is a good ground to attribute these
flows to the state functions of such systems1). Note that this statement could
by no means ensue from the existing theory of irreversible processes since
this theory, according to the local equilibrium hypothesis, has never treated
the thermodynamic forces as system state variables. Functional relationship
(2.6.5) allows writing the exact differential of the flow Ji as:

dJi = Σi (∂Ji/∂Θi)dΘi + Σj(∂Ji/∂Fj)dFj (j ≠ i). (2.6.6)

Now, to find the displacement flows, this expression should be

integrated between the limits from the equilibrium state (where all Ji = 0) to
the current state:

Ji= Σi ∫(∂Ji/∂Θi)dΘi + Σj∫ (∂Ji/∂Fj)dFj (j ≠ i). (2.6.7)

Note that the equilibrium state with Ji = 0 is attained for the so-called
“memory-possessing media” (viscoelastic and viscoplastic materials,
hysteresis phenomena, etc.) generally at some the so-called “threshold”
force value Fj0 other than zero. Therefore the integration with respect to Fj
will be started from exactly this value Fj0. Taking into account the fact that
the value of any parameter does not depend on the integration path, let us,
for convenience’s sake, integrate (2.6.7) first with respect to Θi at Fj - Fj0 =
0 and then with respect to Fj at Θi = const. In this case the integral of the
first sum in this expression does not at all contribute to the flow Ji since at
equilibrium state (Fj - Fj0 = 0) none of the coordinate Θi variations can cause

1)
This statement is especially evident for isolated systems: all processes running therein are
defined by exclusively the properties of a system as itself.

54
the flow Ji. As for the terms of the second sum, they may be represented
more compactly by designating

Kij ≡ (∂Ji/∂Fj) (j ≠ i). (2.6.8)

Then expression (2.6.7) becomes:

Ji = Σi ∫ Kij dFj . (2.6.9)

Let us consider the most general case, when the phenomenological

coefficients Lij explicitly depend on the forces Fj and thermostatic variables
Θj, i.e. Kij = Kij(Θj,Fj), and. the dependence of Ji on Fj is non-linear. Then
this relation may be represented in the pseudo-linear (linearized) form.
Factoring some values of the functions K̄ ij = K̄ ij (Θj,Fj) averaged between
the limits from Fj0 to Fj outside the integral sign the above expression
becomes:

Ji = Σi K̄ ij (Fj – Fj0), (i, j = 1, 2, ..., nj) (2.6.10)

This expression is the most general form of the equations describing the
transfer of heat, substance, charge, momentum, etc. In many of phenomena
the threshold value Fj0 of the force Fj may be neglected. Then the transfer
laws in the form of 2.6.10 can be rewritten in a simpler form:

Ji = Σi K̄ ij Fj . (2.6.11)

Thus thermokinetics suggests for an investigator a self-consistent form

of representing the results of experiments on finding the transfer equations.
As for the proportionality factors K̄ ij themselves, their finding is the
problem for physical kinetics that uses for this, along with experiment, also
considerations of the molecular-kinetic and statistical-mechanical theories.
The differential relationships between the state parameters and their
functions enable solution of also other problems. In particular, they allow
calculating by experimental data the fundamental state functions of a
system. Although these relationships can not describe the mechanism of
processes under investigation, they impose the restricting conditions that
should be met for each of models. This always clarifies the models because
allows neglecting redundant or non-existent constraints. Therefore, the
differential relationships are a very effective tooling of mathematical
analysis conducted on an object of investigation.

55
 Conclusions to Part 1

One of the most attractive features of the thermodynamic method has

always consisted in its possibility to obtain, based on few primary principles
(“the beginnings”), a great number of consequences covering various
phenomena and having the status of oracles. Being consistently
phenomenological (i.e. empirical), this method enables to reveal the general
behavior of various processes without intruding into their molecular
mechanism and not resorting to modeling on structure and composition of a
system under investigation. All this contributed to the exclusive heuristic
value and potency of the thermodynamic investigation method.
The intention to keep these advantages impelled us to base
thermokinetics on the same methodological principles as a consistently
phenomenological and deductive theory. The methodological concept of
thermokinetics features not only the system approach to objects of
investigation which became possible with introducing the parameters of
their spatial heterogeneity, but also negation of process and system
idealization, exclusion of postulates and hypotheses from the theory
grounds, as well as the strictly deductive method used to obtain all its
consequences. None of the other physical theories offers advantages in such
a combination.
The body of mathematics thermokinetics operates is attractively simple
and not outstepping the framework of college educational packages.
Following the classic thermodynamic method of characteristic functions this
body of mathematics is mainly based on the properties of their exact
differential and “imports” all the rest data on the properties of a system
under investigation “from outside” as uniqueness conditions of a kind. This
imparts the status of oracles to the thermokinetics consequences providing
the uniqueness conditions comply with the framework of experiment. It is
naturally to expect, therefore, that the theory based on these principles will
open up new vistas in cognition of Nature.

56
 Part 2

HEAT ENGINEERING FUNDAMENTALS

The main object of this section is to give a methodologically unitary

substantiation and statement of the principles, laws and equations for a
number of fundamental disciplines constituting the theoretical grounds of
heat engineering (the theory of heat engines, thermodynamics of reversible
and irreversible physicochemical processes and the theory of heat-mass
exchange). Such a statement paves the shortest way to comprehension and
cognition of the above disciplines – the way free of multiple extraneous
features these disciplines have historically acquired in the making. This way
is based on a unitary notional system and further clarifies the
interdisciplinary links facilitating the comprehension of their specificity, as
well as the further in-depth study of the disciplines. At the same time such
an approach excludes the necessity to extrapolate the notional and
conceptual base of these disciplines beyond the rigorous frames of its own
applicability thus facilitating the further critical analysis of the problems
arising in various spheres these disciplines applcations.

Chapter 3

CLASSIC THERMODYNAMICS

Present-day thermodynamics has long outgrown the initial frames of the

heat-engine theory and transmuted into a rather general macroscopic
method for studying kinetics of various transfer processes in their
inseparable connection with the thermal form of motion. However, it is still
rooted in the conceptual system of equilibrium thermodynamics
(thermostatics) remote from the transfer concept and in its body of
mathematics going over into inequalities when considering real (non-static,
irreversible) processes. Even in the current manuals on thermodynamics its
construction quite often starts with describing the theory of ideal cycles and
ideal gases as its working media. Such a “squared idealization” in the theory

57
grounds themselves could not help creating problems in the further
generalization of thermodynamics to systems differing from those idealized.
The escape from the situation may be found in obtaining basic
statements of thermodynamics by deductive way as consequences of
thermokinetics in order to avoid the extrapolation of classic
thermodynamics beyond the framework strictly bounding the applicability
of its initial concepts. This chapter is mainly dedicated to ascertaining the
minimal scope of the corrections necessary to introduce into the
fundamentals of thermodynamics from the positions of thermokinetics.

3.1. Principle of the Excluded Perpetuum Mobile of 1st Sort

(the First Law of Thermodynamics)

R. Clausius, the founder of classic thermodynamics, presumed the

principle of heat Q and work W equivalence generalizing the results of
numerous experiments to be the only reliable basis to thermodynamics. He
formulated this principle as follows, ”In all cases, when heat becomes work
in a cyclic process, the amount of the heat expended is proportional to the
work done and vice versa, work done is converted into an equivalent
amount of heat” (Clausius, 1876). If heat and work are measured in the
same units of the international system of units, SI, the equivalence principle
may be written as a simple relationship:

Qc/Wc = 1, (3.1.1)

or
∫ (đQ + đW) =0 (3.1.2)

where đQ, đW – elementary amounts of heat and work for particular parts of
the cyclic process under consideration.
Clausius was the first who noticed that the above relationship did not
depend on the nature of the cyclic process under consideration. According
to a known curvilinear integral theorem the fact the integral along the
contour within the space of variables x1, x2,... xn becomes zero is a sufficient
evidence that the integrand represents the exact differential of some
function of these variables U(x1, x2,..., xn).
Thus the principle of equivalence meant the existence of a specific
function of state U, which variation was equal to the algebraic sum of heat
and work of a process. Allowing for the rule of signs accepted in
thermodynamics (the heat Q fed into a system and the work W done by the
system being positive), the integrand may be expresses as:

58
 dU = đQ - đW . (3.1.3)

R. Clausius named the function U the total heat of a body (he meant the
sum of the heat Q fed into a system and the “disgregation” work of the
dissipative nature, which replenished the “heat of the body” – the energy
associated with heat motion in the system). Such an understanding reflected
a known duality of the heat concept, which had historically been caused by
consideration of the heat as a form of motion (in the same line with such
phenomena as light, sound, electricity, magnetism). This duality means that
heat shows in some cases as the state function (heat of a body), whereas in
other cases – as the process function (heat of a process). Therefore from
Clausius’ times on the function U has been referred to as the internal energy
of a system. However, despite plenty discussions the said duality in using
the term “heat” survives to this day. Accepting this fact as a historically
preconceived reality, let us name the internal heat energy, for short, the
“heat of a body”, whereas the quantitative measure of heat transfer – the
“heat of a process” designating, for the avoidance of mishmash, the former
through Uq and the latter through Q. In this case, to designate the
infinitesimal increments of internal heat energy as a function of state, let us
use the sign of exact differential d, whereas to designate the elementary
amount of work W or heat Q as a process function – the sign đ (đW, đQ).
Equation (3.1.3) has been named the first law of thermodynamics. This
reflects the constancy of energy of an isolated system (U = const at Q, W =
0) and therefore is one of the statements of the law of conservation of
energy. This law reflects the conservation of the internal energy U at its
interchange among a system and the environment in the form of heat Q or
work W. With regard to a cyclic process (cycle) expression (3.1.3) may be
written as:

∫ đQ = ∫ dU + ∫ đW или Qc = ∫ dU + Wc, (3.1.4)

where Qc, Wc – total heat and work of the cyclic process.

Since the circuit integral of any state parameter, including the system
energy U, is equal to zero, expression (3.1.4) directly gives that Wc = Qc, i.e.
the work done by working medium for a cycle is equal to the heat it
expendes in the cycle. In other words, a cyclic machine is impossible, which
would do work without heat expenditure. This statement was named the law
of excluded perpetual motion of the first kind.
At its early stages classic thermodynamics was restricted to the simplest
thermo-mechanical systems, where expansion work was the only kind of
work. A system having been in equilibrium with the environment (equal
pressures p), that work was determined from mechanical interpretation of
pressure as a force and expressed by the relationship đWр = рdV, where V –

59
volume of the system. From there the so-called analytical expression for the
first law of classic thermodynamics ensued in the form:

đQ = dU + рdV. (3.1.5)

It may be said without exaggeration that the majority of methodological

features intrinsic for thermodynamics, as well as its insufficiency, is caused
by that specific form (3.1.5), in which it involves the law of conservation of
energy. Expression (3.1.5) first of all involves only two forms of power
interchange in a system, viz. heat Q and work W done by the equilibrium
system1). Meanwhile, one more type of power interchange referred to
neither heat transfer nor work take place in a great number of cases of
practical interest. This is mass transfer associated with the kth-substances
interchange between a system and the environment. Such systems are called
open ones. These types of power interchange are not covered by equation
(3.1.5) describing the law of conservation of energy, therefore the principle
of equivalency of heat and work can not serve as a basis for further
generalization of this law to open systems. To do this, a more general
approach is needed, which is exactly what thermokinetics puts forward.
Furthermore, equation (3.1.3), unlike (2.2.5), does not contain any
forces Хi causing the generation of the energy-conversion processes. Its
terms on the right-hand side characterize energy interchange between a
system and the environment in the form of heat and cubic strain work
(confirming the principle that a system can only exchange what it really
has). Therefore it describes only processes of energy transfer, but not
conversion. In other words, the first law of thermodynamics characterizes
the balance of energy only and has nothing to do with the law of
conservation of energy at its conversion. It is even more true, because an
internally equilibrium (homogeneous) system can not do useful work in the
absence of other bodies (environment), with which it is not being in
equilibrium. Besides, the energy balance itself described by equation (3.1.5)
comes valid provided the system being in equilibrium with the environment.
Actually, in the absence of, e.g., mechanical equilibrium between a system
and the environment (when the pressure p in the system is not equal to the
environmental pressure pe, and the expression pdV does not define the
expansion work any more) the analytical expression of the first law of
thermodynamics (3.1.5) goes over into inequality:

đQ ≠ dU + рdV . (3.1.6)

1)
Such work will hereafter be referred to as “non-technical” unlike the useful external
(technical) work done by various machines (by extended systems).

60
 This is the reason why the first law in the form (3.1.5) is valid for only
completely (both internally and externally) equilibrium systems.
One more reason of the insufficiency classic thermodynamics
demonstrates is caused, strange though it may appear, by the notion of
internal energy as itself. energy U belonging to, strictly speaking, entire set
of the interacting (relatively moving) bodies is known to be far from being
possibly attributed to one of the comprising parts considered as the object of
investigation. This may be done to only that part which does not depend on
position and motion of the system relative to other bodies and is defined by
exclusively parameters of the system itself. That is exactly where the
definition of the internal energy U is rooted in. However, it ensues from this
definition that such a system can not do useful external work measured by
the variation of external potential or kinetic energy of a system of
interacting bodies. Therefore the internal energy of an equilibrium system is
actually the anergy (see Chapter 2).
The notion of internal energy is quite often extrapolated to systems
lying in external fields of forces, which are, e.g., dielectric and magnetic
materials. Their energy is known to be dependent on not only internal
variables, but as well on intensities of these fields. However, these are often
associated with the notion of magnetic and dielectric “internal self-energy”
considering it sufficient to subtract the energy of electrical and magnetic
fields from the total energy in the volume occupied with the system. This
makes thermokinetics consider a polarized medium along with a polarizing
field as a whole and, generally, non-equilibrium system. Such an approach
is substantiated by the fact that an attempt to isolate a dielectric or magnetic
field from the surrounding fields of forces leads to change of their state
(relaxation), which excludes a possibility of treating their energy as internal
one (self-energy) and applying equation (3.1.3) to it.
There are also other serious and yet hardly perceived constraints for the
equations operating with the notion of internal energy. In this sense the term
“thermodynamic” may be applied only to the so-called “simple” media,
where long-range and surface forces may be neglected since their energy
can not be attributed to just one of the interacting bodies (Caratheodory,
1964). From this one more constraint ensues, viz. for size of a system. This
exists not only “from above” (for galactic-size systems where long-range
gravitational forces can not be neglected), but also “from below”, e.g., for
usual continuums since when artificially splitting them into elementarily
small zones the surface energy of the continuum elements increases and
may appear non-negligible (Putilov, 1971; Bazarov, 1983).
To exclude all these constraints, thermokinetics considers as the object
of investigation “extended systems” comprising such a set of interacting
(relatively moving) bodies and fields, which energy, to enough degree of
accuracy, may be considered as its “self-energy”. Such systems may be
isolated not disturbing their internal state and energy. Furthermore,

61
thermokinetics offers an absolutely other (and in many ways oppositional)
approach to stating the law of conservation of energy. This approach is
notable for its basing the law of conservation not on the energy balance
equation, which equation (3.1.3) is, but on the representation of the system
energy as a function of a quite definite number of the system state
parameters.
Suppose we deal with a thermo-mechanical system having only two
(thermal and mechanical) degrees of freedom. These two degrees of
freedom are described with two extensive state coordinates, which we
formally designate as Θt and Θe. Their meaning will be explained hereafter.
For now it is enough to know that the total energy of the system (consisting
in this case exclusively of its internal energy U) as a function of state is
expressed in the form of U = U(Θt, Θр), whereas its exact differential may
be written as:

dU = (∂U/∂Θt)dΘt + (∂U/∂Θe)dΘe . (3.1.7)

Designating Ψt ≡ (∂U/∂Θt) и Ψe ≡ (∂U/∂Θe) gives a thermo-dynamic

identity:

dU ≡ ΨtdΘt + ΨрdΘр . (3.1.8)

According to this identity the energy of the system varies with variation
of the said parameters Θt and Θe irrespective of what causes the variation –
either energy interchange or internal processes in the system. Undoubtedly,
in the latter case the terms ΨtdΘt and ΨedΘe do not reflect any more heat
exchange or cubic strain work. To define them, thermokinetics offers other
ways (see Chapter 2). That enabled thermokinetics to refuse the
classification of processes by the type of energy interchange. It was the said
classification that engendered the problem of thermodynamic inequalities.
As a matter of fact, the principle of classification by the type of energy
interchange demands the availability of state coordinates that would remain
unvaried in the absence of heat transfer processes and expansion work.
However, this is excluded in a non-equilibrium system, where a number of
its parameters can vary spontaneously. For this reason classic
thermodynamics is known to be restricted to equilibrium states and quasi-
static processes not practically disturbing this equilibrium.
Unlike it, thermokinetics classifies processes by exclusively those state
variations they cause. Such a classification provides the fundamental
equation of equilibrium thermodynamics in the form of (3.1.7) retains the
nature of identity even in the case when the parameters Θt and Θр vary
spontaneously, i.e. in the whole range of real processes.

62
 3.2. Principle of Entropy Existence (Second Law for Reversible
Processes)

Let us clarify the meaning of the variables Θt and Θe introduced above

formally (by definition). Let us start off with the coordinate Θe of expansion
work. It is considered quite natural to accept for this the volume V of a
system. However, in this case the volume-correlated system potential Ψe ≡
(∂U/∂V) according to relationship (2.2.2) takes on the meaning of pressure
reversed in sign (Ψр ≡ – р). According to equation (2.3.2) this entails
opposite-to-actual sign of the motive force of gas flow transfer process (as
known from experience, gas flow is transferred by pressure-gradient force
toward the lower-pressure zone – just like with potentials of other nature).
Furthermore, the volume (positive or negative) does not comply with the
requirement of thermokinetics for potential and coordinate of some process
simultaneously becoming zero. Lastly, the volume of a system can not serve
as a cubic strain coordinate in open systems because it necessarily varies at
mass transfer with the density ρ of the system remaining constant. For
systems, where the processes of composition variation run (due to diffusion
of the kth substances across the borders of the system at constant mass M of
the system), the cubic strain coordinate issue grows even more complicated
since the system density ρ = Σ сkρk can vary with constant density ρk of any
of the kth substances, but with the variation of the fractions сk of their total
masses. Therefore the cubic strain issue appears to be, in actual fact, not that
plain.
From thermokinetics any state coordinate Θi is considered as a
quantitative measure of the corresponding form of energy in a system and
therefore increases with increase of this energy. If, for the sake of
simplicity, to be confined to only closed systems (M = const) of a constant
composition (сk = const), this demand is met by the difference between the
theoretical compression limit Vo = 0 (this is the state corresponding to zero
of the cubic strain coordinate Θe) and the current value of the volume V, i.e.

Θp = (Vo – V). (3.2.1)

It is easy to see that Θe complies with all requirements for the cubic
strain coordinate. In particular, dΘe = – dV, and

Ψр ≡ (∂U/∂Θр) = – (∂U/∂V) = р. (3.2.2)

which provides further the concordance of signs for all potentials and forces
of a system.
In equilibrium systems, where the reversible expansion work done by a
system is the only reason of changing the volume of the system, according
to (3.2.1) đWp = рdV, which corresponds to expression (3.1.4). In more

63
general case of non-equilibrium systems the term pdV characterizes the
process of uniform compression or expansion of a system (its uniform cubic
strain) irrespective of what causes it – either the compression (expansion)
work done by the system or the energy dissipation (including that from
overcoming frictional forces or from spontaneous expansion into void).
The next challenge R. Clausius, the founder of thermodynamics, faced
on the way to representing the law of conservation of energy (3.1.3) in
terms of state parameters was to find the heat transfer coordinate as a value
remaining unvaried in adiabatic processes. R. Clausius found that
coordinate thru splitting the arbitrary cycle of the heat engine by series of
adiabatic and isothermal lines into a number of elementary reversible
Carnot cycles. Designating the elementary amounts of heat being received
and delivered in such elementary cycle at temperatures Т' and Т",
respectively, thru đQ' и đQ" gives the following form of the thermal
efficiency ηt for each of such cycles:

ηt ≡ 1 – đQ"/đQ' = 1 – Т"/ Т' . (3.2.3)

From this it follows that the sum đQ'/Т' + đQ"/Т" of the so-called
“reduced heats” đQ'/Т' and đQ"/Т" is equal to zero over all elementary
cycles, i.e. the circuit integral, in its limit, of the reduced heat

∫ đQ/T = 0 (3.2.4)

appears to be equal to zero irrespective of the cycle configuration. This

means that the integrand đQ/Т is the exact differential of some function of
state, which Clausius named the entropy:

dS = đQ/Т или đQ = ТdS. (3.2.5)

It was thereby proved the existence of a reversible process coordinate,

i.e. a state parameter which variation is a necessary and sufficient criterion
of running the heat transfer process. That statement R. Clausius referred to
the second law of thermodynamics as the law of entropy. The further
investigations (T.A. Afanasieva–Erenfest, 1928; A.A. Guhman, 1947; N.
Petrov, J. Brankov, 1986) have shown that the law of entropy comprises two
quite independent statements referred to, respectively, reversible and
irreversible processes, viz. the laws of existence and increase of entropy.
With all this going on, relationship (3.2.4) being the mathematical
expression of the second law of thermodynamics for reversible processes
has been proposed as the law of existence of entropy (A.A. Guhman, 1947).
With the lapse of time the concept of entropy has crossed the borders of
not only thermodynamics, but also physics, and penetrated into the inmost

64
of human brain. Therefore, many investigators have remained unsatisfied
with the close connection between the concept of entropy and Carnot ideal
gas cycles. Other ways have been tried for a long time to substantiate
existence of entropy. The Caratheodory’s system (Caratheodory, 1909) is
commonly considered as the mathematically strictest and logically most
consistent of all others. C. Caratheodory based his substantiation of
existence of entropy and absolute temperature on an “axiom of adiabatic
unattainability”, according to which “in any vicinity of the initial state
arbitrarily prescribed there are such ones which can not be however
accurately approximated by adiabatic state variations”. The implication of
this axiom is most distinct from the Afanasieva–Erenfest’s (1928)
statement, “Provided đQ = 0 on infinitesimal way between two infinitely
close states of a thermally homogeneous system, then none of mere
adiabatic quasi-static ways is possible as bypass between these states”. The
“bare root” of the Caratheodory’s axiom is thus growing from the evident
ground that reversible heat exchange results in such variations of state
which can not be attained by any other way as well reversible. It is easy to
see that this statement is a particular case of the process distinguishability
principle (axiom) we have embedded into the foundation of the process
classification (Chapter 2). This axiom, as applied to heat exchange process,
just logically results in a statement that a parameter exists, which variations
reflect those specific state changes distinguishable in kind and non-
reducible to each others, which the reversible heat exchange causes.
However, Caratheodory was persistent in his intent to prove the existence of
entropy not recurring to notions of non-mechanical nature (in particular, to
the notions of heat and temperature). That was what engendered a known
complexity and awkwardness of his entropy-existence substantiation. The
main point of his approach offered consists in proving the holonomy of the
expression đQ = dU + pdV, i.e. that this the so-called “Pfaffian” form
contains the integrating factor 1/T converting the heat element đQ into the
exact differential of some function of state, which Clausius has earlier
named entropy. The fact is that the heat đQ and the work đW elements as
themselves in the equation (3.1.3) of the first law of thermodynamics are
not exact differentials since their values depend on the nature of a process.
In particular, for adiabatic processes đQ = 0, for isochoric ones đW = 0.
That is what required the representation of the elementary heat in terms of
the exact differential of some function of state, i.e. per se the determination
of the heat exchange coordinate.
As we’ll try to show hereafter, the one can avoid this complexity who
classifies processes not by the energy exchange type, but by those specific
state variations the processes cause. In this case it should be admitted a
specific process exists dependable on a change of the system internal heat
energy irrespective of what causes this change, either the external heat
exchange Q or internal sources of the dissipative heat Qd (from friction,

65
chemical transformations, high-frequency or induction heating, etc1)). Let us
name this process, after K. Putilov (1971), the thermal process. Then it
directly ensues from the theorem of degrees of freedom proven in Chapter 1
that a specific coordinate of the said process exists designated as Θt in the
previous chapter. All we have to do now is to clarify the relationship
between the thermal process coordinate and the heat exchange coordinate,
viz. the Clausius’ entropy S. With this purpose let us use the equilibrium
conditions determination method, which idea belongs to J. Gibbs (1885).
Let us consider the conditions, at which the thermal equilibrium occurs
between two parts (subsystems) of a system isolated in whole. The
subsystems have intrinsically different empirical temperatures τ' and τ" and
are separated with a rigid diathermic (thermo-penetrable) partition. Since
during the process of equilibrium setting the energy of such a system
remains unvaried, the equilibrium condition according to equation (3.1.4) is
expressed as the absence of the energy variation đU for the system as a
whole at any energy variations δU' and δU" for the subsystems (which is
caused by the dynamic nature of thermal equilibrium):

δU = δU' + δU" = Ψt'dΘt′ + Ψt"dΘt" = 0. (3.2.6)

Taking into account that in the thermal equilibrium state there are no
any internal sources for the Θt' and Θt" coordinates, while the system as a
whole is isolated (Θt = const), gives that possible variations Θt' and Θt" in
the subsystems obey the evident constraint:

δΘt = δΘt' + δΘt" = 0. (3.2.7)

Considering (3.2.6) jointly with equation (3.2.7) of imposed constraints

may lead to a conclusion that at thermal equilibrium the equality of
potentials Ψt' and Ψt" in both subsystems takes place:

Ψt' = Ψt". (3.2.8)

It is known from experience that thermal equilibrium is reached under

equality of the empirical temperatures τ' and τ" in subsystems measured in
an arbitrary temperature scale (Celsius, Réaumur, Fahrenheit, etc.). Hence,
the potentials Ψt' and Ψt" are some functions of these temperatures, i.e. Ψт' =

1)
Note that from the positions of equilibrium thermodynamics, where heat has been
construed as a function of process, it would be incorrectly to qualify the “heat of a body” as a
quantitative measure of the internal heat energy Ut. However, in thermokinetics, like in
thermal physics, the heat exchange is construed as a process of exchanging the internal heat
energy Ut among bodies (confirming the principle that a system can only exchange what it
really has). This considerably facilitates the comprehension of thermal process specificity.

66
Ψt'(τ'); Ψt" = Ψt"(τ"). Since this statement is a generality and does not
depend on the kind of substances in subsystems, the above functional
dependence must be unified for all substances (universal) 1). Furthermore,
equality (4.2.7) is valid as long as heat exchange between subsystems is
possible, i.e. until the heat motion in subsystems caused the particular type
of energy exchange has ceased. From this it follows that the temperatures τ'
and τ" differ from zero as long as heat exchange exists between any
cogitable bodies. This means that the potentials Ψt' and Ψt" must be
measured in the so-called absolute temperature scale, which zero
corresponds to the total degeneracy (disappearance) of heat motion. It is
the Kelvin scale that is known to meet these requirements. Designating the
temperature in this scale thru T leads to a conclusion that the variation of
the thermal process coordinate Θt in equilibrium systems is related to heat
exchange via the same relationship as the Clausius’ entropy:

dΘt = đQ/Т = dS . (3.2.9)

From this it follows that the thermal process coordinate Θt coincides

with the Clausius’ entropy S to an accuracy of some constant (which may be
set zero, though). This allows expressing thermodynamic identity (3.1.7)
thru informal variables:

dU ≡ TdS – рdV. (3.2.10)

Fundamental difference of this identity from the joint equation of the

first and second laws of thermodynamics, where unequal sign figures
instead of identity sign (for irreversible processes), lies in the fact that the
former remains valid for also irreversible processes. Undoubtedly, in this
case the terms of this identity do not characterize any more heat exchange
and expansion work as it would be for reversible processes. The reason of
this, as indicated above, is that entropy and volume have internal sources
caused by dissipative processes running in heterogeneous systems. The
variations of volume and entropy are, in this case, caused by not only the
expansion work and the heat exchange, but by even more general processes,
viz. cubic strain and thermal process both including an irreversible
component. Such is the price that has to be paid for retaining the body of
mathematics in thermodynamics in the form of equalities.

1)
This chain of discourse makes quite evident the necessity to measure in absolute
scale not only temperature, but also pressure, chemical, electrical and any other potential of a
system under investigation. It is a matter for regret that this statement has not yet become
common property of a great number of investigators.

67
 3.3. Principle of Entropy Rise (Second Law for Irreversible
Processes)

In accordance with the law of existence of entropy the entropy means a

state parameter, which variation in reversible processes is equal to the
process heat Q related to the process average temperature T. However, the
name of this parameter given by R. Clausius (entropy in Greek means
“internal conversion”) emphasized a quite other and unusual for science of
that time property of entropy to increase also in the absence of heat
exchange (due to a spontaneous conversion of ordered forms of energy into
the heat energy). First S. Carnot in his “Réflexions…” (1824), then R.
Clausius in “The Dynamic Theory of Heat” (1850) showed by different
ways that if a heat engine was arranged so that in its reverse-direction
operation all mechanical and heat effects were converted into their inverses,
the engine would do the maximal amount of work. That meant that “the
mechanical energy thus expended might be returned to the initial state” (R.
Clausius, 1950). Thus the concept of reversibility as a possibility to return a
system to its initial state appeared immediately having taken the character of
an initial postulate. From how this term was applied, the classicists
construed it as a possibility to recover the “motive force of heat”. In
particular, W. Tomson in his article “On Dynamic Theory of Heat” directly
writes, “When heat or work is obtained with irreversible process, the
dissipation of mechanical energy occurs and its complete return to the initial
state is impossible”. Since the mechanical energy is measured by the
amount of work that a body (system) can do, the irreversibility, as the
founders of thermodynamics understood it, was a synonym for loss of
capability to do work by the body (or, as we term it now – “dissipation” of
energy).
Since all known thermodynamic systems being isolated are tending to
equilibrium (where any macro-processes cease), the capability of such
systems to do work decreases when spontaneous processes are running. To
describe such a behavior mathematically, R. Clausius considered two
“mating” heat engines – one engine working by direct cycle, while the other
– by reverse cycle. He took for granted that the thermal efficiency ηt of any
irreversible heat engine was less than in the reverse Carnot cycle (at the
same temperatures of heat absorber and heat source). If so, then the equal
sign in relationship (4.2.3) must be replaced by the unequal sign:

ηt ≡ 1 – đQ"/đQ' < ηtК = 1 – Т"/ Т' . (3.3.1)

In this case, repeating the same operations we immediately come to a

conclusion that entropy of a system increases if any irreversible processes
are running in the system:

68
 dS > đQ/Т, (3.3.2)

Thus the law of increase of entropy appeared. It reflects the unilateral

directivity of spontaneous processes and has acquired the status of the
second law of thermodynamics. According to this law the entropy of an
isolated system increases when any irreversible processes are running in it.
R. Clausius, not having conceived restrictions to this law, extended it to the
entire Universe. Such “absolutization” of the law of increase of entropy is
best of all highlighted in his winged words, “The energy of the Universe is
invariable. The entropy of the Universe is increasing”.
Clausius’s contemporaries immediately traced far-reaching effects in
that conclusion – from “the Creation” up to inevitability of “heat death” of
the Universe. A good few of investigators even at that time treated such an
extrapolation of the law of increase of entropy as extremely unconvincing
(J.M. Gelfer, 1979). It is impossible to textually duplicate here even the
least of the objections ensued. Under these circumstances the deduction of
the law of increase of entropy from the positions of thermokinetics is a
matter of great interest. The definition of the internal heat energy Uт given
above makes quite natural and predictable its increase not only due to the
heat exchange Q, but also due to internal sources of the dissipation heat QD:

dUт = đQ + đQд. (3.3.3)

Since the quotient obtained when the exact differential of any function
of state (in this case – of Ut) is divided by any parameter Ψt (in this case –
by absolute temperature T) is also the exact differential, the expression dΘт
= dUt/Т does not demand the proof of holonomy and directly leads to the
entropy balance equation I. Prigogine (1947) set forth:

dS = dеS + duS . (3.3.4)

Here dеS = đQ/Т ≤ ≥ 0 – reversible part of the entropy variation caused

by the external heat exchange đQ; duS = đQd/Т ≥ 0 – irreversible part of the
entropy variation caused by internal sources of the dissipation heat Qd ≥ 0.
Under this expression the entropy S of adiabatically isolated systems (where
đQ = 0) increases when any internal sources of the dissipation heat appear
in them.
At the same time expressions (3.3.2) and (3.3.3) manifest clearly the
inapplicability of the law of increase of entropy to irreversible processes,
where the work counter the dissipation forces affects other components of
the internal energy and does not initiate internal sources of the dissipation
heat Qd. In particular, metal cutting and material crushing processes involve

69
changing the internal potential energy depending on the system structure.
This is estimated in practice by the heat output ratio which is the dissipation
heat Qd related to the work expended W. For many processes the ratio Qd/W
is less than unit. Thus, from the positions of thermokinetics, it becomes
absolutely evident that the thermodynamic entropy is not a measure of
“each and either” irreversibility, but just reflects the fact of increasing the
internal heat energy at the expense of other “non-entropy” forms of energy.
The understanding of this circumstance is considerably facilitated by
comparing the exact differential of the internal heat energy (4.3.3) dUt = ТdS
with the similar expression for the kinetic energy dEk = vdP (where v –
velocity of a system, P – its momentum):

dUт = ТdS ; dEk = v·dP . (3.3.5)

It is easy to notice that the entropy S behaves toward the internal heat
energy Ut of a system in the same way as the momentum P of the system
toward its kinetic energy. In other words, the entropy of a system is a
thermo-momentum of a kind, i.e. the sum of momentum magnitudes of the
particles comprising the system, which has lost its vector nature due to the
randomness of heat motion. It is natural therefore that the entropy as a
thermo-momentum of a system increases not only at its transfer from the
environment during heat exchange with the same, but at any dissipation
processes involving the conversion of ordered forms of energy into the
internal heat energy. The interpretation of the entropy as a thermo-
momentum is a matter of no small consequence for conception of
thermodynamics and its further generalizations. In particular, it becomes
evident that the thermodynamic entropy can not serve as measure of the
disorder not relating to heat motion, including the disorder of structure or
uncertainty of information. In this respect the conversion of the ynergy of a
system into its anergy (see chapters 7, 8 and 14) is a much more general
criterion of an isolated system evolving toward equilibrium. Such a
conversion has an experimental confirmation, which may be, e.g. the so-
called “effect of increasing heat content measured” (L. Brovkin, 1960,
1964). Those experiments revealed an increase of the mean integral
temperature during the relaxation of gaseous or solid medium (paper,
cardboard, rubber) with heterogeneous temperature field, whereas that
temperature should seemingly have decreased. Characteristic are herein the
experiments with a roll of paper, where throughout the length a resistance
thermometer was inserted. After the tightly packed roll had been non-
uniformly heated from an external source of heat, the system was lagged
(heat insulated) and the resistance variation of such a “spread” thermometer
was recorded. Series of such experiments revealed a considerable increase
of the mean integral temperature of the roll for the initial period of its
“cooling”, which amounted to 17.2% thru 36.4% depending on the degree

70
of its heating. For rubber that effect appeared to have been even higher.
That unambiguously evidenced the fact that the equilibrium component of
the internal heat energy of the body (its anergy) Ūt = ТS increased whereas
the Ut itself remained constant, i.e. that the thermal ynergy existed and
transformed into the anergy during the relaxation of the system. Hereafter
we will time after time advert to the concept of inergy as more “physic”,
more informative and more general criterion of evolution that entropy.

3.4. Principle of Excluded Perpetual Motion of the Second Kind

Operation experience on heat engines and their theoretical investigation

have led to the comprehension that there should exist in them, along with a
heat source, also a heat absorber (or, to use the common locution – a “cold
source”). Scientists have come to this conclusion by different ways: S.
Carnot (1824) – from analogy of heat engines with hydraulic engines using
the inlet-outlet water level differential; W. Thomson (1847) – from
impossibility for working medium to be cooled below the ambient
temperature; R. Clausius (1850) – from necessity of the “compensation” for
converting heat into work as partial removal of this heat to surrounding
bodies; W. Ostwald (1901) – from impossibility of using inexhaustible heat
supply from the ocean. It was him, who offered to name engines lacking
heat absorber the “perpetual motion of the second kind” (unlike the
“perpetual motion of the first kind” lacking heat source and thus disturbing
the first law of thermodynamics). The postulates stating impossibility of
creating such engines make up an inseparable part of the second law of
thermodynamics and are united by the “law of excluded perpetual motion of
the second kind”.
However, the initial statements of this law referred to cyclic heat
engines only and did not take into account the wide process variety of
converting energy from one form into another. Therefore to many
investigators the consideration of non-cyclic and non-heat engines has
always seemed violating the “don’ts” of the mechanical theory for heat
engines. This causes numerous discussions that have periodically arisen in
scientific and popular press. In this context it is a matter of interest to clarify
the limits of validity of the said postulates from the positions of a more
general theory which thermokinetics is.
Let us consider a heat source with a temperature of Т1 capable of heat
exchange with the environment, but incapable of ceaseless mechanical
work. In this case, to do work, a working medium has to be applied to
capable of both heat-exchange and work, i.e. having both the thermal and
mechanical degree of freedom. To avoid expenditure of working medium,
let us make it perform a cyclic process (Fig.3-1). Providing such a working
medium (e.g. steam or gas) is homogeneous in its physical properties

71
(internally equilibrium), while the processes comprising the cycle are quasi-
static (i.e. not disturbing this equilibrium), the joint equation of the first and
second laws of thermokinetics (2.2.5) becomes (3.1.5). According to (3.1.4)
the work Wc of such a cycle is equal to the heat of the cycle:

Wc = Qc = ∫ TdS . (3.4.1)

From this equation it

T follows that the work Wc in the
cyclic process under
T1 consideration is other than zero
1 2 only in the case, when the
temperature T at various stages
T2 of the cycle accepts different
values. Otherwise, factoring it
outside the integral sign (3.4.1)
S1 S2 S and taking into account that the
Fig.3.1. Generalized Cycle of Heat Engine
circuit integral of entropy, like
of any state parameter, is equal to zero, gives immediately Wc=0. Thus
working medium in a cyclic engine must periodically contact with at least
two heat sources having different temperatures T1 and T2. In other words,
both a “hot” and a “cold” heat sources are needed. However, it does not
mean at all that the working medium can not take heat from both of them.
To show that the cold source must be a heat absorber, let us note there is
part 1–2 in the cycle, where the entropy increases and the heat Q1= ∫T1dS >
0, i.e. is supplied to the working medium, and part 2–1, where the entropy
decreases and the heat Q2 is removed (Q2 = ∫T2dS < 0). Thus to run a cyclic
process, both heat sources and absorbers are needed, which conceptualizes
the law of excluded perpetual motion of the second kind. It is worth noting
that this statement, as itself, directly ensues from the thermokinetic identity
in the form of (2.2.5), where the terms of the addend sum characterize
nothing but the useful work done by the system. These terms are structured
as Хi·dZi, which is a direct evidence that only heterogeneous systems (Хi ≠
0) can do useful work, while this work itself involves transfer of energy
carrier between the parts of such a system (dZi ≠ 0).
Following the chain of our discourse one can easily come to a concept
of the degree of convertibility of heat into work. If to construe Q1 and Q2 as
the heat, respectively supplied to and removed from the cycle, then
according to (3.4.1) the work of an arbitrary cycle Wc = Q1 – Q2 and is
expressed by the cyclic area, while the ratio ηt of the cyclic work Wc to the
heat Q1 supplied from a hot source is:

ηt ≡ Wц/Q1 = 1 – Q2/Q1 < 1. (3.4.2)

72
 This ratio was named the thermal efficiency of heat engine. According
to (3.4.2) it is always less than unit. This fact is sometimes erroneously
ascribed to shortcomings of heat engine leaving out of account the
circumstance that removing a part Q2 of the supplied heat Q1 to the
environment is not a loss, but the necessary provision to close the cycle.
Only that part Q2 may be considered as losses, which is removed to heat
absorber in excess of the minimum required. To define this minimum, let us
represent Q1 and Q2 as:

Q1 = ∫T1dS1-2 = T1 (S2 – S1); (3.4.3)

Q2 = ∫T2dS2-1 = T2 (S1 – S2), (3.4.4)

where T1 , T2 – the so-called mean thermodynamic temperatures of,
respectively, heat supply and heat removal in a cycle. In Fig.3-1 those are
expressed as the height of a rectangle equal in area to, respectively, the
curvilinear trapeziums S1–1–2– S2 and S1–2–1–S2. This allows expressing
the efficiency of any cycle thru these temperatures by the relationship:

ηt = 1 – T2 / T1 < 1. (3.4.5)

From here it directly follows that the efficiency of the cycle under
consideration will be maximal providing the temperatures of heat supplied
and removed are constant and equal to, respectively, maximal and minimal
temperature of the working medium in Fig.3.1. Such a cycle consisting of
two isotherms and two adiabats was first set forth by S. Carnot and named
after him. The removed heat Q2 is there minimal and equal to Q2min. Hence
only the heat difference Q2–Q2min may be referred to as losses. It also ensues
from (3.4.5) that the thermal efficiency of heat engine is defined by
exclusively mean temperatures of heat supplied and removed in a cycle and,
providing those are equal, does not depend on the working medium of the
engine1). This statement is in substance equivalent to the Carnot theorem
which he proved from the theory of thermogen and the assumption of
reversibility (ideality) of his cycle. In our case we have come to all these
statements without any assumptions regarding the cycle configuration and
equilibrium of the processes comprising the cycle. From our consideration

1)
This fact means that the value of ηt does not depend on features of heat engine and its
design perfection, but is defined by exclusively those resources that nature provides for the
human, including the temperatures of heat source and heat absorber. Therefore some
investigators have reasonably proposed to name ηt not “efficiency”, but “degree of
reversibility” of the heat supplied.

73
undertaken it also follows that the “compensation” for the Clausius-
mentioned conversion of heat into work consists in removing the heat Q2
into the environment (heat absorber) and in increasing its entropy by the
value (S2 – S1) theoretically equal to the decrease of the entropy of the heat
source. In other words, to provide the heat conversion process, it is
necessary to arrange a flow of entropy from the hot source to the cold one
like the water flow in hydraulic engines. Yet S. Carnot noted this analogy
between heat and hydraulic engines. Looking now back one can not choose
but regret that after the theory of thermogen as the “indestructible fluid”
failed this term has never been used to describe the carrier of the heat form
of motion.

3.5. Principle of Unattainable Absolute Temperature Zero

(Third Law of Thermodynamics)

This fact means that the value of ηt does not depend on features of heat
engine and its design perfection, but is defined by exclusively those
resources that nature provides for the human, including the temperatures of
heat source and heat absorber. Therefore some investigators have
reasonably proposed to name ηt not “efficiency”, but “degree of
reversibility” of the heat supplied.

Thorough investigation tests on substances to study their low-

temperature behavior were undertaken in the early 20th century. As a result
of these studies it has been found (W. Nernst, 1929) that on approaching the
absolute temperature zero the entropy of any equilibrium system in
isothermal processes ceases to depend on whatever thermodynamic state
parameters Θi and in its limit at T=0 accepts a value the same for all
system, which may be taken zero (I.P. Bazarov, 1991). This is
mathematically expressed by the relationship:

lim (∂S/∂Θi)T = 0. (3.5.1)

T→0

The validity of this statement named the third law of thermodynamics is

presently substantiated for all equilibrium systems. Apparent deviation from
this law revealed for some substances (glycerol, CO, NO, some alloys)
appeared to have been attributed to a “freezing” of them in a meta-stable
non-equilibrium state that passed off in some (sometimes very long) time.
Practical value of the third law consists in its advanced facilitation of
calculating the thermodynamic functions. Before that law stated, to
calculate entropy, it was necessary to know the dependence of heat capacity
on temperature and the thermal state equation. Now this has become
needless, since according to the heat capacity definition

74
 Cv = T(∂S/∂T)V ; Cp = T(∂S/∂T)P (3.5.2)

integration of these equations gives:

S(T,V) = ∫( Cv/T)dT ; S(T,p) = ∫( CP/T)dT , (3.5.3)

where the integration is performed from the absolute temperature zero.

Since according to the third law the entropy is finite at any temperature, the
integrals (3.5.3) must be converging. From this it follows that the isochoric
and isobaric heat capacities at Т→ 0 tend to zero faster than the temperature
(Einstein, 1966).
The availability of relationships (3.5.3) substantially facilitates
calculating the entropy of bodies, when the dependence of their heat
capacity on temperature is known.
At the same time the third law is a matter of no small theoretical
consequence in the context of its interpretation as the law of unattainability
of absolute temperature zero. From the positions of equilibrium
thermodynamics such a conclusion may be drawn taking into consideration
that on approaching the absolute temperature zero all isothermal processes
are becoming simultaneously adiabatic. Hence at T=0 heat exchange
T 1 Z1 becomes impossible and this state
Z0=0 Z2 can not be attained yet by the
reversible heat exchange. This
3 Z3 becomes even more evident taking
5 2 into account the irreversibility of
7 4 heat exchange, i.e. the provision
6 that a temperature differential
between a body and the
environment is required to provide
it. In this case, to attain the
absolute temperature zero, bodies
need to be available with a
S temperature below zero, which do
Fig.3.2. To the Law of Unattainability of not occur in nature. This is what
Absolute Temperature Zero
W. Nernst, who did not welcome
the notion of entropy, interpreted
as the unattainability of absolute temperature zero.
From here it does not yet follows the impossibility of approaching the
absolute temperature zero by whatever other way. It is possible to lower
some potential Ψi (including temperature T) by varying one of the
thermodynamic forces Xi acting in a system. This, in particular, makes the
basis for the method of attaining extremely low temperatures by the
adiabatic demagnetization (V. Sychev, 1977). This method is based on the

75
magneto-caloric effect, viz. on the phenomenon of decreasing the
temperature T of magnetic material at its magnetization Zм depressing, the
decreasing rate being defined by the derivative (∂Т/∂Н)S,p. At quite low
temperatures this derivative reaches a noticeable value for a number of
magnetic materials, which allows realizing the method offered by Debye
and Gioque in 1926. Fig.4.2 explains the process by a T–S diagram. The
diagram depicts a family of magnetization curves Zм = const. Since with the
temperature decreasing, the entropy ceases depending on whatever
parameters of magnetic material (including Zм), all curves at T = 0 converge
to the same point. In this case the character of the curves themselves
depends on the behavior of the heat capacity cp of the magnetic material
nearby the absolute temperature zero. Since with the temperature
decreasing, the heat capacity cp of a number of magnetic materials decreases
even faster, the curves at T=0 have a common vertical tangent. Let point 1
(S > 0) on the T–S diagram characterizes the initial state of the magnetic
material. Then running a single adiabatic demagnetization process (1–2) to
the state Zм = 0 we are still quite far from the state T=0. However we can
isothermally magnetize the material once again simultaneously removing
the heat herein released to an intermediate coolant (2-3) and then repeat the
adiabatic demagnetization. Iterating the process it is possible to become
asymptotic to the absolute temperature zero. Temperatures of 0.001K have
been reached by this way at present time. Thus one can only insist on
unattainability of absolute temperature zero for a single process. In this
context it is advisably to put forward one more, let us say, philosophical
sanction. The unattainability of absolute temperature zero or any other
generalized potential means the “indestructibility” of heat or any other form
of energy at all and by whatever way. It follows from here that all known
forms of energy either have always existed or were “created” for ever and
aye by a disposition of Providence so that new forms of energy can not
occur in nature. Antiscientific character of such a deduction is evident.
Therefore, from the positions of thermokinetics it is more preferable to
adhere to the ”asymptotic attainability” of absolute temperature zero as a
result of infinite sequence of adiabatic-isothermal processes (1-2-3-4-5-6-7-
etc.). From this viewpoint the heat or any other form of energy can
“degenerate” under certain conditions. For the heat form of motion featuring
a “symbiosis” of kinetic and potential energies of particles this can occur in
two events – at infinitely high compression, when kinetic energy of all kinds
of particle motion degenerates, and at infinite expansion (scattering) of
particles, when potential energy of their interaction degenerates. Both
events apparently occur in the Universe, which makes the extension of the
thermodynamic laws to the Universe an impermissible extrapolation.

76
 3.6. Principle of Entropy Maximum as Equilibrium Condition

The equilibrium self-non-disturbance principle is one of the primary

postulates of classic thermodynamics. It reads that an isolated system with
time always comes to a state of thermodynamic equilibrium, and only an
action from outside can make it out of this state. Being a result of the
experience accrued, this statement is a basis of thermodynamics as a whole
and defines the limits of its applicability.
From the viewpoint of statistical physics internal equilibrium in a
system corresponds to such a state of the incessantly moving micro-particles
of the system which occurs most frequently and thus is most probable. From
this it follows that the spontaneous transition of a system to equilibrium is
not an absolute law of nature, but expresses just the most probable behavior
of the system. Based on this principle, thermodynamics restricts the
spectrum of the systems that may be considered as thermodynamic and
leaves out of consideration the ones that develop missing equilibrium. Such
systems, in particular, include microscopic objects, which never ceasing
motion leads to their spontaneous deviation from the most probable state
(system fluctuations). Here come also systems of galactic size, where, due
to lag of interaction, instead of dying the fluctuations down their “build-up”
may arise like in the regulation systems with positive (regenerative)
feedback. Such systems become the subjects of thermokinetics allowing for
processes of “self-ordering” of a number of degrees of freedom against
disordering of other ones and never coming to the total equilibrium state.
However, this discipline is also rested on the statistical foundation of the
collective behavior of a great number of particles (collective degrees of
freedom) resulting in the macroscopic (observable) character which the
microscopic motion acquires. From this it follows that the thermodynamic
and statistical approaches are not alternative, but mutually complementary.
In accordance with its laws classic thermodynamics deals with
transitions in-between two and more equilibrium states disregarding the
kinetics of these transitions. Therefore the concept of equilibrium and
equilibrium conditions are top-priorities in this theory.
The theory of system equilibrium and stability was first developed by
Lagrange for mechanical systems. That was based on the virtual
displacement principle. It says that a mechanical system at ideal constraints
remains in equilibrium providing the sum of works done by all forces at any
virtual (possible) displacement of the system is equal to zero. That theory
was then extended by Gibbs to thermodynamic systems. The peculiarity of
that principle as applied to thermodynamics was that depending on the
conditions of the interface between a system and the environment there
were several equilibrium conditions in thermodynamics instead of one
condition for mechanical systems. In particular, the condition of stable
equilibrium for isolated systems means maximality of their entropy:

77
 δS = 0, δ2S <0 , (3.6.1)

where δS, δ2S – respectively, first- and second-order variation of entropy.

The equality of the first-order variation of entropy δS to zero means
equilibrium, while the negativity of its second-order variation δ2S
(maximality of entropy) characterizes the stability of equilibrium.
The general method of definition of equilibrium conditions set forth by
Gibbs may be instantiated by a system that has, besides the thermal and the
mechanical degrees of freedom, also some ith degrees of freedom (related to,
e.g., interchange of the kth substances between the parts of the system). Let
us consider a heterogeneous in whole system consisting of two
homogeneous subsystems (phases or zones). Designating the parameters of
these subsystems with correspondingly one or two primes let us express,
according to (2.3.6), the variations of energy δS for both subsystems as:

δU' = T'δS' – p'δV' + Ψi' δΘi' , (3.6.2)

δU" = T"δS" – p"δV" + Ψi" δΘi" . (3.6.3)

For the sake of simplicity we have only considered here one additional
degree of freedom with the coordinate Θi and the potential Ψi, which,
however, does not affect the generality of consideration. Since the system in
whole is isolated, the variations of the extensive coordinates in (3.6.2) and
(3.6.3) have the constraints imposed:

δU' + δU" = 0; δS' + δS" = 0; (3.6.4)

δV' + δV" = 0; δΘi' + δΘi" = 0. (3.6.5)

To use the principle of entropy maximality in equilibrium conditions,

let us represent equations (3.6.2)-(3.6.3), allowing for (3.6.4)-(3.6.5), in the
form:

δS = (1/T' – 1/T")δU' + (p'/T' – p"/T")δV' + (Ψi'/T' – Ψi"/T")δΘi' = 0. (3.6.6)

Since the variations δE', δV' and δΘi' caused by, respectively, possible
heat exchange, increase of volume of one part at the expense of other and by
transfer of the ith energy carrier, do not depend on each other and can take
any values, the necessary and sufficient criteria of equilibrium assume the
form:

T' = T" (thermal equilibrium); (3.6.7)

78
 p' = p" (mechanical equilibrium); (3.6.8)
th
Ψi' = Ψi" (i - kind equilibrium). (3.6.9)

However, the equilibrium conditions above defined do not allow for

some zones which may form in each of the subsystems and featuring new
properties, e.g., drop formation in supersaturated steam, graining in
oversaturated solution, etc. Additional equilibrium conditions excluding
such processes are usually called equilibrium stability conditions (R. Haase,
1964). If transition of such a kind is impossible, the equilibrium (or the
system) is referred to as stable. If the transition under consideration may
occur at only discrete (finite) variations of state, the system is relatively
stable (meta-stable). If the said transition occurs at infinitesimal variations
of state, the system is absolutely unstable (labile).
The above-mentioned maximum of entropy is a starting point for stating
the stability conditions. In this case, however, along with the absence of the
first-order variation of entropy the inequality sign appears đS ≤ 0 allowing
for unilateral state variations. In many cases, to define the equilibrium and
equilibrium stability conditions, characteristic functions (thermodynamic
potentials) appear more convenient, such as the internal energy U (at S, V,
Θi = const), the enthalpy H = U + pV (at S, Θi = const), the Helmholtz “free
energy” Γ = U – TS (at V, Θi = const), the Gibbs “free enthalpy” G = U +
pV – TS (at T, p, Θi = const), etc. Their variations form a system of 8
inequalities of the type described above (R. Haase, 1964). Since each of
these functions is the potential for only definite set of independent variables,
the choice of the corresponding function is required for a particular set of
variables. To provide this, it is necessary to address a number of mnemonic
codes facilitating storage of the equilibrium criteria relevant for particular
sets of variables. It will be shown hereafter (in chapter 9) that this difficulty
may be avoided using the inergy E of a system as a universal
thermodynamic potential. Dividing energy into inergy and anergy allows
also integrating the first and the second laws of thermodynamics and
expressing its fundamental law most concisely, “When whatever
spontaneous processes are running in isolated systems, their inergy goes
over into anergy with their sum remaining constant”.
Consideration of the fundamental laws of equilibrium thermodynamics
as deductions of the unified theory of real processes allows presenting these
laws in the most evident form. The assumptions laid in the foundation of the
real-process theory become more distinct, while the reasons of its
narrowness – more cognizable. This enables tracing out ways toward
generalization of the classic theory and overcoming the difficulties revealed.
These ways will be threaded hereafter in the consecutive chapters of this
book.

79
 The same entropy maximum is a starting point for laying down the
stability conditions. In this case, however, along with the first variation of
entropy being absent the inequality sign đS ≤ 0 appears, which considers the
possibility of unilateral state variations. In many cases, in order to lay down
the equilibrium and stability conditions, characteristic functions
(thermodynamic potentials) appear to be more appropriate. The definition of
such functions is essentially simplified with the basic equation of
thermokinetics used in the form of (2.2.5) wherein the expression Σi Хi·dZi
characterizes the elementary useful work đW the system does. Basing on
this and applying the Legendre transformation Ψi dΘi = dΨiΘi – ΘidΨi to the
terms of its first sum with consideration of (2.5.6) gives:

đW = – d(U – ΣiΨiΘi) + ΣiΘidΨi . (3.6.10)

For the particular case of a thermo-mechanical system with two degrees

of freedom (Ψ1≡ Т; Θ1 ≡ S; Ψ2 = р; Θ2 = –V) the function U – ТS + рV has
been termed as the Gibbs energy G. Then at р,Т = const

đW = – [dG]p,T , (3.6.11)

i.e. the work the system does at р,Т = const besides the expansion work, is
equal to the Gibbs energy loss. Based on this fact, the value G is also called
the isobaric-isothermal potential or free enthalpy G = H – TS (where H ≡ U
+ рV – enthalpy). With V,Т = const the isobaric-isothermal potential gives
way to the isochoric-isothermal potential Г ≡ U – TS Massieu first
introduced (1869) and H. Helmholtz called as free energy:

đW = - [dГ] V,T . (3.6.12)

Eventually, with V, S = const the role of potential is played by the

system internal energy U:

đW = - [dU] V,S . (3.6.13)

Another significant property of thermodynamic potentials is their

capability to play the role of characteristic functions. The term
“characteristic function” was introduced by Massieu (1869) – the first one
who realized the practicability of using those functions Г(T,V) in order to
find thermodynamic properties of a system as its derivatives. In particular,
the system pressure p can be found as the first derivative of the system
internal energy U with respect to the volume V (providing the entropy S
being constant), while the temperature T – as the derivative of the system
internal energy U with respect to the entropy S (at V = const):

80
 р = - (∂U/∂V); Т = (∂U/∂S). (3.6.14)

Similarly the system entropy S can be found as the first derivative of the
free energy F with respect to the temperature T, while the isochoric heat
capacity Сv – as its second derivative:

S = – (∂F/∂T), Сv = –T(∂2F/∂T2). (3.6.15)

A system at equilibrium is incapable for work. This means that the

thermodynamic potentials at this state reach their minimal value equal to
zero. This allows finding the equilibrium and stability conditions for non-
isolated thermodynamic systems proceeding from minimum of
thermodynamic potentials as a condition:

(δ2H)S > 0 ; (δ2G)p,T > 0; (δ2F)V,T > 0. (3.6.16)

Since each of these functions is the potential at only a certain set of

independent variables, each particular set of the independent variables
requires a particular function to be chosen from their total number. Taking
into account that this number reaches eight in the general case of open
systems, we are thus led to a number of mnemonic rules to facilitate
remembering the equilibrium criterion acceptable for a particular case. This
difficulty may be avoided later on (see Chapter 14) through application of
the system inergy E as a universal thermodynamic potential.
It is worth noticing in conclusion that the substantiation of the basic
laws pertaining to equilibrium thermodynamics as consequences from
thermokinetics allows to refuse their postulation and to expose the very gist
of them. This makes the assumptions laid into the foundation of this theory
clearer and the reasons of its inferiority more understandable. This allows
outlining the ways how to generalize the classic theory and to overcome the
difficulties revealed. These ways will be realized in subsequent chapters of
the book.

81
 Chapter 4

THERMODYNAMICS OF IRREVERSIBLE PROCESSES

The rise of engineering interest in phenomena at the interface between

various disciplines and perception of their close relation with the
phenomena of the energy dissipation led to creating, in the early 20th
century, the thermodynamic theory of real process rate. This field in
macroscopic physics of the 20th century was named the theory of
irreversible processes (L. Onsager, 1931; I. Prigogine, 1947, 1955; H.
Cazimir, 1945; K. Denbigh, 1951; S. De Groot, 1952, 1962; J. Meixner,
1954, I. Gyarmati, 1960, 1970; R. Haase, 1962, etc.). It has enriched the
theoretical mind of the 20th century with a number of general physics
principles and notably contributed to cognition of the in-depth interrelations
between different-type phenomena. That contribution was appreciated by
two Nobel prizes awarded in the field (Onsager, 1968; I. Prigogine, 1977).
However, the theory of irreversible processes (TIP) was constructed by
extrapolating classic thermodynamics beyond the strict framework of
applicability of its system equilibrium and process reversibility concepts.
This has led the TIP to its losing the rigor and completeness intrinsic for the
classic thermodynamic method. In this context it becomes a question of
significant importance to provide a consistent thermodynamic substantiation
of the TIP fundamentals from the more general positions of
thermodynamics.

4.1. Theory of Dissipation Process Rate in the Making

More than centenary had passed before it became clear that

“thermodynamics unaware of time” (to a Brian’s figurative locution) was
substantially thermostatics wherein only Fourier’s, Navier’s, Ohm’s, Fick’s,
Darcy’s, Newton’s, etc. equations prefigured the coming theory of non-
static (running with finite rate) processes, viz. thermokinetics. However, the
development in that direction demanded introducing in thermodynamics the
transfer concepts intrinsically extraneous for it. One of these belongs to N.
Umov (1873), who wrote the law of conservation of energy in terms of the
mass elements of resilient media as:
∂ρε /∂t + ∇⋅jeо = 0 , (4.1.1)

where ρ, ε – density of a system and its specific energy, respectively; jeо –

local density of the energy flow across the stationary borders of the system;
t – time.

82
 The flow concept in application to entropy (G. Jaumann, 1911) became
another stride. Note, that application was quite novel, because of the
statistic-mechanical interpretation of entropy as a measure of state
probability for which the transfer concept is absolutely senseless. G.
Jaumann set forth the equation of entropy balance:

∂(ρΣs)/∂t + ∇⋅jsо = σs , (4.1.2)

where the specific entropy variation rate s (J/kg⋅K) is represented as a

consequence, on the one hand, of its transfer across the system borders by
the entropy flow with the density js (W/m2⋅K), while, on the other hand, of
availability of its internal sources with the volume density σs (W/m3⋅K). A
little bit later De Donder (1927) related the entropy source σs with the rate
of the rth chemical reactions in the volume unit wr of the system and with
the affinity Ar of those reactions:

σs = Σr Ar wr /T . (4.1.3)

That was how the concepts of flow and process rate started their
introduction into thermodynamics. The most decisive move in that direction
was not, however, made, until 1931 (L. Onsager). Onsager built his theory
of physicochemical process rate (he named it “quasi-thermodynamics”)
based on the expression for the entropy generation rate, having thus
emphasized the irreversible part of real phenomena. The entropy S of a
closed adiabatic isolated system in equilibrium state is known to be
maximal. If the parameters x1, x2 ...,xn (temperature T, pressure p,
concentrations сk of kth substances, etc.) of non-equilibrium state differ from
their equilibrium values x1o, x2o,..., xno by a value of αi = xi - xio, it is
naturally to assume that the difference between the entropies of the current
S and equilibrium So states ΔS = S – S0 is a some function of α1, α2,...,αn. In
this case the reason of the ith scalar process generation (the scalar
thermodynamic force Xi) and the generalized rate of this process (named by
L. Onsager the flow Ji) could be found from the expression for the entropy
generation rate:
dS/dt = Σi(∂S/∂αi) dαi/dt = Σi XiJi , (4.1.4)

where

Xi = (∂S/∂αi); Ji = dαi/dt . (4.1.5)

Thus L. Onsager endued the force Xi with a meaning quite different

from that in Newton’s mechanics and construed it as a parameter measuring

83
the deviation of a system from internal equilibrium. At the same time L.
Onsager postulated that for minor deviation from thermodynamic
equilibrium any of the flows Ji obeyed the law of linear dependence on all
the thermodynamic forces Xj active in the system:
Ji = Σj Lij Xj . (i,j = 1, 2, …, n) . (4.1.6)

L. Onsager referred to those laws of relaxation processes, as well as the

associated coefficients Lij, as “phenomenological” (i.e. practice-based). The
off-diagonal summands in expression (4.1.6) were introduced by Onsager to
allow for various “superposition” (interrelation) effects of different-type
irreversible processes running simultaneously in the same spatial zones.
The proof of reciprocity relationships between the “off-diagonal”
phenomenological coefficients Lij and Lji (i ≠ j) was most important in the L.
Onsager’s theory:

Lij = Lji . (4.1.7)

These symmetry conditions are called the reciprocity relationships.

They reduce the number of the coefficients Lij to be experimentally defined
from n (for mere empirical description) down to n(n+1)/2 (where n – a
number of independent flows) and lead to setting up a before unknown
relationship between the rates of different-type irreversible processes. L.
Onsager was afterward awarded the Nobel Prize (1968) for his studies in
that field. Those studies attributed to non-equilibrium thermodynamics just
as much as the R. Clausius’ studies to the making of classic
thermodynamics. They have embodied the odds and ends of concepts and
facts representing them in an accessible and understandable form. Their
publication made a good start to the intensive development of the TIP in
macroscopic and statistical physics. After Onsager, H. Cazimir (1945)
extended the Onsager’s theory to cover vector processes, having herein
proved that in case the α– and β–type forces (even and odd time functions)
acted simultaneously, the reciprocity relationships (6) would go over into
the anti–symmetry conditions:

Lij = –Lji (4.1.8)

A little bit later (1956-62) I. Prigogine (the 1977 Nobel prizewinner)

based on Curie’s symmetry law showed that in case the scalar and vector
processes ran simultaneously, only the processes of the same (or even)
tensor range could interact (superimpose). Insufficiency of the Onsager’s
postulate (see 4.1.6) was thus revealed. Besides, I. Prigogine showed that
the stationary (unvaried with time) state of non–equilibrium systems was
characterized by minimal rate of the entropy generation (minimal entropy
production dS/dt) and corresponded to disappearance of the flows Jj
84
subscript–similar to the non–fixed forces Xj. He also showed that the part of
the entropy production dXS/dt associated with the force variation decreased
as having approached the stationary state. That allowed further solving a
number of problems associated with the evolution of non-equilibrium
systems.
An especially rapid development of the TIP started post–World War II.
The interest in that field of knowledge was caused not only by its general
theoretical significance, but rather its important applications having been
outlined in the forties and fifties and having involved the thermal diffusion
isotope separation, the allowance for additional terms in the hydrodynamic
equations for missile art and plasma physics, the development of membrane
technique, biophysics, etc. Due to the resumptive studies of H. Cazimir
(1945), I. Prigogine (1947, 1976), K. Denbigh (1951), S. De Groot (1952,
1962), J. Meixner (1954), I. Gyarmati (1960, 1970), R. Haase (1962), etc.
this theory has become a separate field of thermodynamics with its own
method and the certain spectrum of the problems to be solved.
Important investigations in this field have been carried out by domestic
scientists. In particular, in 1947 M. Leontovich and L. Mandelshtamm
developed a thermodynamic theory of acoustic relaxation distinguished
from the Onsager’s theory. L Landau and E. Livshits greatly contributed to
the TIP, when showed in 1951 that under the symmetry conditions (4.1.7)
the phenomenological coefficients in Onsager equation were terms of a
substantially positive matrix and therefore had the constraint imposed:
(Lij + Lji)2 < 4LiiLjj . (4.1.9)

Domestic scientists were among those, who much advanced practical

applications of that theory to various processes, viz. chemical (Bakhareva,
1967; Bulatov and Lundin, 1984), metallurgical (Veinik, 1966; Gurov,
1978), biological (Rubin, 1984; Gladyshev, 1988, etc.), as well as a
popularization of that field of knowledge (Zhukovsky, 1979; Burdakov,
1985, etc.). They especially contributed to having developed statistical
methods of substantiation and construction of the linear and non-linear
irreversible process theory (Zubarev, 1971; Stratonovich, 1985, Kvasnikov,
1987; Bazarov, 1989, etc.).
The development of the phenomenological and statistical TIP has
advanced the approximation of the heat-mass transfer theory to
hydrodynamics, electrodynamics and continuum mechanics. It appeared to
have been especially useful to study phenomena at interfaces between those
disciplines. However, those theories never did eliminate the
abovementioned line of demarcation between thermodynamics and the heat
transfer theory. The reason is that the TIP is restricted to studying the
dissipation processes like heat conductivity, electric conductivity, diffusion,
as well as effects of their superposition, but does not deal with the processes

85
of useful energy conversion, which are the main object of investigation in
thermodynamics. This is the reason why the necessity appeared to call
thermodynamics into being as a unified theory of energy transfer and
conversion rate and capacity, which, unlike W. Thomson’s “pseudo-
thermostatics” and L. Onsager’s “quasi-thermodynamics”, would not
“unfile” some part of a phenomenon, but would cover the entire spectrum of
real processes.

4.2. Motive Forces and Generalized Rates of Transfer Processes

As it follows from expression (4.1.4), the selection of the motive forces

Xi and the flows Ji in the TIP depends on how the expression for the entropy
generation rate is broken up into separate components. The only demand
herein is that the flows to be linearly independent and, at equilibrium,
become zero along with the thermodynamic forces. Such arbitrariness was
acceptable till the moment the different-type summands Xi⋅Ji characterized
the power of the same energy dissipation processes. However, that arbitrary
rule appeared to have been absolutely unacceptable in terms of the useful
energy conversion processes, e.g. in biophysics, since that would lead to an
equivocal assessment of the energy converter efficiency (see chapters 16,
17). Furthermore, expression (4.1.4) is inapplicable to evaluate the
reversible component of the motive forces, since the reversible processes
are known to have no influence on entropy (M.P. Vukalovich, I.I. Novikov,
1968; I.P. Bazarov, 1991). If, e.g., chemical reactions are described running
in an ideal fuel cell (without entropy production due to thermodynamic
irreversibility of chemical reactions), expression (4.1.4) gives zero value for
these forces. For this reason it does not allow finding the true value of the
motive forces including a reversible component. Moreover, in some
instances expression (4.1.4) does not allow defining even the sign of these
forces. If useful work is done on a system, the product of the subscript-
similar forces Xi and flows Ji appears to be always negative, whereas the
value dS/dt in (4.1.4) being always positive. Furthermore, all summands of
phenomenological laws (4.1.5) are always positive, whereas for useful
energy conversion processes they are partly negative (the flow Ji decreases
as the forces Xj being overcome increase). Besides, from relationship (4.1.7)
some relative efficiency constraints ensue, which are alien for real energy
converters. As a result of the said, the vastest spectrum of real processes
with a relative efficiency above zero appeared to have been “extra vires” the
TIP.
Thermodynamics allows avoiding this arbitrariness, when referring the
forces Xj and flows Ji to the non-equilibrium state parameters, and enables
finding them irrespectively of what causes them – either the dissipation or
useful work done. As shown in chapter 2, any vector processes arise

86
providing there is no internal equilibrium in a system, i.e. they are
heterogeneous (M.P. Vukalovich, I.I. Novikov, 1968). This allows finding
the thermodynamic forces of any nature Xi as the derivatives of the system
energy E with respect to the corresponding distribution moment

Хi ≡ – (∂E/∂Zi), (4.2.1)

while the flows Ji – as the partial time derivatives with respect to these
moments:

Ji = ∂Zi/∂t = Θivi . (4.2.2)

Such a definition of thermodynamic forces univocally represents their

local values xi in terms of negative gradients of the generalized potentials xi
≡ -∇ψi. This approach endues the thermodynamic forces with a concrete
physical meaning as intensive measure of system heterogeneity. Their
“global” (for a system in whole) analogs Xi are herein the system-averaged
values of these gradients. As shown in chapter 2, the forces Xi mean the
forces Fi in their traditional (Newtonian) interpretation per a unit amount of
the energy carrier Хi = Fi/Θi, i.e. are analogs of the specific mass, volume,
surface, etc. forces in hydrodynamics. Unlike them, the thermodynamic
forces in the TIP may be represented as ∇ψi, Т-1∇ψi, ∇(1/ψi), θi∇ψ/Т, etc.
(De Groot, 1956). This, naturally, deprives them of an explicit physical
meaning and impedes comprehension of the backbone of the phenomena
occurring.
The said appertains as well to flows that may have different
dimensionalities and meanings depending on the way the product Xi·Ji is
broken up into cofactors. Besides, flows in the TIP are often construed
(after L. Onsager) as the generalized rates of scalar processes wi = dΘi/dt,
e.g. of chemical reactions. Thermodynamics “restitutes” the general
physical conception of flow as a vector value thus allowing to distinguish it
from the generalized rate wi of scalar process not only physically, but also
analytically.

4.3. Entropy Balance Equation and Dissipation Rate

The core of the TIP comprises the balance equations for mass,
momentum, charge, angular momentum and energy of multi-component
systems, which allow further determining the entropy balance and finding
from this the motive forces and generalized rates for a system under
consideration. Setting up these balance equations is the most burdensome

87
and laborious part of the TIP and its applications, especially taking into
account that each of the equations mentioned has a local (spatial)

∂ρi/∂t + ∇⋅jiо = σi (4.3.1)

and a substantial (material) form :

ρdθi/dt + ∇⋅ji = σi , (4.3.2)

where jiо = ρθivi , ji = ρθi(vi – vо) – density, respectively, of local and

substantial flow of the field value Θi; vi , vо – transfer velocity, respectively,
of the Θi value and selected volume element in the stationary (laboratory)
coordinate system; σi – density of internal source for the value Θi.
To set up such balance equations for a particular field value Θi,
thorough knowledge of an applicable scientific discipline is necessary,
which should precede the associated TIP application. Naturally, one can
hardly expect some corrections to such disciplines would be done.
Next step toward a TIP application denotes setting up energy balance
equations – for both kinetic (translation and rotation) and potential
(mechanical, electromagnetic, chemical, etc.) energies in some form or
other (spatial or material). These equations are similar to expressions (4.3.1)
and (4.3.2) except that θi , jiо ,ji and σi are construed as specific values of a
relevant energy form, densities of their flows and internal sources (if
applicable). Since classic thermodynamics does not operate in terms of time
as a physical parameter, the above data is also entirely taken from outside,
viz. from applicable scientific disciplines. Only then thermodynamics
proper starts off with using the joint equation of the first and second laws of
classic thermodynamics for open systems in the form of the Gibbs
relationship (see 2.5.1). Using it for irreversible processes under
investigation is based on the local equilibrium hypothesis which assumes
this equation to be true locally (for continuum elements) despite the
gradients of various potentials presenting in the elements and the
irreversible processes running in them (internal sources of the entropy σs
presenting there). From joint consideration of equation (2.5.1) and the
abovementioned balance equations for mass, momentum, charge and
angular momentum the energy balance equation is formulated, which, due
to the above, now includes time and an irreversible part can be derived from
it. This part formally looks like equation (2.5.1) that contains the sources of
all actual values. This is the entropy balance equation that is formulated
based on it relating the density of the entropy internal sources σs with the
generalized rates of scalar and vector processes wi and ji and their motive

88
forces Ar and xi. Along with the entropy increase rate in a system the
dissipation function Tσs

Tσs = Σi ji ⋅xi + ΣrArwr , (4.3.3)

is often used, defining the energy dissipation rate in the system.

Next step denotes the formulation of Onsager’s (4.1.6) kinetic equations
(phenomenological laws) for particular processes under investigation. These
equations are then considered jointly with Onsager’s symmetry conditions
(4.1.7) or (4.1.8) in order to define the relationship between the flows ji and
jj arising from the static character of irreversible processes. The last
investigation stage reveals the expressions for the so-called “stationary
superposition effects” associated with disappearance of one of the actual
flows ji while the system non-equilibrium state maintained.
As it follows from the above, the superposition effect definition based
on the TIP is a quite complicated multi-stage problem involving the whole
intellectual arsenal of special disciplines. This makes thermokinetics even
more attractive as a short cut method of finding a solution to the problem.
This becomes possible due to the fact that the thermodynamic equations of
(2.3.1) type already contain time, flows and thermodynamic forces and,
therefore, do not need setting up complicated mass, charge, momentum,
energy and entropy balance equations. These equations eliminate whatever
arbitrariness in choosing the physical values defining motive forces and
generalized rates of scalar and vector processes. Their meaning and
analytical expression are unequivocally determined by choosing the
coordinate Θi as a quantitative measure of a particular energy form carrier.
In this case the thermodynamic forces, according to (2.2.8), are expressed in
terms of negative gradients of the generalized potential Θi-related, while the
flows Ji or ji – by the product of these values and the rate of their transfer
under the action of the forces Xi or xi. Applications of this method to
various problems will be specifically instantiated in the chapters following
hereafter. Other advantages of this approach will simultaneously become
evident.

4.4. Phenomenological Laws of Transfer Processes

The Onsager’s assumption that the around-equilibrium generalized rate

of a relaxation process (he named this rate the flow Ji) is a linear function of
all thermodynamic forces Xj acting in the system (L. Onsager, 1931) is one
of the postulates lying in the TIP foundation. This statement named the
linearity law is reflected in Onsager’s phenomenological laws (see 4.1.6).
Such a (matrix) form of kinetic equations differed from Fourier’s,
Ohm’s, Fick’s, Darcy’s, Newton’s, etc. laws by the presence of additional

89
(off-diagonal) terms (with subscripts j ≠ i). L. Onsager introduced those
terms to allow for the interrelation between different-type phenomena,
which he had found experimentally and explained by the “superposition” of
different-type flows. He expressed that interrelation in his well-known
“reciprocity relationships” which stated the symmetry of the
phenomenological coefficients matrix (see 4.1.7).
L. Onsager obtained those relationships based on a known statement of
statistical mechanics regarding the reversibility of micro-processes in time
under the assumption that the coefficients Lij and Lji were constant, while
the subscript-dissimilar flows Ji and Jj were linearly independent and
became zero with disappearance of the forces Xi and Xj.
The above postulate was not objected to for more than a half a century
and was reproducible in all study guides on irreversible process
thermodynamics but with a proviso that according to Curie’s symmetry law
only processes of the same (or even) tensor order could be interrelated. The
question did not anyhow arise in that context how the flows Ji and Jj found
to (4.1.4) as the time derivatives of the system independent state parameters
dαi/dt could be interrelated. There were no objections either to the fact that
some (moreover the simplest) equations appeared to have been laid as a
foundation of the thermodynamic theory which was known to have
imported particular data on system properties (expressed in terms of state
and transfer equations) “from outside” as uniqueness conditions of a kind
while its body of mathematics itself was universal as based on the properties
of exact differential of a number of functions of state. The investigators
were not at all embarrassed either by the fact that the above postulate
undermined the centuries-old buttress of mechanics which stated that each
independent process (movement, acceleration, tending to mechanical
equilibrium, etc.) was associated with the only (resultant) force and ceased
as soon as the force disappeared. Meanwhile, the existence of such a force
followed from those equations of anisotropic heat conductivity and electric
conductivity which, to Onsager’s confession, prefigured his
phenomenological laws (1). In fact, for anisotropic heat conductivity and
electric conductivity the motive forces Xj were components of the only
force – the negative temperature gradient vector ∇Т and the electric field
intensity E (j = 1,2,3), respectively. There were also other reasons to doubt
the Onsager’s postulate was adequate to the point. L. Onsager and his
followers, based on equations (4.1.6), derived numerous thermo-
mechanical, thermo-electrical, thermo-diffusive, etc. effects from the
interrelations between the generalized rates of irreversible processes ran in
the same spatial zones, i.e. from the “superposition of flows” or their
“mutual entrainment” (N. Bulatov, A. Lundin, 1984). Meanwhile, the above
effects are known to reach maximum in the so-called stationary states, when
the flows not fixed by an external constraint disappear and, therefore,
evidently can not superimpose on those remained. E.g., in electrolytic

90
solutions, where electric conductivity and diffusion phenomena take place,
the electric potential difference (Kwinke effect) is maximal whenever
current ceases (R. Haase, 1967). This is the case with the Soret effect –
arising gradient of the kth substance concentration in a system initially
homogeneous with the temperature gradient building up, the concentration
gradient reaching maximum as soon as diffusion flows disappear. Hence the
reason of such superposition effects should have been searched from the
very first not in the interaction between independent flows, but in the
superposition of unlike forces as it takes place in mechanics and
electrodynamics. Furthermore, the laws (see 4.1.6) L. Onsager named
“phenomenological” (i.e. practice-based) are really not so. First of all, the
coefficients Lij in Fourier’s, Ohm’s, Fick’s, etc. laws depend on the state
parameters of a system (its temperature, pressure, composition, etc.), i.e. are
inconstant, whereas the requirement the coefficients Lij to be constant is a
substantial part of Onsager’s linearity law and bears a principal character. In
fact, the forces Xi in the Onsager’s theory are functions of the system
parameters (temperature T, pressure p, concentration ck, etc.), therefore the
dependence of the coefficients Lij on them would mean their dependence on
also the forces Xj, i.e. non-linearity of the phenomenological laws (4.1.6).
Furthermore, according to practice many phenomena arise only
when the force reaches some “threshold” value Fjо depending on the
activation energy for a process under consideration (see chapter 2).
However, this circumstance can not be allowed for in Onsager's laws since
in this case the flows Ji would disappear earlier than the forces Xi become
zero. Lastly, the terms of equation (4.1.6) not necessarily always have the
same sign – in many cases the process rate (flow Ji) decreases with increase
of the force Xj “being overcome” in this process. So are, as it will be
hereafter shown, the phenomenological laws of all processes involving
useful energy conversion.
All the said means that Onsager’s kinetic laws are not
phenomenological, while his postulate itself does not hold water. This
nothing but enhances interest toward finding their allowable form
proceeding from the demands thermodynamics imposes on the uniqueness
conditions of transfer processes. It is easy to show then that the so-called
Onsager’s phenomenological laws are just a particular case of
thermodynamic kinetic equations (2.7.10), when the coefficients Кijср do not
depend on the variables Θj,Fj. As a matter of fact, introducing the
thermodynamic force Xi ≡ Fi/Θi instead of Fj and assuming Fjо = 0 gives:

Ji = Σj Lij Xi, (i, j = 1, 2, ..., nj) (4.4.1)

where Lij = ΘjКij – constant coefficients L. Onsager named

“phenomenological”; ji – density of the flow Ji; xj – specific (per unit of the

91
Θj value being transferred) thermodynamic forces Xi in their “ energy”
representation (in terms of generalized potential).
The same laws may be expressed as functions of the flows Jj и jj :

Xi = Σj RijJj or xi = Σj Rijjj , (4.4.2)

where Rij – resistance coefficients inverse to the conductivity coefficients

Lij.
As one can see, they actually integrate the thermal Θj and the kinetic Кij
factors. This is the reason why the coefficients Lij or Rij can be referred to
neither state parameters nor mere kinetic coefficients (R. Haase, 1967, S.R.
De Groot, R. Mazur, 1964, etc.). The values of these coefficients for a
particular system are determined experimentally or, as the simplest case,
based on statistical theories.
Representation of the phenomenological laws in the form of (4.4.2)
allows generalizing the concept of inertia to non-mechanical processes. If to
write Newton’s law F = Mа = dP/dt in the same form of (4.4.2) substituting
F/M for xm and the acceleration а = dv/dt for jm (i.e. to represent the
acceleration of a body as the generalized rate of its momentum variation
process), then Newton’s second law will appear to be a particular case of
phenomenological laws (4.4.2), where the off-diagonal coefficient Rji = M.
In this case the mass M as a measure of inertia of a system takes a meaning
of one of the coefficients of resistance the system offers to running a
relevant process (in this case the acceleration process). Thus the concept of
inertia as a measure of resistance a system offers to a process running in it
acquires the general physics meaning.

4.5. Reciprocal Relation in Transfer Processes

As it has already been mentioned above, the substantiation of

reciprocity relationships (4.1.7) was one of the most important statements of
the Onsager’s theory of irreversible processes (TIP). To substantiate them,
he had to apply to the fluctuation theory, microscopic reversibility principle
and additional postulate on linearity of the fluctuation damping laws. These
three statements outstep the framework of thermodynamics. Therefore in
the phenomenological TIP reciprocity relationships (4.1.7) are usually
adopted as an additional primary postulate sometimes named (after D.
Miller) the “forth law of thermodynamics”.
At present statements often and often occur that these relationships do
not need any thermodynamic proof since they are statistically substantiated.
Discussions about the L. Boltzman’s and J. Gibbs’ statistical theories,
“which grounds are obscure in many ways” (R. Cubo, 1970), are seemingly
sunk into oblivion. So, one of the basic statements of statistical mechanics

92
regarding the system ergodicity 1) has remained a hypothesis so far. By the
early nineties of the 20th century as a result of the critical analysis conducted
on the body of mathematics of statistical mechanics, as well as a result of
digital experiments on high-performance computers, it became clear that
only hypothetical systems of non-interacting particles might be ergodic.
Interaction between particles (e.g. Coulomb or Van der Waals forces) leads
to losing ergodicity, therefore the real systems of interacting particles
should be described not by statistical, but dynamic methods.
The fact that these reciprocity relationships often vindicate their validity
far beyond the restrictions imposed by the method itself of their
substantiation evidences the means of proving the relationships L. Onsager
offered are inadequate to the point. In fact, the microscopic reversibility
principle with one of its statement about equal rates of any direct and
reverse molecular processes is valid, strictly speaking, for only equilibrium
states. This principle is indubitably inapplicable to the processes of transfer
from one non-equilibrium state to other because a system returns to
equilibrium state just because the frequency and amplitude of the micro-
processes running toward equilibrium dominate. Therefore the reciprocity
relationships, providing these really vindicate the principle above, should
have been valid in close proximity to the equilibrium state and violated
more and more as the system was withdrawing from it. Meanwhile, as it
will be instantiated for the helium and argon interrelated filtration and
diffusion processes in graphite membranes, to vindicate the reciprocity
relationships, just a partial equilibrium is enough, when just a part of the
processes running in a system ceases.
Another weak point has been an assumption that the microscopic laws
of fluctuation damping obey the same linear phenomenological laws (4.1.6)
as the macroscopic processes of heat, substance, charge, etc. transfer.
Meanwhile, the real laws of fluctuation damping demonstrate rather
exponential character. Even kinetic equations (5.1.6) are, strictly speaking,
non-linear. This is especially evident for Fourier’s, Ohm’s, Fick’s, Darcy’s,
Newton’s, etc. laws written in the integral form (in terms of the temperature
T, pressure p, concentration ck, etc. differentials), when the proportionality
factors appear to be dependent on the temperature, pressure, concentration,
etc. fields.
Lastly, should the linear laws be really necessary to prove the
reciprocity relationships, the latter would be violated each time whenever
the phenomenological laws cease to be linear. However, as it will be shown
hereafter, the reciprocity relationships may be valid as well for the systems,

1)
The system is referred to as ergodic, where space averaging of a physical value gives the
same result as time averaging.

93
where only the off-diagonal terms of equations (4.1.6) describing the
superposition effects are linear. As a result, both the Onsager’s theory and
TIP in whole fail to reach those rigor and completeness which are intrinsic
for the classic thermodynamic method. Therefore C. Truesdall (1975) was
positively right, when confirmed that “once the reciprocity relationships
are valid, then the possibility of their merely phenomenological deduction
should as well exist”. Let us show that the Onsager’s symmetry conditions
ensue directly from the differential relationships of thermodynamics. It
follows from independence of the second derivative of the system energy U
with respect to the coordinates Zi and Zj:

(∂Zi/∂Хj) = (∂Zj/∂Хi). (4.5.1)

These equalities remain valid as well after deriving the total derivatives
of both parts with respect to the time t:

d(∂Zi/∂Хj)/dt = d(∂Zj/∂Хi)/dt . (4.5.2)

In the stationary redistribution processes (at Θj, Хi = const) the time

differentiation involves only the coordinates Zi, Zj. Their time derivatives in
the absence of the transfer (recharging) and reorientation processes (Θi, φi =
const) define the flows Ji and Jj. Allowing for this and changing in (4.5.2)
the sequence of differentiation with respect to the time t and forces Хj, Хi,
gives:

(∂Ji /∂Хj) = (∂Jj/∂Хi). (4.5.3)

These equations may be well referred to the differential reciprocity

relationships between flows and forces (V. Etkin, 1991, 1999). They state
the cross-impact balance among different-type flows and their motive forces
and therefore may be rather named the reciprocity relationships than the
symmetry conditions of phenomenological coefficients matrix (4.1.7).
Risselberghe (1962) was the first who postulated the existence of such-type
relationships and proposed to name them the “generalized reciprocity
relationships”. This is justified since the above Onsager’s symmetry
conditions may be obtained as a corollary of these relationships for a
particular case of linear systems. In fact, applying (4.5.3) to equations
(4.1.6) one can obtain:

(∂Ji /∂Xj) = Lij = (∂Jj/∂Xi) = Lji . (4.5.4)

Thus the Onsager’s reciprocal relations (symmetry conditions) ensue as

a corollary of more general differential relationships of thermodynamics and
94
do not need involving whatever statistic-mechanical considerations. Though
attempts to substantiate these relationships were made time and again
(Gyarmati, 1958, 1960; Li, 1058, 1988; Pitzer, 1961; Risselberghe, 1962),
they failed till the lacking coordinates of redistribution processes were
introduced. Note, relationships (4.5.3) do not impose any constraints on the
process irreversibility degree and the system remoteness from equilibrium.
They do not depend either on particular form of the state or transfer
equations and, as it will be shown hereafter, are equally applicable to
reversible and irreversible processes.
It is worth particular mentioning that from the positions of
thermodynamics Onsager’s symmetry conditions (4.1.7) are valid as well in
the case, when only additional (off-diagonal) terms of equations (4.1.6) are
linear, i.e. only the “cross” coefficients Lij or Lji (i ≠ j) are constant. This
significantly extends the applicability of the differential reciprocity
relationships once the diagonal terms of the matrix comprising the
phenomenological coefficients Lii or Ljj may remain herein any functions of
the thermodynamic forces Xi or Xj subscript-similar to them. This is what
Fourier’s, Fick’s, Ohm’s, Darcy’s and Newton’s laws are. As for the
additional (off-diagonal) terms, these describe less order infinitesimals and
in a number of cases may be considered as linear. Thus approaching the
problem from the positions of thermodynamics reveals excessiveness of the
demand the coefficients Lii or Ljj in equations (4.1.6) be constant, which is
substantial in all preceding theories of irreversible processes. The demand
of close proximity of the system to equilibrium Onsager has laid down in
his theory appears to be the same excessive. In fact, the symmetry
conditions (see 4.1.7) ensuing from the differential reciprocity relationships
(see 4.5.3) are valid for arbitrary large values of the forces Xi or Xj in the
diagonal terms of the phenomenological laws. This is what explains why the
Onsager’s theory appears to be applicable far beyond the restrictions
imposed by the method itself of their substantiation.
Finally, according to (4.5.4) symmetry conditions (4.1.7) may give
place to the anti-symmetry conditions Lij = –Lji (named the Cazimir’s
relationships in the TIP) providing the terms in equation (4.1.6) have
opposite signs. This occurs, when in a transfer process the flow Ji or Jj is
directed against the forces Xj or Xi, i.e. “overcomes” them. Such are, in
particular, all energy conversion processes, where energy source works
against loads. This manifests itself in the occurrence of the so-called “no-
load conditions”, when with increase of the force Xj or Xi being overcome
the flow Ji or Jj becomes zero. A demonstrative example to the case is loss
of the secondary current in the welding transformer with arc extinction. The
above statement as well outsteps the TIP, where laws (4.1.6) were
postulated by Onsager with only the same sign for all terms.
Thus the reciprocity relationships do not need assumptions on close
proximity of a system to equilibrium, constancy of all phenomenological

95
coefficients and linearity of the fluctuation damping laws, which are laid in
the foundation of their statistic-mechanical substantiation. According to the
above, they ensue from the first laws of thermodynamics and, therefore,
feature universal character. This enhances the heuristic value of these
relationships and makes them a reliable tool to analyze the interrelations
between different-type processes in the real world.

4.6. Law of Minimum Entropy Production

The insufficiency of formalism intrinsic for the linear TIP has caused
numerous attempts to give other – variational statement of non-equilibrium
thermodynamics. Calculus of variations is a quite universal mathematical
method that allows, subject to a suitable model selection, detailing a process
under investigation and deriving the associated math laws (including the
transfer equations) based on the only variational principle. However, these
principles themselves need their substantiation, while their corollaries each
time have to be experimentally checked.
L. Onsager set forth the first variational principle of non-equilibrium
thermodynamics (1931). He took note of the fact that for the dissipation
processes (Jj = Jjр) his phenomenological laws

Xi = Σj Rij Jjр . (i, j = 1, 2, …, n) , (4.6.1)

(where Rij = Lij-1 – resistance coefficients inverse to the phenomenological

conductivity coefficients Lij) along with the reciprocity relationships

Rij = Rji (4.6.2 )

are equivalent to the statement that the expression ΣjXidJjр is the exact
differential of some potential function Ф(Jiр ,Jjр) = ½ Σj Rij Jiр Jjр, the first
derivatives of which with respect to the flows Jjр give linear
phenomenological laws (4.6.1), while the mixed derivatives (∂2Ф/∂Jiр∂Jjр)
give reciprocity relationships (4.6.2). The function ПX (Xi ,Xj) = ½ Σj Lij Xi Xj
features the same properties, which leads to phenomenological laws (4.1.6)
and reciprocity relationships (4.1.7). In linear systems the functions Ф(Ji ,Jj)
and ПX (Xi ,Xj) are equal to half the dissipation function P =Тσs

ΦJ (Jiр, Jjр) = ПX (Xi , Xj) = ½ P (4.6.3)

and are named the local dissipation potentials (expressed in terms of flows
and forces, respectively).

96
 It is easy to make sure that the variation of the potentials ΦJ (Jiр, Jjр) and
ПX (Xi ,Xj) with respect to flows at all forces Xj being constant gives the same
value of Xj so that the extreme condition is met:

δ[Y(Xj ,Jjр) – ΦJ (Jiр, Jjр)]Xj = 0, (4.6.4)

where δ – variation symbol; the subscript to the function being varied

denotes the value remaining constant at variation.
The extreme determined by (4.6.4) can be only the maximum owing to
positive determinacy of the dissipation potentials. Therefore L. Onsager
named statement (4.6.4) the “principle of least energy dissipation”.
Another form of the same principle was later offered by I. Gyarmati
(1974) for the functional Υ(Xj ,Jjр) - ПX (Xi ,Xj) expressed in terms of forces at
constant flows:
δ(Y – ПX ) = Σj (Jjр – ∂П/∂Xj)δXj = 0. (4.6.5)

I.Gyarmati further merged both forms of that principle in a single

variational condition:

δ(Y – ФJ – ПX ) = 0, (4.6.6)

where variation may be conducted with respect to both forces and flows.
This condition reflects the extremity of the so-called “Onsager-Machlup’s
function” (the expression in brackets) and includes all statements of
Onsager’s “quasi-thermodyanmics”.
A number of other variational principles convenient for particular
applications were offered by G. Tzigler (1966). As I. Bakhareva then
showed (1967), all the said variational principles could be obtained by
simple transformations of the same expression:

Σj(∂ФJ / ∂Jjр – Xj)δαi = 0. (4.6.7)

This eliminates the necessity to provide thermodynamic substantiation

for each of the principles mentioned – it is enough to substantiate the
validity of condition (4.6.7). This may be done based on some (linear or
non-linear) phenomenological laws. In fact, whereas differential reciprocity
relationships (4.5.3) do not depend on character of these laws, the
differential form

Σj Jjрd Xj = d ПX (4.6.8)

is always the exact differential of the dissipation function П. Applying the

Legendre transformation, it can be easily shown that the expression ΣjXjdJjр

97
is as well the exact differential of other function of state of the non-
equilibrium system Ф. From this it follows that ∂Ф/∂Jjр = Xj. Expression
(4.6.7) is thereby valid with the same generality degree, i.e. not only for
linear systems, but also in a more general case, when the differential
reciprocity relationships are valid. In this case P = ФJ + П even when ФJ ≠
ПX.
The law of minimum entropy production occupies a special place
among the TIP extreme principles. Its first statement belongs to I. Prigogine
(1947) and refers to discontinuous systems. According to the Prigogine-
proven theorem the minimal production of entropy σs = min in a stationary-
state system with the constant forces X1, X2,…,Xk (k < m) corresponds to
the state, where flows with the indices j = k+1, k+2,…, m disappear. The
alternative statement of this law for continuum systems reads that in the
stationary state compatible with the external constraints the dissipation
function P is minimal:

Р = ∫Tσs dV = Σj ∫ Xj Jj рdV = min; δР = 0. (4.6.9)

This statement is usually substantiated involving linear

phenomenological laws (4.4.1) and the Onsager’s reciprocity relationships
and is, therefore, considered valid for only the systems being in close
proximity to equilibrium. From the positions of thermodynamics this law
acquires a more general character. According to the evolution criterion (see
3.6.6) it is quite evident that if an external force prevents a system from
reaching equilibrium, the system comes to a halt in the minimal-dissipation
state. Using flow balance equation (2.4.8) the condition of inergy constancy
for a stationary-state system may be written in the form:

dUi/dt = –∫ xi·(jiе + jiр)dV = 0, (4.6.10)

From this it directly follows that if some of the external flows jiе
supporting the stationary state (xi = const) disappears (jiе = 0), the subscript-
similar relaxation flow jiр also disappears along with the associated
component of the dissipation function P. This statement does not depend on
character of the phenomenological laws and is so evident that hardly needs
some theorems to be proved.
Thus all basic statements of the current linear TIP may be obtained as
deductions of thermodynamics without appeal to whatever postulates and
hypotheses. This makes superfluous formulating complicated balance
equations for mass, charge, momentum, energy and entropy, which has
always been the most time-consuming part of the theory. It is not less
important that such an approach excludes the necessity for preliminary and
thorough knowledge of a number of fundamental disciplines to formulate
the balance equations. This allows one to solve problems being not

98
overburdened with hardened paradigms and dogmas associated with these
problems. Furthermore, this approach opens up new possibilities to
subsequently overcome the profound narrowness of non-equilibrium
thermodynamics that is restricted to only linear systems and close-to-
equilibrium states.

Chapter 5

HEAT-MASS TRANSFER THEORY

The development of the theory of irreversible processes (TIP) has

created the prerequisites for a cardinal rapprochement between
thermodynamics and the heat-mass transfer theory. Firstly, this theory has
from the very beginning operated with the heat-mass transfer concepts:
fields (both stationary and non-stationary), energy flows (including heat),
gradients of potential (temperature), uniqueness conditions, etc. Secondly,
thermokinetics is based on the equations of a more general form, from
which the equations of heat conductivity, electric conductivity, diffusion,
etc. ensue as a particular case. Thirdly, it also uses the balance equations for
extensive values and expresses their sources in terms of the measurable
parameters of a system. In a word, the TIP already contains the whole
arsenal of tools necessary to describe and investigate the heat-mass transfer
processes. Thermokinetics spreads still further and allows obtaining all
statements of this theory essentially from thermodynamic. Therefore the
TIP fundamental principles, laws and equations are most advisable to be
stated from the positions of thermokinetics.

5.1. Elimination of Strange Delimitation of Thermodynamics

and the Heat Exchange Theory

No sooner had the heat theory appeared, it immediately separated into

two directions. In 1822 a known J. Fourier’s work appeared, which laid the
foundation of the heat transfer theory; in 1824 not less famous S Carnot’s
work laid the foundation of thermodynamics. Both works were based on the
afterward-rejected theory of thermogen as the indestructible fluid, both
considered temperature as some potential which gradient conditioned the
heat transfer direction or conversion of heat into ordered forms of energy.
However only Fourier operated with time as a physical parameter, and that
left an imprint on the whole further development of those theories. The

99
concept of entropy introduced by R. Clausius in 1850-1865 as the
coordinate of reversible heat exchange and the S. Carnot’s method of
reversible cycles widely used in thermodynamics connected
thermodynamics for a long time with the concept of equilibrium and quasi-
static character (infinitesimal rate) of processes under investigation as the
condition of their reversibility. A lot of time had passed before it became
evident that “timeless” thermodynamics (to Brayan’s locution) was
substantially thermostatics.
Meanwhile the J. Fourier’s ideas were running their course. Yet in 1822
the L. Navier’s work appeared having laid the foundation of hydrodynamics
of viscous liquids, G. Ohm derived his famous law in 1827, A. Fick set
forth the diffusion equation in 1855. Those equations and the like described
kinetics of momentum, charge, substance, etc. transfer. However both
directions mentioned developed quite independently without any points of
contacts. Their difference showed not only in terminology, but it rather
rooted in a basic methodological nature. Carnot-Clausius’ thermodynamics
maintained aloofness from the transfer ideas and the heat exchange rate
concept. The heat exchange theory, on the contrary, had nothing to do with
the conversion of heat into other forms of energy and considered entropy as
an extraneous concept. The so “fancy separation of two directions within
the same area in macroscopic physics” (to K. Denbigh’s locution) was too
hard to be overcome by the TIP as well. Even today the definition of heat
concept remains different in thermodynamics and the heat exchange theory.
In thermodynamics this is the part of energy exchange caused by
exclusively the temperature difference between bodies and not associated
with substance exchange between them (Thermodynamics. Terms, 1973).
The heat exchange theory, on the contrary, considers heat as the part of
internal energy associated with random motion (because a system can
exchange just what it has) and studies, along with heat conductivity, the
heat transfer carried out by substance and enabled by heterogeneity of the
fields of other physical values (Heat Transfer. Terms, 1980). Such a
situation demands searching more cardinal means to unify the two said
fundamental disciplines.

5.2. Conductive Heat Exchange (Heat Conductivity)

The heat exchange processes may be classified into three classes:

conductive, convective and radiant heat transfer. The conductive heat
transfer (heat conductivity) features the absence of observable
(macroscopic) motion in a heat conductive medium with direct contact
between bodies or parts of a body having different temperatures. The basic
law of heat conductivity was stated, as it has been already mentioned, by J.
Fourier having proceeded from interpreting heat as the indestructible fluid.

100
According to this law the heat flow density jq is directly proportional to the
temperature gradient ∇Т reversed in sign:

jq = –λ∇Т, (5.2.1)

where λ (W/m⋅K) – proportionality factor named coefficient of heat

conductivity.
Classic thermodynamic interprets the heat flow as a conductive (not
associated with the substance transfer) flow of the internal energy u with the
density jq. Here the coordinate of a heat transfer process is construed as the
entropy S. Therefore it is more correctly, from the positions of
thermokinetics, to refer to rather the entropy flow with the density js = jq/T
than to a heat flow. In this case Fourier’s law for stationary heat
conductivity directly ensues from kinetic equations (2.6.11) written down in
the local form:

js = Lsxs , (5.2.2)

where xs = –∇Т – motive force of the process; Ls = λ/T – coefficient that

should be rather named coefficient of entropy conductivity. Note that the
“thermokinetic” form of heat conductivity law representation (see 4.3.1) is
much closer to the original form (see 4.3.1) than the one ensuing from
Osager’s laws, where the thermodynamic force is used in the form of xq = -
Т-1∇Т or xq = lnT even when choosing the heat flow jq as the generalized
rate of the process.
The differential equation of non-stationary heat conductivity ensues
from thermokinetics equally as directly. It is enough just to use the balance
equation for the arbitrary field specific value θi of generalized form (4.3.1)
assuming there θi ≡ u and σi ≡ σq/Т. Using the definition of the specific heat
at constant volume (isochoric) сυ ≡ ∂u/∂Т in the absence of mass transfer
and work (i.e. at constant ρ и сυ) one can find:

ρсυ∂Т/∂t + ∇⋅jq = σq . (5.2.3)

This expression is the most general form of the differential equation of

heat conductivity valid for both linear and non-linear processes. It is
available in a more ordinary form if considered jointly with (6.2.1)
assuming constant λ and introducing, to shorten the formula, the thermal
diffusivity α q = λ/сυρ:

∂Т/∂t = α q∇2Т + σq/ρсυ . (5.2.4)

This equation relates the heating rate of a body directly proportional to

the temperature field and heat sources in it. The latter is very important as
101
allowing for chemical reactions running in a heat-conductive medium (see
4.1.3), high-frequency and induction (eddy-current) heating, friction, etc.,
i.e. the availability of any entropy sources σs.
Just like other differential equations of thermokinetics, expressions
(5.2.3) and (5.2.4) describe the entire group of the like phenomena. To
extract a process under investigation from this entity and to completely
describe it mathematically, it is necessary to take into consideration
particular properties of the object. These are defined by uniqueness
conditions including for a general case of non-stationary processes the
initial, geometrical, physical and boundary conditions (including the initial
temperature distribution, body configuration and size, body thermo-physical
properties, surrounding relations, etc.). The problem thus stated is solved by
either analytical, or numerical, or theoretical-experimental methods with an
appeal to the heat exchange phenomena similarity theory. These methods
are considered in special disciplines (A. Lykov, Y. Mikhailov, 1963).

5.3. Convective Heat Exchange

Convective heat exchange (or heat transfer) is usually understood as a

heat exchange between a moving medium and a stationary one without a
substance exchange between them. This process is always accompanied by
heat conductivity dominating nearby the stationary surface. Therefore the
basic law of convective heat exchange may be derived based on the same
kinetic equations of thermokinetics in the form of (6.2.2) substituting herein
the coefficient of heat conductivity λ for some empirical value α multiplied
by the thickness δт = ∆n of the thermal boundary layer of liquid (i.e. the
liquid layer within which the liquid temperature varies from the wall
temperature Тс to the temperature Tw of the bulk liquid flow). Taking into
account that in stationary conditions Тw = Тw(n), i.e. depends on only the
normal coordinate n to the heat exchange surface, and substituting on this
basis ∇Т ≡ ∂Тw/∂n for dТw/dn expression (5.2.1) may be represented thru
separation of variables in the form:

dТw = ( jqδТ /αq)dn. (5.3.1)

Integrating this with respect to δт one can find:

jq = αq ∆Т, (5.3.2)

where α = λ/δт (W/m2⋅К) – proportionality factor having been referred to as

heat transfer coefficient; ∆Т = (Тw – Тс) – thermal head. The further problem
comes down practically to the definition of how this coefficient depends on

102
various factors, viz. liquid properties, liquid flow pattern, body
configuration and streamline, etc.
Differential equation of convective heat exchange may be found in the
same way as expression (5.2.3). It is only necessary to take into account that
the temperature of a moving liquid is a function of not just time, but also the
coordinate r of this liquid, i.e. Тw = Тw(r, t). Therefore its total variation in
the time domain dТw/dt, along with the local variation ∂Т/∂t, includes the so-
called convective variation (∂Т/∂r)dr/dt = ∇Т⋅v, defined by the flow velocity
v. Allowing for this component equation (5.2.3) assumes the form:

dТw/dt = аq∇2Т + σq/ρсv . (5.3.3)

This equation is usually named differential equation of energy in order

to distinguish it from the other differential equation characterizing the heat
exchange conditions at the wall–moving liquid interface. To derive this
equation, let us heed the fact that there is a thin layer of stationary liquid
nearby the wall (n = 0), wherein the heat transfer is provided by exclusively
heat conductivity. Therefore, along with (53.2), one can write jq = – ∇Тn = 0,
wherefrom it directly follows:

αq = –λ∇Тn=0 / ΔT. (5.3.4)

This is the differential equation of heat exchange manifesting that

convective heat exchange so many times exceeds heat conductivity in
intensity, as temperature gradient in the boundary layer exceeds thermal
head in value.

5.4. Radiant Heat Exchange

Radiant heat exchange is usually understood as a heat exchange thru

radiation and absorption of electromagnetic waves within thermal-frequency
range by bodies separated with a transparent medium. This range is small
and corresponds to waves with a wavelength of 0.8 micron thru 0.8 mm.
Generally speaking, the radiation is an ordered (directional) process
pertaining to one of the useful work categories. This is confirmed by the
fact that the absorption of the radiant energy is accompanied by
photosynthesis of organic compounds, ionization, dissociation, photo-effect,
and the like phenomena. Therefore it makes sense to refer to the radiant heat
exchange providing a radiant energy dissipation process occurs. This
dissipation of electromagnetic energy takes place, generally speaking, not
only in bodies themselves, but also in the medium separating them.
Therefore it should be advisably considered as a “photon heat conductivity”

103
phenomenon (G.F. Muchnik, I.B. Rubashov, 1974). Thus the differential
form of the radiant heat exchange law may be obtained based on the same
kinetic laws of thermokinetics in the forms of (2.6.10) and (2.6.11). In the
last case the law may be written as:

jq = λf ∇Т, (5.4.1)

where λf – coefficient of “photon heat conductivity”. For optically dense

media this coefficient is defined by the Henzel equation:

λf = 16 kб Т 3/3æг , (5.4.2)

where æг – absorption factor of a medium.

For the stationary heat exchange ∇Т may be represented, as shown

above, in the form of the fraction dТ/dn. Separating on this basis the
variables in (5.4.1) and integrating this expression with respect to the ray
path length ℓ in-between the bodies with temperatures Т1 and Т2, one comes
to the radiant heat exchange law expressed as:

jq = [kб/(1+ æгℓ)] (Т14 – Т24) . (5.4.3)

This expression is a good approximation for media with low optical

density (weak absorption).
Thus all integral and differential equations of the heat exchange theory
may be derived from thermokinetics. This is what provides a
methodologically unified approach to thermodynamics and the heat
exchange theory.

5.5. Heat - Mass Exchange in Open Systems

Classic thermodynamics is known to have always distinguished only

two kinds of system–environment energy exchange, viz. heat exchange and
work. The former has been considered as a random form of energy supply,
whereas the latter – as its ordered form. We owe the application of the
thermodynamic method to open systems with mass transfer therein to J.
Gibbs, who introduced the parameters Mk and Nk as additional independent
state variables and derived the joint equation of the first and second laws of
thermodynamics for open systems in the form of relationship (5.5.1) which
can be more conveniently expressed in terms of the mole numbers of the kth
substances:

dU = TdS - pdV + Σk μk dNk , (5.5.1)

104
where U, S, V – internal energy, entropy and volume of a system,
respectively; T, p – absolute temperature and pressure, respectively;
μk ≡ (∂U/∂Мk)S,V,Nm – component potential determined at constant S, V and
mole numbers of all other components Nm mass transfer process involving a
variation of the system composition due to the kth substances transfer across
the system borders and from the mass transfer process involving the
substance transfer without system composition variation.
The last term in this expression, by Gibbs’ assumption,
characterizes an independent process irreducible to heat exchange and cubic
strain and involving a composition and mass variation for a system as a
whole. He also assumed that the term TdS still (as in closed systems)
characterized the elementary heat exchange đQ. In fact, he writes, “…if a
system consists of parts supposedly not thermally interconnected, then any
entropy reduction in either of these parts should be deemed impossible since
such variations can not occur without heat transfer”. However, as
investigators were striking into the issue, they clarified the necessity to
exclude from the total entropy dS and volume dV variations for open
systems a part caused by the kth substance transfer. Despite the existing
disagreement in identification of this part (S.R. De Groot, R. Mazur, R.
Haase, 1967), the great majority of investigators have come to construing
the heat exchange and work in open systems as the part of energy exchange
caused by exclusively the temperature gradient and not associated with the
substance transfer across the system borders. In the same way heat is
understood and referred to in the collection of terms recommended by the
Academy of Science, “Process heat is the energy transferred at interaction
from a one body to another depending on exclusively temperature of the
bodies and not associated with substance transfer between them”
(Thermodynamics. Terms, 1973).
To confirm the validity of such an approach, let us proceed from the
joint equation of the first and second laws of thermodynamics derived for
one mole of the pure kth substance and, therefore, valid for both closed and
open systems. Designating the molar energy, entropy and pure substance
volume as ukо, sko and υko, respectively, and considering them as temperature
and pressure functions (skо = skо(р,Т), υkо = υkо(р,Т, rk)) the joint equation of
the first and second laws of thermodynamics has the form:

dukо = đqk - đwk = Tdsko – рdυko . (5.5.2)

Here đq = Tdsko, đw = рdυko – elementary neat and expansion work of

an open system under consideration.
Multiplying all terms of this equality by the mole number of the kth
substance and applying the Legendre transformation Nkdukо = dUkо – ukоdNk,
equation (6.2.2) may be modified as:

105
 dUkо = đQk - đWk + ukоdNk, (5.5.3)

where đQkо = Nkđqkо = NkTdsko and đWkо = Nkđwkо = Nkрdυko – respectively,

elementary heat exchange and expansion work of the same pure kth
substance taken to the amount of the Nk moles if the heat and work concepts
maintain their classic meaning. It is easy to see that once the terms TdS and
pdV in Gibbs’ relationship (5.5.1) are still understood as the heat exchange
đQ and expansion work đW, then in a particular case of a single-component
system (sk = sko; υko = υko ; uk = ukо и μkо = ukо – Tskо + pυkо) this relationship
does not go over into (5.5.3) as should be expected. Hence the terms of
(5.5.1) do not characterize independent processes as Gibbs assumed and the
principle of effect distinguishability demands, while the term ΣkμkdNk does
not determine the energy exchange caused by the kth substance transfer
across the system borders.
Thus there arises a problem in open systems how to distinguish the heat
transfer, work and mass transfer. The decision to the problem supposes
finding such coordinates of these processes which remain independent from
processes of other kind simultaneously running. It is quite evident that the
total entropy S and the total volume V of an open system are not any more
coordinates of heat transfer and expansion work therein since they vary at
mass transfer as well (variation of the total number of moles N at constant
composition of the system). The specific entropy s = Σkskrk and specific
volume of the mixture υ = Σk skrk are not these coordinates either since they
vary even at N = const due to the system composition variation (variation of
the mole fractions rk of the kth components). The partial molar entropies sk
and the volumes υk of the kth components can not either serve as coordinates
of heat transfer and work in open systems since they vary with the variation
of relationship between the mole numbers Nk of the kth substances, i.e. sk =
sk (р,Т, rk) и υk = υk (р,Т, rk) 1). Lastly, not quite adequate are the coordinates
skо(р,Т) and υkо(р,Т, rk) since they do not allow for the heat and bulk effects
arising at the irreversible mixing of the introduced component with the
substance of the system. Thus the concepts of “heat” and “work” loose their
intrinsic meaning in the systems where diffusion takes place (M. Tribus,
1970). In this connection the classic division of the external energy
exchange into heat transfer and work looses its heuristic value and should
give place to other process classification having nothing to do with the
energy exchange means. Such is, in particular, the “heat process” featuring
a temperature variation in a system at constancy of its volume, mass and
composition irrespective of what causes the process – external heat

1)
Remember that partial molar function of the kth substance means an increment in adequate
extensive value when introducing into the system one mole of this substance at constant
temperature, pressure and other potentials.

106
exchange or internal heat sources. A coordinate of this heating process may
be the entropy of a mixture irreversibly composed So = ΣkNksko (in the
absence of the heat mixing processes), where sko is considered a function of
exclusively intensive parameters (temperature and pressure as the simplest
case). Then:

đQ = ΣkNkTdsko = TdΘSo, (5.5.4)

where dΘSo – part of the entropy variation caused by exclusively heating and
not associated with the system volume and coordinate variation (including
Nk). This heat keeps the same meaning as in classic thermodynamics, i.e. is
defined by internal energy variation in the system at constancy of its volume
V and the coordinates Zi of all other kinds of work (đQ = [dU]V,Zi). It is
easily seen that the heat Q absorbed or released by the body is defined in
this case in terms of the “isocoordinate heat capacity” СΘ = (∂U/∂Т)V,Zi of
the system, which is a conceptual generalization of the isochoric heat
capacity CV for the case of a polyvariant system. Since the derivatives of
some parameters of the system (U in this case) with respect to its other
parameters (T in this case) are also its state parameters, the system
isocoordinate heat capacity СΘ becomes state parameter (like CV), while the
heat đQ – measure of internal heat energy variation:

đQ = dUт = СΘdТ. (5.5.5)

Similarly the expansion work for an open system đWр may be found as:

đWр = ΣkNk рdυko = рdψVo, (5.5.6)

where Vo = ΣkNk υko – volume of a reversibly composed mixture (in the

absence of the bulk mixing effects); dψVo – volume variation due to
exclusively bulk deformation of the system components and not associated
with the system intensive parameters’ variation ψi (including р, Т and сk).
The above proposed definition of heat and work in open systems as
notions keeping their classic meaning corresponds to the axiom of process
distinguishability and allows to distinguish these processes as independent
(irreducible to each other). In this case the heat-mass exchange process
should be first of all distinguished from the diffusion process of the kth
substances across the system borders, which involves the variation of
system composition with the system mass invariable. Accordingly, the heat-
mass exchange will be construed as an exchange of internal heat energy
between bodies if caused by the substance transfer across the system
borders and having nothing to do with the variation of the system
composition.

107
 It is easy to see that the term TdS in relation (5.5.1) includes the
component ΣkTskodNk caused by the transfer of the pure kth substances across
the system borders. This is the component that ought to be construed as the
analytical expression of the heat-mass exchange đQm:

đQm = ΣkTskodNk . (5.5.7)

Similarly, the term pdV in open systems, along with the usual expansion
work đWp = ΣkNkрdυko, also includes the so-called “input work” đWin =
Σkрυko dNk :

đWin = Σkрυko dNk. (5.5.8)

Substituting (5.5.7) and (5.5.8) into the Gibbs relation (5.5.1) gives:

dU = đQ + đQm – (đWp + đWm) + Σk μk dNk . (5.5.9)

From this it follows that the term ΣkμkdNk in the Gibbs relation
characterizes the heat-mass exchange with the diffusion of the kth non-
reacting substances across the system borders đUd. This process should be
called the diffusion energy-mass exchange 1), unlike the diffusion in its exact
meaning. In such a case the total energy-mass exchange ΣkukоdNk can be
represented as an algebraic sum of the heat-mass exchange đQm, the input
work đWin and the diffusion energy-mass exchange đUd.
Thus the heat exchange in open systems includes the conductive and
convective components of the heat flow caused by the heat conduction jqc
and the heat-mass exchange jqk, respectively. The latter is inseparable from
the density of the kth substance flow across the system borders jk = ρkvk
(mol/m2⋅s) and according to (5.5.4) equals jqк = Tskо jk. Then the basic law
of heat exchange in open systems becomes:

jq = –λ∇Т + ρk Tskо vk . (5.5.10)

This expression fundamentally differs from that traditionally used in the

heat-mass exchange theory because it contains not the enthalpy hk of the kth
substance input having nothing to do with the internal heat energy of the
substance, but the associated energy Tskо of the substance input.

1)
The term “diffusion” (from the Latin “diffusio” – spreading) relates, strictly speaking, to
the other process – the equalization of the component concentrations all over the bulk of the
system with its composition and mass invariable.

108
 Conclusion to Part 2

As mentioned in this section, all the basic integral and differential

equations of classic and nonequilibrium thermodynamics, as well as of the
heat-mass exchange theory, can be obtained as consequences from
thermokinetics without any recourse to whatever hypotheses, postulates and
statistical-mechanical considerations. Thereby not only the methodological
principles of the thermokinetic construction acquire a reliable
substantiation, but also the basic laws these disciplines state. Then the
assumptions which underlie some theory or other become clearer, as well as
the applicability limits of the basic concepts and principles pertaining to
these theories.
The deductive interpretation of the heat engineering theoretical
grounds eliminates the strange historical delimitation of various branches in
the heat doctrine and restores merits to thermodynamics as a consistently
phenomenological theory, which consequences have the status of oracles. It
is not less significant that such an interpretation gives the crux of the matter
in the most concise and compact form free of the burden of historical
accretions. Besides, the synthesis of the heat engineering theoretical
grounds on a unitary methodological basis exposes the reserves for their
further generalization. All this essentially improves the thermodynamic
investigation method in its heuristic value.

109
 Part 3

ELIMINATION OF NEGATIVE CONSEQUENCIES

OF THERMODYNAMICS EXTRAPOLATION

The permanent merit of the classic thermodynamic method, viz. an im-

mutable validity of its consequences, engenders the investigators’ natural
desire to apply thermodynamic methods to solution of new problems and,
furthermore, to extend these methods to new fields of knowledge not relat-
ing to thermodynamics. The main technique used for that is an “adaptation
to classic” with the help of various hypotheses, postulates and remote ana-
logs. As a rule, such a technique does not involve attempts to revise and
generalize the basic concepts and thermodynamic laws, since specialists are
oversensitive in apprehending these attempts. Such extrapolation results in
so intricate entwinement of truth and errors that it often is impossible to be
untwined due to multiple “metastases” – unforeseen remote consequences.
The objective of this Section is to consider from more general positions of
thermokinetics a number of consequences obtained beyond the applicability
of the classic thermodynamics and resulted in a number of paralogisms ap-
peared in it, viz. fallacious statements, though, seemingly, quite credible.

Capter 6

CORRECTION OF THERMODYNAMICS OF OPEN AND

POLYVARIANT SYSTEMS

Classic thermodynamics in accordance with the equilibrium self-non-

disturbance principle has always excluded from consideration the spontane-
ous state variations including the internal processes of variation in system
composition, structure, chemical properties, etc. There are, meantime,
chemical thermodynamics, thermodynamics of solutions, phase transitions,
etc. We owe such an extension of thermodynamic methods’ applicability to
J. Gibbs (1875), who replaced the internal irreversible processes by the re-
versible processes of system–environment exchange of the kth substances.
With this purpose in mind he represented non-equilibrium system as a set of

110
equilibrium open subsystems, viz. phases and components exchanging sub-
stance with the environment. Thus thermodynamics was supplemented with
one more type of the external heat exchange reducible to neither heat, nor
work and usually named a diffusion of the kth substances across the system
borders1).
Since the classic thermodynamic methods may be used for studying
only closed systems which interaction with the environment is restricted to
heat exchange and work, such a generalization of thermodynamics engen-
dered numerous difficulties far from being publicly recognized. They were
partly overcome by Gibbs with the mass Mk or the number of moles of the
kth substances Nk used as additional state variables. However, some of the
difficulties arisen have lasted out till the present and manifested themselves
in especially the indeterminacy of the concepts of heat and work in open
systems, unsuccessful attempts to solve the “Gibbs paradox”, loss of poten-
tial properties by free energy, inapplicability of the law of the excluded per-
petual motion of the second kind to open systems, etc. Some processes in
open systems appeared to have been so peculiar that “to explain and con-
firm their behavior from the classic concepts does not look like possible”
(M. Mamonov, 1970). Such are the diffusive processes with uncontrollable
substance flow across the system borders, chemical processes of isothermal
adiabatic fuel combustion in calorimeter, etc. There are, in particular, sys-
tems not allowing for even imaginary impenetrable membrane placed in
their borders with the process being undisturbed (M. Tribus, 1970). In-depth
investigation of such systems leads to a conclusion that “thermodynamics of
variable-mass bodies…has regarding classic thermodynamics the same de-
gree of the methodological self-sufficiency as thermodynamics of irreversi-
ble processes” (M. Mamontov, 1970).
Of more interest is, therefore, to investigate open systems from the posi-
tions of thermokinetics enabling consideration of internal processes in iso-
lated systems not recurring to the formalism above described.

6.1. Discrimination between Ordered and Unordered Works

Wishing to generalize the joint equation of the first and second laws of
classic thermodynamics in the form of Gibbs’ relationship (6.1.1) to sys-

1)
The term “diffusion” (from the Latin “diffusio” meaning spreading) refers, strictly
speaking, to the irreversible process of equalizing concentrations of some substances
when distributing all over the system volume. This process features composition invari-
ability of the system as a whole (unlike diffusion of the kth substances across the system
borders). Therefore hereafter we will distinguish the diffusion in a strict meaning of this
word from the selective.

111
tems doing other kinds of work beside expansion work the authors of many
guidelines in thermodynamics express this equation as:

dU = ΣiΨi dΘi , (i = 1,2,…,n) (6.1.1)

where U – internal energy of the system; Ψi, Θi – values named the “gener-
alized forces” and “generalized coordinates”.
Such an “extension” of the law of conservation of energy in the form of
(2.2.7) using the concepts of generalized force and generalized coordinate
introduced yet by Lagrange seems to many investigators so natural that it is
often applied without whatever substantiation. In particular, equation (6.1.1)
is often applied to the analysis of surface phenomena representing work as
(V. Sychev, 1986):

đWf = σf df , (6.1.2)

where Ψi is construed as the surface tension σf, while the coordinate Θi – as

the body surface f. Meantime, a multiphase system is spatially heterogene-
ous and internally non-equilibrium, which is evidenced by existing pressure
gradient on the phase boundary:

đWf = (р' – р")dV' , (6.1.3)

where р', р" – gas pressure in the cohabiting phases (e.g. steam and liq-
uid). Hence, this expression reflects processes in spatially heterogeneous
media and refers to the terms of the second sum (2.2.5).
The same often the generalized forces appear to include in their number
the intensities of electric E, magnetic H and other external fields, while the
generalized coordinates Θi – the vectors of polarization P and magnetization
M. However, dielectrics and magnetics are internally non-equilibrium sys-
tems – the fact which may be easily confirmed if observing the vector re-
laxation processes after their isolation from the external force fields. Their
spatial heterogeneity is manifested in the opposite sign the unlike charges or
poles displace when electric and magnetic dipoles are generating.
Expression (6.1.1) is often applied to the complex strain processes by
substituting the pressure P for the pressure tensor P, while the volume V –
for the strain tensor T. But for all that it is omitted that this tensor contains
the components characterizing the deviation of the system from equilibrium,
i.e. describing processes in non-equilibrium systems. This may be simply
instantiated by the uniform extension of a rod, which strain work is usually
expressed as:

đWℓ = Fℓ dℓ . (6.1.4)

112
 The value Ψi is herein construed as the extension force modulus Fℓ,
while Θi – as the rod length ℓ. Meantime, a stressed rod is as well non-
equilibrium system where the state of different parts varies opposite (the
displacements dℓ of different halves of the rod in its center-of-mass system
have opposite signs: dℓ' ≤0, dℓ" ≥ 0). Therefore the work to extend the rod
actually refers to the terms of the second sum in (2.2.5), which have vector
nature:

đWℓ = Fℓ'·dℓ'/2 + Fℓ"·dℓ"/2 = Fℓ·dℓ . (6.1.5)

It comes down even so far that equation (6.1.1) is applied to systems

within the gravity field, where Ψi is construed as the gravity field intensity
g, while Θi – as the distance between the body and the “field source” Rg.
However, the work to displace a body within the gravity field does not in-
fluence its internal state and may not enter the internal energy balance equa-
tion (R. Haase, 1967; I. Bazarov, 1991).
Therefore the transfer of the laws of thermodynamics to mechanical and
electromechanical phenomena involving the external energy of systems un-
der investigation should be done with great caution. The coordinates Ri of
centers of mass, inertia, gravity of a system and of whatever energy carrier
Θi refer to the external parameters of the system (I. Bazarov, 1991) and,
hence, characterize the external energy of the system. Therefore a formal in-
one-group integration of processes relating to principally different catego-
ries involving separately the internal and external energies of the system and
described by the two different sums of equation (2.2.5) can not be construes
otherwise than just an “adaptation to classic”. This has a lot of negative
consequences.
First of all, including “other kinds of work” most often means a tacit
change to consideration of spatially heterogeneous media. Such media ac-
cording to (2.2.6) have a double, while in a more general case of (2.2.5),
even a triple number of degrees of freedom. This means that each new form
of energy of the system, to be described, demands three additional parame-
ters of Θi, Ri and φi types to be generally introduced. However, only one in-
dependent coordinate Θi in (6.1.1) corresponds to any additional degree of
freedom. This generally leads to an “under-determinacy” of the system with
all methodological and mathematical errors ensuing.
The main feature of this kind of work is that it is associated with trans-
fer of the ith kind energy Ui by its energy carrier Θi from the environment
into the system (or visa versa). Denoting, according to (2.6.1), this part of
the coordinates Θi variation in the system and environment in terms of deΘi'
and deΘi", respectively, while the potentials Ψi – in terms of Ψi' и Ψi", re-
spectively, the elementary work đWidis may be herein expressed as:

đWi' = Ψi'deΘi' ; đWi" = Ψi"deΘi". (6.1.6)

113
 Since deΘi" = - deΘi', then at Ψi' = Ψi" works đWi' and đWi" are numeri-
cally equal. This means that in the reversible process of disordered work the
energy transfer across the system borders occurs without energy form varia-
tion. Therefore the work of such kind may be defined as a quantitative
measure of energy transfer process. In particular, bodies, at the reversible
heat transfer, exchange internal heat energy, while at the reversible expan-
sion –strain energy, at the electrization – electrostatic energy, etc.
Hereto should also be referred the “gas introduction work” Win =
∫ψυdМ, where ψυ = pυk – the so called “hydrodynamic” potential; the work
of charge Θе introduction into a zone with the φ potential (named here for
short the electrization work) Win = ∫φdΘе; the so-called “chemical work”
Wkd= ∫ζkdМk characterizing the energy exchange at diffusion of the kth sub-
stance into the system; the work due to varying kinetic energy of reciprocal
(diffusive) motion of mixture components Wkdif = Σk∫wkdPkα (where wk, Pkα
(α = 1,2,3) – velocity and momentum component of the kth substance in its
motion relative to the center of mass); the work of momentum turbulent
transfer inside the system Wωαт = ∫ωkαd(ζk ωkα), where ωkα, Мkωkα – compo-
nents of, respectively, angular velocity and angular momentum of the kth
system component. The heat exchange Q = ∫ТdS should also be referred to
this category if construed as a “micro-work” done on gas- or liquid-borne
particles at their spontaneous motion acceleration. All these kinds of work
vary only the system internal energy and are not associated with overcom-
ing the resultant Fi of whatever forces. Therefore in thermokinetics demand-
ing maximal generality we refer to these kinds of work as non-technical
work, while in a more general case of non-technical disciplines – disordered
work (failing a more adequate term) and denote them as Widis.
Work of quite other kind is done when the elementary forces Fik cause
the same-sign displacement drik of the objects these forces are applied to,
i.e. the work has an ordered directional character. In this case Σk eik ≠ 0, and
the forces Fik have a resultant Fi even though the system remains homoge-
neous. Such is the work done by technical devices (machines) intended for
the purposeful conversion between kinds of energy. Therefore such work is
often named technical (A.I. Andrushchenko, 1975). However, the work of
such kind is generally done by not only mechanical devices, but also bio-
logical, chemical, astrophysical, etc. systems. Therefore from the positions
of thermokinetics this should be preferably referred to as the useful or or-
dered work Wiе.
The useful work is also done when the system is heterogeneous, but the
forces Fik are applied to subsystems with different signs of the Θik property,
e.g. to positive and negative charges, north and south poles of magnetic di-
poles, electrons and holes in metals, zones with a reduced and increased (as
compared with average) density of the Θi parameter, etc. Such kind of work
is done at the dielectric and magnetic polarization, gas dissociation and

114
ionization, rod extension, redistribution of charges, substances, momentums
and entropy between the parts of an intrinsically homogeneous system. Di-
viding for such a case the sum ΣkΘikeik into two parts having the same sign
of the Θik and eik values the resultant force Fi other than zero is again re-
ceived. As shown in Chapter 2, coordinates of such kind processes (i.e. pa-
rameters necessarily varying when the process running) are the “displace-
ment vectors” ∆Ri or the “distribution moments” Zi = Θi∆Ri relating to, re-
spectively, the resultant forces Fi or their specific values xi = Fi/Θi = – ∇ψi
(thermodynamic forces). The analytical expression for the useful work Wiе
in a steady process (where (dRi·∇)ψi = dψi) may, according to (2.2.9), obtain
a more simple form as:

Wiе = ∫ Хi·dZi = – ∫ Θidψi = – Θi(ψi" – ψi'), (6.1.7)

where ψi', ψi" – generalized potential for the energy carrier Θi entering the
system and leaving it, respectively.
As can be seen, the useful work is associated with the variation of this
kind energy value Ui = Θiψi when the energy carrier crosses the system bor-
ders, i.e. with the conversion of energy from one form into other. Therefore
the useful work is a quantitative measure of energy conversion. This kind of
work features an ordered (vector) character.
Along with these categories of work one more category exists featuring
a conversion of energy from ordered form into disordered one. Such work is
usually named dissipative Widis. It features the initial form of energy having
its resultant Fi, however, in course of energy conversion the dissipative
forces have to be overcome, which do not have a resultant. Therefore in
course of the dissipative work ordered forms of energy are converted into
disordered ones (including heat). Thus the dissipative work may be ex-
pressed thru parameters of both initial (ordered) and final (disordered)
forms of energy. This directly ensues from the equations of balance for
these works (2.6.2) and (2.6.3), according to which:

đWdis = – Σi ∫jip·∇ψidV = Σi ∫ ψiσi dV, (6.1.8)

where jir – relaxation component of the ith energy carrier flow Θi; σi – den-
sity of the internal sources for this value.
In particular, in terms of system parameters the dissipative work may be
expressed caused by dying-out of relative motion of components at their dif-
fusive mixing (not influencing the momentum of the system in whole) Wkdif
= Σk∫wkdPk, the material destruction work Wddis = ∫ σfdf associated with the
material surface f increase (where σf – surface tension) at breaking of solid
and liquid materials, metal cutting, etc. Thus the energy dissipation gener-
ally consists not only in the conversion of ordered forms of energy into heat,
but also into any other disordered kind of energy. Mathematically this is ex-

115
pressed in the “scalarization” of the energy components in a system, i.e. in
the conversion of its inergy Е(Zi) into the anergy Ū(Θi).
The fact that the experimental values of the useful work done in chemi-
cal reactions conform to the analytical values of the same calculated by the
free energy decline promotes to a considerable degree the incomprehension
of the fundamental difference between the augend and the addend in expres-
sion (2.2.5). Such conformity is not accidental since dissipation heat is
equivalent to useful work. Therefore the incorrectness of such authentica-
tion is completely expressed in only the conclusion of the “complete con-
vertibility” of any kind of work (J. Szargut, R. Petela, 1968). In this respect
thermokinetics is remarkable in that it explicitly reveals the specific charac-
ter of the augend and the addend of (2.2.5). This specificity is that the ad-
dend’s terms have the tensor order different from that of the augend’s terms.
Physically this means that vector processes remain internally non-
equilibrium even being quasi-static (infinitely slow). This situation is abso-
lutely extraneous to classic thermodynamics where the adjectives quasi-
static and reversible have always been synonyms. Non-equilibrium proc-
esses naturally demand for their description specific non-equilibrium pa-
rameters to be introduced unavailable in equation (6.1.1) and extraneous to
classic thermodynamics in principle. Their absence forces to refer the vector
processes of polarization, magnetization to just a category of “other kinds of
work” done by a polyvariant system. This has the result the investigator
does not distinguish the ordered and disordered forms of external action on
a system. Here the incomprehension roots that the energy dissipation gener-
ally consists not only in the conversion of ordered forms of energy into heat,
but also into any other disordered kind of energy. Mathematically this is ex-
pressed in the “scalarization” of the energy components in a system, i.e. in
the conversion of its inergy Е(Zi) into the anergy Ū(Θi).
The fact that the experimental values of the useful work done in chemi-
cal reactions conform to the analytical values of the same calculated by the
free energy decline promotes to a considerable degree the incomprehension
of the fundamental difference between the augend and the addend in expres-
sion (2.2.5). Such conformity is not accidental since dissipation heat is
equivalent to useful work. Therefore the incorrectness of such authentica-
tion is completely expressed in only the conclusion of the “complete con-
vertibility” of any kind of work (J. Szargut, R. Petela, 1968). In this respect
thermokinetics is remarkable in that it explicitly reveals the specific charac-
ter of the augend and the addend of (2.2.5). This specificity is that the ad-
dend’s terms have the tensor order different from that of the augend’s terms.
Physically this means that vector processes remain internally non-
equilibrium even being quasi-static (infinitely slow). This situation is abso-
lutely extraneous to classic thermodynamics where the adjectives quasi-
static and reversible have always been synonyms. Non-equilibrium proc-
esses naturally demand for their description specific non-equilibrium pa-

116
rameters to be introduced unavailable in equation (6.1.1) and extraneous to
classic thermodynamics in principle. Their absence forces to refer the vector
processes of polarization, magnetization to just a category of “other kinds of
work” done by a polyvariant system.
One of such errors consists in a deep-rooted belief that homogeneous
systems are as well capable for useful work. In the homogeneous isobaric-
isothermal chemical reactions this work is numerically expressed in terms
of the Gibbs’ energy decline G = U + pV –TS. In reality the useful work in
chemical reactions can be done if only the chemical potential gradients are
available, which demands a spatial division of the reagents. This is what oc-
curs in Van’t Hoff’s box, galvanic or fuel cells. The homogeneous chemical
reactions inevitably become irreversible and involve the dissipation heat re-
lease, which is allowed for in TIP by an adequate increment in entropy
(4.3.3).

6.2. Substantiation of Second Law Applicability

to Non-Heat Engines

It is often confirmed as going without saying that the maximal effi-

ciency ηj of any non-heat engine consuming the energy from a well in the
form of work is always equal to unity:

η j = Wj /Wi = 1. (6.2.1)

Here Wj – work done by a non-heat engine; Wi – energy Ui΄ supplied to

a system by doing work on it. Such an assertion of the “exclusiveness” per-
taining to heat engines is based on the belief that heat and work are in prin-
ciple unequal because the latter can “directly apply to augment any form of
system energy, whereas heat – to replenish the internal energy only” (I.P.
Bazarov, 1991). The “substantiation” of this assertion usually appeals to
such kinds of work as bar extension, charge displacement in electric field,
load lifting, body acceleration, polarization, magnetization, etc, i.e. to useful
(ordered) kinds of work. However, the fact is missed in this case that the
useful external work is done on a spatially heterogeneous (expanded) sys-
tem as a whole and replenishes the ordered part of its internal energy
(inergy) measured by the amount of this work exactly. As shown above, the
useful external (ordered) work as itself is a quantitative measure of the en-
ergy conversion process. Therefore raising the question about the converti-
bility rate of the energy as applied to a heterogeneous system through doing
useful work on it is the tautology as itself. It is quite another matter if con-
cerns the convertibility rate of the energy as applied through doing unor-
dered kinds of work, to which, in accordance with the above, the “mi-
crowork” at heat exchange should be attributed. This is exactly the case re-

117
vealing that the “idea about impossibility of the perpetual motion of the 2nd
kind if shrunk to the statement about the exclusive properties of heat wells
is unjustified from the methodological standpoint” (A. Gukhman, 1947).
To make sure that the extrapolation of the energy conversion laws to
unordered kinds of work is inadmissible with the energy supplied to a sys-
tem in the form of ordered work, let us consider some working medium ca-
pable for expansion, heat exchange and exchange of the kth substances, elec-
trical charge, etc with the environment. If such a body (subsystem) is spa-
tially homogeneous, the replenishment of its unordered energy (anergy) as a
result of an unordered work Wi done on it is expressed, according to (2.2.5),
as:

dUi = Ψi dΘi . (i =1, 2, …, n) (6.2.2)

where Ψi – generalized force (absolute pressure р, chemical, electrical, etc

potentials of the system); Θi – generalized coordinates (system volume with
the opposite sign –V, number of moles Nk of the kth substance, bulk electri-
cal charge Θе, etc).
To avoid the expenditure of a working medium during the conversion of
the energy Ui received from a well, let us compel the medium to run a cyclic
process. The work of such a cycle is

Wц = ∫ ψ i dΘi (6.2.3)

It follows from this expression that the useful work Wj = Wc of the cyclic
machine under consideration is different from zero in only the case when the
potential ψi takes different values for different parts of the cycle. Otherwise,
factoring ψi outside the integral sign (6.7.2) and considering that the cir-
cuital integral of any state parameter ∫ dΘi = 0 gives Wj = 0. Thus in any
cyclic machine the working medium must contact at least two wells of the
convertible energy Ui with different potential values ψi΄ and ψi˝ (analogs of
the heat well and heat sink in the theory of heat engines). Then dividing the
circuital integral on the right-hand side of (6.7.2) into two parts where dUi΄=
ψi΄dΘi΄> 0 (energy is supplied to the working medium)) and dUi˝ = ψi˝dΘi˝<
0 (energy is withdrawn from it), respectively, and denoting the mean val-
ues of the potentials for these parts as ψi′ and ψi′′ (Fig. 3.1) with considera-
tion that dΘi˝ = – dΘi΄ gives :

đWj = ( ψi′ – ψi′′ )dΘi΄ или Wj = ( ψi′ – ψi′′ )Θi* . (6.2.4)

where Θi* = ∫dΘi˝ – amount of energy carrier from its well to its sink.
From this it follows that the absolute efficiency ηj of a cyclic engine
converting the arbitrary ith form of the homogeneous system energy may be

118
expressed in much the same way as the heat engine thermal efficiency
(3.2.3):

ηj = Wj /Ui΄ = 1 – ψi′′ / ψi′ . (6.2.5)

From this expression it can be seen that the efficiency for a cyclic con-
verter of unordered energy in any form is defined by exclusively the relation
between the mean integral values of the potential corresponding to supply
and withdrawal of the energy being converted and is always below unity
since the values ψi′′ = 0 or ψi′ = ∞ are physically unrealizable.
By way of example let us consider a cyclic-action electrostatic engine
converting external electrostatic energy of some charged body. The engine
is a plane capacitor with sliding plates. In such a device the electrostatic en-
ergy Uе΄ is supplied by entering the electric charge Θе with the elementary
work dUе΄ = φ΄dΘе done in this case to enter the charge dΘе into the area
with the potential φ΄, while the energy is withdrawn by doing the work dUе˝
= φ˝dΘе to bring this charge out. The useful work is done here through trans-
fer of the elementary charge dΘе from a charge well with the potential φ΄ to
a charge sink with the potential φ˝ and is equal to đWе= d(Uе΄– Uе˝) = (φ΄ –
φ˝)dΘе, which in relation to the well energy loss dUе΄ gives the part

η j = 1 – φ˝/φ΄, (6.2.6)

corresponding to relation (6.7.5). It is significant that the efficiency of such

an electrostatic energy converter η j for the quite real case of φ΄ = 108 V and
∆φ = 103 V will amount at the most to one thousandth of a percent irrespec-
tive of the electrostatic generator design. Thus we stand very far from the
possibility to use the “entire electric energy”.
Let us consider now an “expansion” machine doing work through the
pressure difference between the working medium p΄ and the environment
p˝. The work done in this case will be defined as đWре = (p΄ – p˝)dV, while
the well energy loss will amount to dU = p΄dV according to (6.7.2). Hence
the work done in this case makes up only a part of the well energy loss
and is determined as (6.7.5). This part is defined by the environmental
pressure which is known to be different from zero.
The same result can be obtained from consideration of non-cyclic
open-type engines where the working medium is the substance of the en-
ergy well itself, while the sink is the environment wherein the working
medium enters after the work done. In such machines the energy dU΄ is
supplied by entering the kth substance in the amount of dNk moles with the
molar enthalpy hk΄, while the energy dU˝ is withdrawn by bringing this sub-
stance out with the enthalpy hk˝. Then dU΄ = hk΄dNk , dU˝= hk˝dNk and the
work done đWk = d(U΄– U˝) = (hk΄– hk˝)dNk. Thus the efficiency of the
open-type engines (e.g. magneto-hydrodynamic energy generators operating

119
in open circuit) is characterized through the so-called “enthalpy” efficiency
(Favorsky, 1978):

ηi = 1 – h˝/h΄ , (6.2.7)

where h΄, h˝ – enthalpy of the working medium at the generator inlet and
outlet, respectively. Since the enthalpy of dissociated and ionized gas is by
no means proportional to its temperature, this efficiency will essentially dif-
fer from the thermal efficiency at the same temperatures of working me-
dium.
The said directly concerns the engines using the external kinetic and po-
tential energy. The “adaptation to the classics” for such systems as force
fields or velocity fields usually means that the intensities of these fields (ex-
pressed in terms of the potential gradients ∇ψi) are taken for the “general-
ized forces” Ψi, while the value of the energy carrier Θi*=ΔΘi˝ transferred
from some field point to another – for the “generalized coordinates” Θi
themselves. Meanwhile, equilibrium thermodynamics if correctly applied
to them requires due consideration of the fact that these fields are formed
by heterogeneously distributed material bodies with the masses М1, М2,
…, Мn or the charges Θ1, Θ2, …, Θn, which change their state in opposite
way: some of them receive a charge or a mass, whereas others lose them,
on the contrary. Thus, to set the problem in the same form as for heat en-
gines, some of these subsystems should be considered as energy wells,
whereas others – as energy sinks. By way of example let us consider an
engine that uses the gravitational energy Ug. A body with a mass of М1 in-
troduced into this field has the gravitational potential Ψg = Ug/М1 = –
GМ2/R = Ug/М1, where М1, М2 – masses of the field-forming bodies, G –
gravitation constant, R – center-to-center distance of the gravitating
masses.
Let now one of the bodies with a mass of М1 give the part dМ1 of its
mass to the working medium passing from a field point with the gravita-
tional potential ψg΄ to a point with the potential ψg˝ doing at that the work
đWg = (ψg΄– ψg˝)dМ1. This value is less than the energy dUg΄ = ψg΄dМ1 the
working medium receives in this case so that the relation between them ex-
actly corresponds to expression (6.7.5). Note that the absolute efficiency of
such a “gravitational” engine is below unity since the value ψg˝ = 0 corre-
sponds to the infinite distance in-between the interacting bodies and is,
hence, unrealizable. Note furthermore that the same conclusion can be ob-
tained as a short cut through relation (6.7.7).
Thus the laws of energy conversion in any unordered form appear to be
unitary in classic thermodynamics if, in accordance with its methodology,
the object of investigation is construed as a homogeneous system all parts of
which similarly change their state in processes under investigation. In this
case a necessity arises in another body available not being in equilibrium

120
with the energy well and capable to play the part of the energy sink, i.e. a
necessity in the availability of a nonequilibrium system. Such a nonequilib-
rium system classic thermodynamics considers may be instantiated as the
so-called “expanded” systems comprising the environment as energy sink.
In this case even at φ˝ = 0 there is direct evidence of the second criterion of
the “compensation” for energy conversion process R. Clausius mentioned –
the transfer of some part of energy carrier (which quantitative measure is
the parameter Θi) from energy well to energy sink with changing the state of
not only the well, but also other subsystems involved in the process (in op-
posite way though). Note the question here is not the inevitable losses, but
rather the principal unattainability of 100% conversion of the energy ob-
tained through the ψi dΘi impact, i.e. by the energy transfer (unordered
work done). The other challenge issued as the conversion of the energy sup-
plied to a system through such an action as the useful work Хi·d Zi (Е·d P,
H·d B, etc) is a departure from the classic problem definition.
Meanwhile, this is exactly the way investigators choose when construe
the parameter Ψi as the voltage Δφ = φ΄– φ˝ across the capacitor plates or
the fuel cell electrodes and the coordinate Θе – as the charge ΔΘе transferred
in-between them (which is, by the way, the function of process rather than
state). In this case, naturally, the “deduction” is inevitable of complete con-
vertibility of electric or other energy. This does not come as a surprise since
in this case the electric energy dUе΄ = ΨidΘi supplied to the converter ap-
pears to be identical with the useful work đWе = (φ΄– φ˝)dΘе it does. How-
ever, it is easy to show that such a conclusion means the violation of the
equilibrium thermodynamics applicability conditions. Indeed, any extensive
parameter in equilibrium thermodynamics characterizes the state of an en-
tire system and varies uniformly in all its parts. This statement is evidently
violated in the abovementioned nonequilibrium systems, different parts of
which (subsystems) change their state oppositely in the course of energy
conversion. Such subsystems may include not only the heat wells and heat
sinks in the classic theory of heat engines (which oppositely change their
entropy), but also the opposite charges moving in the opposite directions or
opposite poles of polarized and magnetized bodies, the electrons and holes
in semiconductors, the opposite plates of capacitors, the positive and nega-
tive ions in electrolytes and plasma, the spin systems and atomic lattice in
crystals. It is easy to make sure that all these systems are nonequilibrium
when observing the arbitrary variation of their state after having been iso-
lated from the environment. The consideration of such systems, strictly
speaking, outsteps the applicability of classic thermodynamics since their
behavior conflicts with the fundamental “equilibrium self-non-disturbance
principle” (general law) of classic thermodynamics. According to this prin-
ciple the state of a thermodynamic system can be changed by only an im-
pact from outside. Since this condition is evidently violated in nonequilib-
rium systems, the latter, from the position of classic thermodynamics,

121
should be divided into spatially homogeneous areas, phases or components,
to which the equations of equilibrium thermodynamics are solely applica-
ble. This is the violation of this condition that underlies the erroneous divi-
sion of various energy forms into “completely convertible” and “incom-
pletely convertible” (A. Dolinsky, V. Brodyansky, 1991). Any form of en-
ergy is convertible inasmuch as it is internally nonequilibrium and the rela-
tion between its ordered and unordered parts is a measure of this nonequi-
librium state.
It is possible to prove that the principle of excluded perpetual motion of
the 2nd kind has the universal character for also continuums where wells and
sinks of whatever kind of energy can be hardly distinguished. To do so, it is
just necessary to use the body of mathematics for thermokinetics operating
the notions of forces and flows. Indeed, as follows from relations (6.7.9), to
do useful work, it is necessary here also to arrange a transfer of the energy
carrier Θi from a subsystem with the potential ψi΄ to another one with the
potential ψi˝. This means that some areas of the continuum are wells,
whereas others – sinks of the ith energy carrier. This is what means the
“compensation”, which necessity R. Clausius has stated. In this case the
work Wiе according to (6.7.4) makes up relative to the supplied energy Ui΄ =
ψi΄Θi the same part (6.7.5) as for heat engines. This fact evidences the uni-
tary of the conversion processes for energy in any forms.
Some investigators recognizing the universality of expression (6.7.5) as-
sociate, nonetheless, the possibility to reach the 100% efficiency with the
arbitrariness to choose the zero reference potential of energy carrier sink
(ψi˝ = 0) for non-heat forms of energy. Such arbitrariness really takes place
in mechanics and electrodynamics operating external energy. For them the
zero reference point is a matter of agreement. However, it is absolutely in-
tolerable in thermodynamics, where the question as itself arose regarding
the degree of heat convertibility into work. As a matter of fact, if it were
not for the third law of thermodynamics (principle of unattainable absolute
zero of temperature), then we, having the heat sink temperature taken for
zero, would immediately come to the conclusion that the thermal effi-
ciency could also be equal to unity. However, from the positions of ther-
modynamics the state of any bodies involved in processes under investiga-
tion or varying their state in whatever manner can not be taken as zero ref-
erence point for any of the system parameters. Otherwise, the system in-
ternal energy would vary also in the absence of energy exchange with
bodies under consideration (only due to state variation of these reference
bodies). That would violate the first law of thermodynamics which de-
clares the internal energy existing as a state function and not depending on
the position or movement of the system relative to those bodies. Hence the
zero of whatever potential corresponds in thermodynamics to the “degen-
eracy” of movement (interaction) of a particular kind and to the loss of a

122
particular degree of freedom by the system, which makes it inapplicable as
energy sink.
Thus the conclusion of 100% efficiency for any non-heat engines is a
consequence of a number of methodological errors, viz. departure from
the classic problem definition regarding heat engine, confusion of energy
and anergy as notions, identification of technical and nontechnical (or-
dered and unordered) kinds of work, arbitrary choice of zero reference po-
tentials.
It is worth noticing that along with the machines converting some ith
form of energy also such ones exist which use simultaneously several
kinds of energy or several wells of the same form of energy. In all such
cases the notion of efficiency also needs clarification since the work of
such a machine is defined by the difference of several potentials Ψi. In
such cases one can speak of only the conversion degree of the energy sup-
plied from all its wells ΣjUi΄, which corresponds to the efficiency expressed
as:

ηi = Wi/ΣjUi΄ . (6.2.8)

The universal character of the principle of excluded perpetual motion of

the 2nd kind becomes much more understandable with the ban it declares
removed and superseded by the condition necessary and sufficient to create
heat engines. This condition finds its reflection in a famous dictum by S.
Carnot, “everywhere with a temperature difference available the thermal
motive force is possible to appear”. In a more complex systems being in
partial equilibrium the absence of some process or other may follow from
the mutual compensation of thermodynamic forces Xj (i, j = 1, 2, …, n).
This relates also to systems featuring heterogeneous temperature fields and
being in gravitational, centrifugal, etc force fields. For such systems the de-
viation from internal equilibrium (the generalized potentials ψi unequal in
some parts of the system) is already insufficient criterion for the absence of
external equilibrium (ceased flow of energy of the ith kind between the sys-
tem and the external field and its conversion into other forms of energy).
That demanded to clarify the Carnot principle and to declare that external
equilibrium with the force field to be necessarily absent. Such a generaliza-
tion, by analogy with the Carnot principle, may be expressed as follows,
“everywhere with equilibrium being absent the useful conversion of energy
is possible”. This statement is valid also for systems in external force fields.

6.3. Correction of Material Equilibrium Conditions

Based on relationships (5.5.1) Gibbs was he first who defined the condi-
tions of equilibrium distribution of the kth substances in heterogeneous sys-

123
tems. Those conditions consisted in equality of temperatures T, pressures p
and chemical potentials μk of the kth substances in all parts of such systems,
which corresponded to heat, mechanical and material (“chemical” after
Gibbs) equilibrium in them. Since the equilibrium in thermodynamics is
construed as a state featuring a cessation of some macro-processes, it would
be logic to assume that the above kinds of equilibrium mean the cessation
of, respectively, heat transfer, cubic strain and mass transfer between the
parts of a system under consideration. However, when that question was
considered from the theory of irreversible processes (TIP), it became clear
that the equality of chemical potentials μk was not yet a sufficient condition
for the cessation of the kth substance redistribution (I. Prigogine, 1947; S. De
Groot, 1956; R. Haase, 1967; etc.). It turned up that with the chemical po-
tential gradients for all system components having become zero, i.e. with
the Gibbs’ material equilibrium conditions met, the kth substances redistri-
bution did not cease, the transfer of those substances could occur also due to
temperature gradient (thermo-diffusion, thermo-osmosis), pressure gradient
(pressure diffusion, reverse osmosis), electric potentials (electro-osmosis,
electrophoresis, electroplating), etc. However, the so-called “stationary
states” were allegedly possible, when the flow of some kth substance disap-
peared not because its motive force, viz. its chemical potential gradient, be-
came zero, but due to other motive forces maintained in the system, which
mitigated the process. On the other hand, according to TIP, when the redis-
tribution of the kth substances is running, the thermal equilibrium can not be
either reached since the heat-mass transfer (transfer of internal energy of kth
substances’ spontaneous motion) is possibly maintained in this case. A more
than paradoxical situation has arisen, when the cessation of a process is no
more the criterion of equilibrium reached, which violates the thermody-
namic grounds themselves. Nevertheless, none of the investigators doubted
the conformity of the J. Gibbs-found equilibrium conditions to the real cir-
cumstances. Meantime, that fact could have been noticed a long time ago
due to the impossibility to define the chemical potential μk = (∂U/∂Nk)S,V,Nm
in such a way as expected from relationship (6.2.1), i.e. as a partial deriva-
tive at constant entropy S and volume V of the system. Actually, neither the
volume V = ΣkNkυk, nor the entropy S = ΣkNksk particularly, being extensive
values, do not remain, generally speaking, invariable with introducing the
kth substance changing not only the composition (Nk/N), but also the mass
(mole number) of the system (N). Unfortunately, instead of correction of the
Gibbs’ relationship the investigators preferred the stationary state theory
having developed, where at the process cessation the phenomenological co-
efficients appeared as a function of the process rate and construed as neither
thermodynamic variables, nor pure kinetic values (Chapter 4).
Meantime, the true form of the component potential is easy to be found.
For this let us take into account that introducing the kth substance should not
vary the parameters sko and υko. Therefore considering the internal energy of

124
the kth component Uk as a function of the independent variables sko, υko, Nk
and noting that at Nk = const the derivatives (∂Uk/∂Sko) = Т and (∂Uk/∂Vko) =
- р (i.e. mean the same as in closed systems) gives:

dU = ΣkNkTdsko - ΣkNkрdυko + ΣkςkdNk, (6.3.1)

where ςk= (∂Uk/∂Nk) – potential of the kth component corresponding to the

osmotic diffusion of the kth substance across the system borders and there-
fore named the osmotic potential (V. Etkin, 1999) 1). The first two sums of
this expression characterize the independent processes of heat exchange and
expansion work in a constant-composition system. They do not include the
heat and bulk effects associated with introducing the kth substance and the
mixing processes involved. These effects are inseparable from the substance
introduction and, therefore, refer to the last term of (6.3.1).
To find a relation between the osmotic and chemical potentials, let us
rearrange Gibbs’ relationship (6.2.1) subtracting the terms ΣkTskodNk and
ΣkрυkodNk from the first two of its components and simultaneously adding
them to the last component (so that not to disturb the energy balance):

dU = ΣkNkTdsko -– ΣkNkрdυko + Σk (μk +Tsko – pυko)dNk , (6.3.2)

wherefrom it follows that

ςk = μk + Tsko -– рυko = uk – T(sk –sko) + р(υk – υko). (6.3.3)

Representing the chemical potential μk by the known way in terms of

the partial molar internal energy uk, the partial molar entropy sk and the par-
tial molar volume of the kth component υk (μk = uk – Tsk + pυk) one can find
that for not interacting substances (sk = sko and υk = υko) the osmotic poten-
tial ςk is equal to the molar energy uk transferred by unit pure kth substance
across the system borders. Thus the thermokinetic analog of Gibbs’ rela-
tionship (6.3.1), unlike (6.2.1), changes to relationship (6.2.3) for, in par-
ticular case, a single-component open system, which proves the correctness
of the revision applied to the component potential expression.
Now let us consider the diffusion process in its “pure” form as a mixing
of components in a system with a constant mass M and volume V, which
may be both continuous and separated into two parts with a semipermeable
membrane or a gate. This mixing process as well involves heat effects due
to interaction between components. At V = const these are variables sko, V
and Nk that become independent, therefore the component potential is de-
fined as:

ζk = (∂U/∂Nk)Sko,V,Nk. (6.3.4)

125
 This potential is advisably to be named diffusive. Its relation to the
chemical potential may be revealed coordinating Gibbs’ relationship (6.1.1)
with new uniqueness conditions. Assuming in (6.1.1) dV = 0 and rearrang-
ing the term ΣkTsk dNk gives:

dU = ΣkNkTdsko + Σk ζk dNk . (6.3.5)

where
ζk = hk + T(sko – sk). (6.3.6)

In even more particular case, when the components do not interact and
the partial molar enthalpy and entropy of the kth component are equal to
those for pure substances (hk = hkо, sk = sko), the enthalpy hkо becomes the
component potential. This is characteristic, e.g., for filtration process. For
such processes the fundamental equation of thermokinetics in open systems
becomes:

dU = ΣkNkTdsko + ΣkhkdNk . (6.3.7)

As can be seen, the variety of uniqueness conditions for processes

under investigation results in a variety of component potential, which is
quite naturally. From here the variety of material equilibrium conditions
also ensues. In particular, at p,T = const (or equivalent conditions sko, υko =
const) using the method described in Chapter 4 one can find that the equilib-
rium distribution of the kth substance among the homogeneous parts of a
heterogeneous system (which parameters are single- and double-primed, re-
spectively) occurs with the equality of osmotic potentials in them:

ςk' = ςk". (6.3.8)

Similarly, at T,V = const the equilibrium condition means the equality

of the diffusive potentials:

ζk' = ζk". (6.3.9)

At last, for a filtration process equilibrium occurs with the equality of

enthalpies of the kth substance in the system and filtrate:

hk' = hk" . (6.3.10)

The definition of a particular form for the component potential in ac-

cordance with uniqueness conditions has far reaching consequences. Firstly,
this allows finding that only motive force providing the transfer of some
substance, which ceases the process if becomes zero. With such a (resultant)

126
force found phenomenological laws (4.4.1) may be starkly simplified by
negating the summation with respect to all “thermodynamic” forces Xi
available in the system, i.e. by reducing these laws to the “diagonal” form
not containing the cross phenomenological coefficients Lij. Afterward this
will allow reducing the number of the phenomenological coefficients in
these equations, expressing the so-called superposition effects in terms of
thermodynamic parameters and extending the TIP methods to non-linear
systems (Chapter 12).

6.4. Definition of Potential State Functions for Open Systems

The thermodynamic potentials, such as Helmholtz’s Г = U – TS or

Gibbs’ G = U – TS + pV free energy, are known to be inapplicable to open
systems. The reason of this becomes more explicit if approaching the ques-
tion from the positions of thermokinetics. According to (2.5.6) the ynergy Y
of any, including open, system, i.e. its reversible (capable for work) part,
may be found both from the “reverse balance”, i.e. by the difference be-
tween the system energy U and its equilibrium part ΣiΨi Θi, and directly in
terms of the non-equilibrium parameters Хi and Zi. At that according to
(2.5.6)

đW = –d(U – ΣiΨiΘi) + ΣiΘidΨi , (6.4.1)

where the expression in brackets defines the system ynergy E found from
the “reverse balance”. In the particular case of an open multi-component
thermo-mechanical system described by Gibbs’ relationship (6.1.1) expres-
sion (2.5.6) becomes:

đW = –d(U –TS + pV –ΣkμkNk) = SdT –Vdp –ΣkNkdμk. (6.4.2)

It may be easily confirmed applying to the terms of Gibbs’ relationship

(6.2.1) the Legendre transformation TdS = dTS – SdT; pdV = dpV – Vdp;
μkdNk = dμkNk –Nkdμk followed by their rearrangement. Since there are no
the Хi and Zi parameters in classic thermodynamics, the open system free
energy may be found from exclusively the reverse balance. In this case with
the traditional definition of the chemical potential as μk ≡ uk – Tsk + pυk the
left-hand side of (6.4.1) becomes zero since ΣkμkNk = U –TS + pV ≡ G. This
is what the known Gibbs-Duhem relationship follows from:

SdT – Vdp –ΣkNkdμk = 0. (6.4.3)

127
 Based on this, a conclusion is usually drawn that there is no the poten-
tial in thermodynamics of open systems, which reduction would define the
work done by a system (I. Bazarov, 1991; G. Gladyshev, 1988).
The paralogism of the situation arisen is that open multi-component sys-
tems are actually capable for useful work. Such are, in particular, the fuel
cells using chemical affinity of fuel components. However, this needs the
components, e.g. oxygen and hydrogen, to be spatially separated and sepa-
rately supplied to the fuel cell electrodes. In other words, a spatial heteroge-
neity of the system is required described in thermokinetics by the specific
parameters Хi and Zi. In this case applying the Legendre transformation
ΨidΘi = dΨiΘi – ΘidΨi to (6.3.1) gives:

d(U – ΣiΨiΘi) = dE = ΣiΘidΨi – ΣiХi·dZi. (6.4.4)

Since in this equation the expression ΣiΘidΨi representing Gibbs-

Duhem relationship (6.4.3) in the brief form becomes zero, the sum of all
kinds of useful works ΣiХi·dZi in an open system is defined by its ynergy
reduction:

đW = ΣiХi·dZi = – dE . (6.4.5)

Thus the reason the Gibbs’ free energy looses its potential properties is
the said “under-determinacy” of open non-equilibrium systems, i.e. the at-
tempt to investigate them with the same set of variables as in equilibrium
thermodynamics. Here the imperishable advantage is rooted, which can be
obtained from introducing the inergy as the most common measure of order
and integrity of both open and closed, simple and complex, homogeneous
and heterogeneous systems.

6.5. Solution to the Problem of Thermodynamic Inequalities

The real processes running with a finite rate are known to represent a
sequence of non-equilibrium states. The relaxation processes inevitable in
this case force spontaneous variations of the entropy S, volume V and some
other state parameters. This fact is reflected in the equations of their balance
like (2.4.1) and (3.3.4). As a result, the classic equations of the process ele-
mentary heat đQ and expansion work đW go over into the inequalities:

đQ ≠ TdS ; đWр ≠ рdV, (6.5.1)

and the so-called problem of thermodynamic inequalities arises not solved

yet.

128
 The intensity of processes increases, inequalities (6.5.1) aggravate, and
the calculation of the system external energy exchange, based on them, be-
comes increasingly non-rigorous. For all of that classic thermodynamics it-
self is incapable to estimate the error associated with disregard of the said
inequalities since their exact analytical expressions remains unknown.
This problem is accentuated for open and polyvariant systems doing,
besides the expansion work, other kinds of work Wj (e.g. the work of intro-
ducing the kth substances into the system, work against surface tension
forces, polarization and magnetization). In this case not only the masses Mk
of the kth substances as system components can arbitrarily vary (due to in-
ternal chemical transformations), but also the coordinates Zi of all kinds of
useful work (due to system relaxation). This fact is reflected in balance
equations (2.4.1) and (2.4.2) and results in the inequalities for also the en-
ergy-mass exchange đUk and all kinds of the useful work đWiе:

đUk ≠ μkdNk; đWiе ≠ Хi·dZi . (6.5.2)

The reason for which the thermodynamic inequalities appear becomes

evident enough if considering the issue from the positions of thermokinet-
ics. With this purpose let us consider the equation of balance of various co-
ordinate “wells” in the form of (2.4.7) or (2.4.8). This balance is provided
due to compensation of the “wells” duΘi of the certain coordinates (the en-
tropy, the kth substance masses, the volume, etc.) by the “sinks” duRj of oth-
ers (i.e. by equalization of distributing the parameters Θi all over the sys-
tem). However, the basic identity of thermokinetics in the form of (2.4.7)
immediately goes over into an inequality along with balance equation
(2.4.8) if excluding some of its actual terms allowing for the sinks duRj.
This is what happens to the equations of classic thermodynamics because of
the absence therein the non-equilibrium state parameters Хi and Zi. Thus the
inequalities are caused by the said under-determinacy of a system, i.e. by an
attempt to take irreversibility into account not allowing explicitly (with the
help of the non-equilibrium parameters and their sinks) for its reason – the
spatial heterogeneity. In fact, there are no reasons in spatially homogeneous
and chemically neutral media, which would cause a spontaneous variation
of whatever thermostatic variables Θi, i.e. their wells or sinks appearing.
Should these wells or sinks nevertheless appear, this means some uncounted
reasons exist.
Equation (2.3.7) suggests the way out when implying the possibility to
represent the external energy exchange directly in terms of the energy car-
rier flow jie across the system borders. Indeed, the variation deΘi of any co-
ordinate Θi due to its transfer across the borders of a stationary system can
be represented in terms of its local flow ji density:

deΘi /dt = – ∫∇⋅jie dV . (6.5.3)

129
 From this it follows that the terms of the first summand (2.3.7) give the
exact analytical expression for the work of unordered character (process ca-
pacity Niun = đWiun/dt), where ψi is a local value of the generalized potential
on the system border:

Niun = ∫ψi∇⋅jie dV = ∫ψi jie df. (6.5.4)

The flows jie can be measured with heat meters, ammeters, flow meters,
and other devices, while the potentials ψi – with temperature sensors, pres-
sure sensors, etc. This enables in principle to solve the problem consisting
in direct definition of the system energy exchange in a variety of conditions
and to retain, at the same time, the phenomenological character of the clas-
sic theory for nonequilibrium thermodynamics.
In spatially homogeneous (internally equilibrium) systems the potentials
ψi are the same for all points of the system and can be factored outside the
integral sign. This corresponds to the infinitely slow (quasi-static) state
variations, i.e. to the so-called reversible processes. Then ∫∇⋅jiedV = dΘi/dt
and relation (6.5.4) is simplified going over into (2.3.2). Table 6.1 illus-
trates the exact analytical expressions for heat exchange, mass exchange,
expansion work and other unordered works in reversible and irreversible
processes as based on (6.5.4). All kinds of unordered works (including
heat exchange and mass exchange) are assumed in the table as positive if
augmenting the system energy.

T ab le 6.1
Exact Analytical Expressions for Unordered Works
In reversible In irreversible
Process Note
proc., đWiun, J proc., Niun, W
jme – substance flow f
Substance Enter đWin = рυdM – ∫рυjme·df
across surface
jυe– surface f dis-
Cubic Strain đWр = – рdV ∫рjυe ·df
placement velocity
Chemical Reaction đWх = Ardξ r – Σk∫ μk jke·df μk– chemical potential
Diffusion of the kth đUk = ζ kDdNk jke – kth substance flow
– ∫ζ k jke·df
Substance (р,Т = const)

Heat Exchange đQ = TdS – ∫Т jse·df jse – entropy flow

Osmosis of the kth đUk = ς kosdNk jke – solvent flow
– ∫ς k jke·d f
Substance (V,Т = const)
Electrization đWе = φdΘе – ∫φjеe·d f jеe – electric current
Star Matter Accre- ψg – gravitational po-
đWаc = ψgdМ – ∫ ψgjme·d f
tion tential

130
 Most of the expressions for reversible energy exchange shown in the
table are known from classic thermodynamics with the exception of
probably the energy-mass exchange at diffusion of the kth substance
wherein the chemical potential μk gives place to the diffusive ζk and osmotic
ςk potentials in line with the corrected notions of heat in open systems (see
Chapter 5). Of interest are also the expressions for accretion work – an incre-
ment in the mass of a star with a substance dropped onto its surface from out-
side, which is the gravitational analog of the gas-enter work.
The table is primarily attractive with the generality of the analytical ex-
pressions for all kinds of unordered works. This emphasizes the fact that, as
a result of this category of processes, a system remains internally equilib-
rium, i.e. acquires inconvertible energy (anergy) and, therefore, remains in-
capable for work. This is the reason why, in order to convert the energy ob-
tained from such energy exchange processes, a system is required not being
in equilibrium with this energy and, therefore, capable to play the part simi-
lar to the heat sink in classic thermodynamics.
The basic equation of thermokinetics (2.3.7) gives also exact analytical
expressions for the useful (ordered) external works Wiе. In accordance with
the flow balance equations (2.4.3) this equation distinguishes that part of the
redistribution processes which is not associated with the relaxation proc-
esses (the relaxation flows jir). However, it is rather problematic to distin-
guish the component jiе in the continuum bulk elements. Therefore, let us
consider the possibility to simplify the terms of the second sum in this ex-
pression. To do so, it is necessary to take into consideration that the redistri-
bution processes correspond to the development ∇⋅(ψijie) at ∇⋅jie = 0. For the
system as a whole this condition takes the form of ∫∇⋅jie dV = 0. Changing in
this expression from the volume integral to the integral jie⋅n df = 0 taken
through the closed surface f of the system gives that at jie ≠ 0 this condition
implies existing the parts f ΄ and f ˝on this surface with the opposite sign of
the product jie⋅n, i.e. existing the flows ji΄ and ji˝ entering the system and
leaving it. Then:

∫ji⋅n df = ∫ji΄⋅n df ΄ + ∫ji˝⋅n df˝ = Ji˝ – Ji΄ = 0, (6.5.5)

where Ji΄ = ∫ji΄⋅n df ΄ ≤ 0 and Ji˝ = ∫ji˝⋅n df ˝ ≥ 0 – energy carrier flows at the
system inlet and outlet, respectively.
Thus, to distinguish in the balance equation (2.6.3) the flow external
component Jiе caused by the useful work We done from the relaxation com-
ponent Jir, it is possible to proceed from the conservation of the energy car-
rier flows Ji΄ and J˝ at the system inlet and outlet, respectively. Substituting
into (2.4.7) the bulk element dV across the flow jie as the product of its cross-
section normal-to-flow vector element df and the normal element dn (dV =
df·dn) with (dn·∇)ψi = dψi for steady flow gives the following instead of the
second sum in (2.4.7):
131
 Ni = ∫(∫ jie·df)dψi = – ∫ Jiеdψi = Jiе(ψi΄ – ψi˝), (6.5.6)

where Jiе = – ∫ jie·df = Ji˝ = – | Ji΄| – energy carrier “transit” flow crossing the
system without changing its value; ψi΄, ψi˝ – energy carrier potentials at the
system inlet and outlet, respectively.
Thus the useful component is associated with the energy carrier “tran-
sit” flow Jiе which crosses a system without changing its value. This state-
ment may be instantiated by the electric machine where input current is
known to be equal to output one. This is the same evident fact for also the
cyclic heat engine since in any cycle the entropy rise ∆S1–2 at heat supply
from heat well is numerically equal to its loss ∆S2–1 at heat rejection to heat
sink. Therefore, for steady state the entropy flows Jsе = ∆S/∆t at the machine
inlet and outlet are the same, too. This is the circumstance that served for S.
Carnot the basis to compare heat engine with mill wheel (hydraulic ma-
chine). The analytical equations derived from the above expression are for
convenience tabulated in Table 6.2.

Table 6.2
Exact Analytical Expressions for Ordered Works
In reversible In irreversible
Process Note
е
processes đW , J processes Niе, W

Heat Transformation –SdT Nq = – ∫ Jsе dТ Jsе – entropy flow

Gas Work in Flow – Vdр Nр = – ∫ Jυе dр Jυе – flow rate
Liquid Injection – Мdh Nm = – ∫ Jmе dh Jmе – liquid flow rate
МА r Nx = – ∫ Jrе Jrе – reagents flow rate,
Chemical Reaction
d(Аr ξ r) А r –affinity
Binary Mixture Sepa-
–Nkdψk Nд = – ∫ Jkе dψk ψk – k th potential
ration
Electric Charge Dis- – Θеdφ Jее – circuit current,
Nе = – ∫ Jее dφ
placement φ – electric potential
Jес – displacement cur-
Dielectric Polarization – Е·dР Nп = – ∫ Jес dφ
rent
Nk = – ∫ Jwе dv
Body Acceleration – P·dv Jwе – momentum flux
v – скорость
Displacement in
– Мdψg Ng = – ∫ Jmе dψg ψg – gravit. potential
Gravitational Field

As follows from the table, the possibility to find exact analytical expres-
sions for ordered and unordered works is based on their representation in

132
terms of the measurable energy carrier flow jie across the system borders 1).
Such a way of the energy exchange definition is valid irrespective of the
fact whether the consequent processes in the system itself are reversible or
irreversible. This gives the most acceptable (to date) solution to the problem
of thermodynamic inequalities.
Besides, a comparison between Tables 6.1 and 6.2 shows that the
true “dividing ridge” in real processes lies not in-between heat and work (as
it has always been in equilibrium thermodynamics), but rather in-between
two qualitatively different categories of effects referred here as to ordered
and unordered works.

6.6. Possible Use of Environmental Heat in Non-Heat Engines

Classic thermodynamics is known to exclude using the heat dissipated

from the environment in heat engines. Here the concept of the perpetual mo-
tion of the second kind as itself is rooted proposed by Nobel Prize winner
W. Ostwald. The ban on creation of such kind heat engines proceeds from
the assumption that the environment is a heat receiver in such machines and
converting it into a heat radiator in the absence of other heat receivers
would mean the creation of a mono-thermal unit using the practically inex-
haustible heat supplies from, e.g. the world ocean.
It should be noted, however, that equilibrium thermodynamics consid-
ered the environment as a single equilibrium whole with no allowance for
the temperature gradients occurred there. Besides, it considered the only cy-
clic heat engines. At such conditions the ban on using the dissipative heat
from the environment directly ensued from the reasoning as described in
Chapter 4 and was quite natural.
The situation radically changes with the world ocean temperature strati-
fication considered. In this case using the ocean as a free energy (ynergy)
source not only does not contradict thermodynamics, but also has been long
used at the oceanic power stations. The similar situation takes place if non-
heat and non-cyclic engines are considered as thermokinetics does.
Let we have an arbitrary flow-type machine converting the flow energy
of some substance Jm = dM/dt, which chemical potential decreases from the
value μ1 at inlet down to μ0 at outlet (at equilibrium with the environment).
The work of such a machine according to (6.3.9) is equal to:

1)
In heat engines the energy carrier transit flow Jiе sometimes changes with additional
entropy flows added to it. This phenomenon is taken into account with the help of the “reheat
factor”. In such cases the flow Jiе can not be factored outside the integral (6.5.6) sign. None-
theless, the expression (6.5.6) is here more practical for use.

133
 Wiе = М*(h1 – h0) – М*(T1s1 – T0s0), (6.6.1)

where h1, h0; s1, s0; T1, T0 – enthalpies and entropies at machine inlet and
outlet, respectively (at equilibrium with the environment); М* – mass of the
working medium passed thru the machine.
Providing the working medium expands reversibly and adiabatically (s1
= s0), the above expression may be written as the algebraic sum of two
components:
Wiе = М*(h1 – h0) - S*(T1 - T0), (6.6.2)

where S* = М*s1 – entropy transferred by the substance flow thru the ma-
chine.
Here the first term of the right-hand side characterizes the work done by
the substance flow at its enthalpy reduction, while the second one – the
work done by the entropy flow at its temperature reduction. This means that
the machine under consideration represents a combined engine with two in-
terconnected flows of energy carriers Jm = dМ*/dt and Js = dS*/dt, wherein
both the intramolecular energy of the substance and its internal heat energy
are converted. If the working medium of such a machine receives heat from
the environment which temperature T1 > T0 and which is thus certainly ca-
pable for work toward the heat receiver, this naturally influences the re-
ceiver-done work. This is, e.g., the case with the Thomson’s heat the
thermo-element electrode absorbs when electric charge is moving therein,
as well as with recovery plants using wasted heat of a relatively low (as
compared with the heat source) potential. Whatever violations of thermody-
namic laws are naturally absent in this case.
Let us now consider the case, when some non-heat engine receives heat
from outside at the T0 ambient temperature. This is, in particular, the case of
an endothermic reaction running in a fuel or galvanic cell. In this case equa-
tion (9.5.2) becomes:

Wiе = М*(h1 – h0) + Q0 , (6.6.3)

where Q0 = М*T0(s0 - s1) – amount of heat received from the environment.

As it follows from this expression, the heat flow, in this case, contribute
to the fuel or galvanic element work. This fact has often been mentioned by
the authors of electrochemical guidelines. However, it would be a mistake
to suppose that the heat energy is here as well converted into ordered forms
of energy. The fact is that in this case the heat flow from outside just com-
pensates the system free energy consumption needed to cover the working
medium bound energy T0(s0 - s1) increasing in course of the chemical trans-
formations. The comprehension of this circumstance is significantly facili-
tated with division of the system energy U into the inergy E and anergy Ū.
According to energy balance (2.5.6) in a reversible process, whenever the

134
system anergy remains unvaried, the useful work Wе done by the system is
equal to the system ynergy decline:

Wе = – ∆U = – (∆E + ∆Ū) = –- ∆E . (6.6.4)

However, if there are some processes running in a system resulting in

its anergy Ū increase (chemical or phase transformations, destructive and
other dissipative processes associated with increasing entropy), the useful
work decreases according to the anergy increase ∆Ū:

Wе = –∆E – ∆Ū ≤ –- ∆E. (6.6.5)

If at these conditions the anergy increase ∆Ū is covered not due to the

ynergy, but due to external anergy sources, e.g. by the heat exchange Q0 =
∆Ū, the work done by the system will still be equal to the system ynergy
decline as in the reversible run of the process:

Wе = – ∆E – ∆Ū + Qo = – ∆E. (6.6.6)

However, the heat supplied to the system in this case is not converted
into other forms of energy, but just replenishes the anergy increase. The ef-
ficiency of such a system must be naturally defined as the work Wе related
to the entirely consumed energy Еi′ = Аr + Qo, but not to only the chemical
reaction affinity Аr = - ∆Е, and therefore is always below unity. This means
that no laws of thermodynamics are violated herein.
Thus the ban on using in heat engines the heat dissipated from the envi-
ronment is once again caused by the arbitrary extrapolation of the thermo-
dynamic laws beyond their applicability. However, the inadmissibility of
such extrapolation at the present state of thermodynamics is far from being
evident.
The in-depth comprehension of the processes running is somewhat fa-
cilitated from the positions of thermodynamics of irreversible processes
(TIP) based on the concept of interaction between various flows and “en-
trainment” of one flow by other (in this case the heat flow entrained by the
substance flow). However, such interaction is restricted in TIP to flows and
forces of the same tensor order, so this theory can not elucidate the connec-
tion between the scalar chemical reactions and vector heat flows. The more
advisable appears to be the consideration of these issues from the positions
of thermokinetics. If from these positions, there is no wonder in the pub-
lished information on the development, patenting and manufacturing of de-
vices using the environmental heat to increase the useful work obtained in
these devices (A. Serogodsky, 1992; S. Klimov,1992; G. Bujnov, 1992; G.
Skornjakov, 1992; N. Zaev, 1992, and others). There are enough of facts
evidencing the increased integrity of fuel, galvanic, electrolysis, etc. devices

135
working on the absorption of heat from the environment. We can not but
hope that thermokinetics will remove the suspicion about them on the part
of official science and will promote searching the most successful engineer-
ing solutions in this field.

Chapter 7

THERMODYNAMIC RESOLUTION OF GIBBS

PARADOX

Among the paradoxes of physics it can hardly be found one more

equally as famous and enigmatic as the “Gibbs paradox”, viz. a statement
that entropy builds up stepwise at changing from identical gases to a mix-
ture of gases arbitrarily little distinguishable in their macro-physical and
micro-physical properties (J. Gibbs, 1950). For a century this fact has not
once become the object of investigation for both physicists and philoso-
phers. To many of its investigators it seemed they could eventually explain
why the entropy jumped with so queer independence from the degree and
character of distingushability between the gases mixed, as well as why the
notion “entropy of mixture” was inapplicable to identical gases. However,
like the legendary sphinx that paradox has been thrashed over on pages of
scientific books and magazines and has not yet left them till nowadays. As a
result, the majority of its investigators have inclined to an opinion that the
“Gibbs paradox is unsolvable on the plane of classic thermodynamics (B.
Kedrov, 1969).
Other vistas open up to this problem from the positions of thermokinet-
ics. This chapter is dedicated to show that the Gibbs paradox is actually
paralogism, viz. an erroneous statement sounding credible due to the statis-
tic-mechanical interpretation of entropy as a measure of “any and all” irre-
versibility. As an alternative, the thermodynamic theory of irreversible mix-
ing processes will be offered as allowing for the dependence of losses from
the nature of gases being mixed.

7.1. Origin and Nature of Gibbs Paradox

In his famous work “On Equilibrium of Heterogeneous Substances”

(1875–1876) J. Gibbs set forth the following expression for the entropy of
an ideal gas mixture:

S = Σk Nk(cνklnT + Rkln υk + sok), (7.1.1)

136
where Rμ – universal gas constant; Nk – mole number of the kth substance;
cvk, υk, sok – isochoric heat capacity, partial volume and entropic constant of
a mole of the kth substance, respectively 1).
Gibbs wrote this expression by analogy with Dalton law to which the
pressure of an ideal gas mixture p is equal to the sum of partial pressures of
the components рk (р = Σkрk). When postulating this “similar principle re-
garding the gas mixture entropy”, Gibbs made no mention whatever of what
the individual characteristics sok and cvk meant having them evidently as-
sumed identical properties of a corresponding pure substance.
It is significant that Gibbs did not at all consider expression (7.1.1) as
rigorously proven. He just assumed it would have been correct to initially
accept this relationship as a fundamental equation describing an ideal gas
mixture and then to substantiate the validity of such definition by properties
which might have been derived from it. Applying this expression to the dif-
fusion at the mixing of ideal gases as two separate masses, each of the gases
initially occupying half a complete volume, he defines that the difference
between the gas mixture entropy S = (М 1R1lnV + М 2R2lnV) and the entro-
pies before mixing constitutes the constant value

S – [М1R1ln(V/2) + М2R2ln(V/2)]= Rс ln2, (7.1.2)

where Rs = М1R1+ М2R2 – universal gas constant of the system as a whole.

Commenting on this result Gibbs notes, “It is significant that the value
of this expression does not depend on kinds of the gases being mixed and
degree of their difference”… since the “value pV/T is entirely defined by
the number of molecules being mixed”. Thus Gibbs himself traced nothing
paradoxical in that result. However as investigators were studying the ques-
tion, they encountered ever growing difficulties, which caused the “Gibbs
paradox” definition.
In voluminous literature dedicated to this question several standpoints
are met regarding the nature of this paradox. A number of investigators (M.
Leontovich, 1951; A. Samoilovich, 1955; P. Chambadal, 1963; S. Fraier,
1973, and others) identify the nature of the paradox with the impossibility of
a limit change to identical gases in expression (18.1.1). In fact, (18.1.1) does
not contain any parameters describing the difference between gases. There-
fore it necessarily follows from this expression that entropy jumps when
portions of the same gas are mixing. Gibbs himself having adhered to the
Boltzmann’s (probabilistic) interpretation of entropy saw nothing queer in
that since a “mixture of the same-kind gas masses in principle differs from
that of the different-kind gas masses” – for lack of information allowing, in
principle at least, to separate them. However such an argument is evidently

1)
Here, unlike the original, 1 mole is adopted as a quantity unit of the kth substance

137
unacceptable from the positions of thermodynamics wherein the initial in-
formation of a system is restricted to definition of the thermal and caloric
equations of state identical for ideal gases.
Some investigators refer to the Gibbs’ theorem itself as a paradox. Ac-
cording to it, the entropy of a gas mixture is equal to the summary entropy
of particular gases, each occupying the volume of the whole mixture at the
same temperature. Gibbs substantiated this statement by an imaginary ex-
periment on the reversible separation of gas mixtures thru semipermeable
membranes. However an imaginary experiment may be used in thermody-
namics to substantiate some statement providing its conclusions do not con-
tradict theory only (K. Putilov, 1974). Therefore many investigators have
not taken this “proof” as convincing. Herein multiple attempts are rooted to
more rigorously prove the entropy additivity in the Gibbs’ concept. The
proofs of the said theorem insomuch offered are reduced to two main cate-
gories:
a) Method of semipermeable membranes, which, besides Gibbs him-
self, was used by Rauleich, 1875; L. Boltzmann ,1878; A. Wiede-
burg, 1894; A. Bik,1903; B. Tamman, 1924; V. Nernst, 1929; P.
Chambadal, 1963; B. Kedrov, 1969, and others.
b) Method of gas column in gravitational field, which, in particular, H.
Lorenz (1927) and E. Schrödinger (1946) used.
All these methods were aimed to define the work of reversible mixture
separation and eventually based, explicitly or implicitly, on the assumption
of ideal membranes capable to provide the so-called “membrane equilib-
rium” (when the gas mixture pressure on one side of the membrane is coun-
terbalanced by the partial pressure of one of the components on its other
side). It is significant that in all imaginary experiments of such a kind after-
investigators discovered a number of inaccuracies and disputable assump-
tions. Furthermore, from such reasoning based on “asymmetrical” semiper-
meable membranes (letting gas through in only one direction) results were
obtained antipodal to the Gibbs’ theorem (P. Chambadal, 1963). Those
proved the additivity of component entropies found at the total pressure and
temperature of the mixture when no entropy jump appeared at all.
A number of investigators (V. Luboshits, M. Podgoretsky, L. Gelfer,
1971, 1975; E. Gevorkian, R. Gevorkian, 1975, 1976) adopt a neutral atti-
tude toward the Gibbs’ paradox considering the entropy jump as quite natu-
ral for gases modifying their properties discretely. In this case it is unclear
how far different (from a thermodynamic standpoint) such substances are
as: isotopes (different in molecular mass, but equal in chemical properties),
isobars (different in chemical properties, but equal in molecular mass), iso-
mers (different only in their life span in excited state), optical antipodes
(different optically due to different spatial grouping of molecules), etc.
Depending on investigators’ attitude toward the Gibbs’ paradox their in-
terpretation of its “solution” varies. The overwhelming majority of investi-

138
gators accept the statement of the mixing process entropy existing as a true
one though sounding somewhat unusual and incredible. These investigators
after Gibbs refer the entropy jump at the mixing to (1) principal impossibil-
ity of after-separation of the same-kind-gas mixture; (2) principle difference
of physical and chemical properties of gases regarding the character of their
variation (M. Planck, 1925); (3) discrete variation of atomic properties (E.
Schrödinger, 1946; R. Kubo, 1970; A. Samoilovich, 1955; A. Sommerfeld,
1955; L. Terletsky, 1966; I. Bazarov, 1976); (3) density discontinuity at the
mixing of various gases (P. Lelouchier, 1975); (4) some work to be done to
create partial pressures (B. Kedrov, 1969), etc. Other investigators see the
solution to the Gibbs’ paradox in proving the fact the mixing process en-
tropy depends on the degree the gases differ from each other (V. Luboshits,
M. Podgoretsky, 1971; Y. Varshavsky, A. Sheinin, 1968; R. Gevorkian, E.
Gevorkian, 1976), e.g., for a mixture of the same gases with a continuously
equalizing composition.
It is just a minor part of investigators (J. Van der Vaals, F. Konstamm,
1911; P. Postma, 1927; P. Chambadal, 1963; A. Veinik, 1967; B. Casper, S.
Fraier, 1973; M. Biot, 1977) including the author of this book (V. Etkin,
1973, 1991), who deny any entropy variations at the mixing of non-
interacting gases, which is the most radical solution to the said paradox.

7.2. Thermodynamic Inadmissibility of the Gibbs’ Paradox

There are a number of arguments evidencing that the Gibbs’ paradox is

actually paralogism, viz. an erroneous statement sounding credible in the
conviction that entropy rises in any irreversible process. It is impossible to
reproduce herein all arguments of various authors to substantiate this thesis.
Therefore we will not go beyond those of the arguments which are of the
methodological character and therefore sound most convincing.
Classic thermodynamics dealing with only closed systems is known to
have been interested in only the variation of entropy, but not its magnitude.
This entropy variation in course of some process does not depend on
whether a system is considered as a mixture of the kth ideal gases or as a set
of the same ideal gases separated with a movable heat-permeable membrane
since from the thermodynamic standpoint all properties of a system are de-
fined by exclusively its thermal and caloric state equations. This entropy
variation for a system with an arbitrary and constant (in whole) composition
is derived from a known expression:

ΔS = ΣkNkRµlnT/To – ΣkNkRµln р/рo, (7.2.1)

where To, рo and T, р – absolute temperature and pressure of the gas mixture
at the beginning and at the end of whatever process, respectively.

139
 It follows from the identity of equations (7.2.1) for a gas mixture or a
set of pure gases that from the positions of thermostatics they have the same
number of degrees of freedom. This number is defined by the number of in-
dependent variables of state and is equal to 2 for gases (thermal and me-
chanical degrees of freedom). Hence the thermodynamic properties of a sys-
tem under consideration (either a gas mixture or a set of pure gases) are to
the full extent characterized by two (thermal and caloric) state equations in
the form:

pV = МRT; U = CνT, (7.2.2)

where R, Cv = ΣkNkсvk – universal gas constant and total isochoric heat ca-
pacity of the system, respectively, found experimentally without knowing
its composition and studying properties of its components separately. At
these conditions the definition of any other properties of the system, e.g. its
composition, is superfluous. Gibbs himself admitted this fact having noted
that for a constant-composition system the “state is completely character-
ized by the total mass M so that the knowledge of the composition of a sys-
tem is not the necessary condition to derive its state equations”. Hence both
the gas mixture entropy and the gas set entropy as functions of system state
were defined by two parameters of system state (T, p or T, V) and due to
their constancy at isobaric-isothermal mixing remained unvaried. Thus from
a thermostatic standpoint none process ran, the more so because none of en-
ergy effects were observed at that. In fact, the aggregate system with two
degrees of freedom yet before mixing was in total (both thermal and me-
chanical) equilibrium. Gibbs quite realized that when noted that the “prob-
lems of thermodynamics refer just to the system states defined by such in-
complete way”. Therefore the Van der Vaals’ standpoint (1911) is quite rea-
soned when he noted regarding the mixture of isotopes, “However, from a
thermostatic standpoint the mixture of such substances should be considered
as a single substance and, since entropy is defined as thermostatic, there are
no reasons to talk of an entropy rise at diffusion”.
Another contradiction is revealed in the thermodynamic approach to the
problem of entropy constants’ additivity at the mixing. In fact, Gibbs’ equa-
tion (10.1.1) was based on the analogy of the fundamental equation with
Dalton law
р = Σk MkRμT/V (7.2.3)

With regard to equation (7.1.1) Gibbs notes that this expresses a known
principle, according to which the pressure of a gas mixture is equal to the
sums of the pressures the components of this gas mixture would have pro-
viding they exist separately at the same volume and temperature. Thus

140
Gibbs was explicitly based on additivity of entropy for each of the compo-
nents as expressed by the relationship S = ΣkNksk.
Let us clarify now whether expression (7.1.1) complies with this rela-
tionship if the value sok assumed constant in processes of system composi-
tion variations. The additivity (summability) of whatever extensive parame-
ter is known to suppose its specific value does not depend on mass. In other
words, additive values are homogeneous mass functions, i.e. comply with
Euler’s theorem which in entropy application has the form:

∂Sk(Nk)/∂Nk = Sk(Nk)/Nk . (7.2.4)

Expression (7.2.4) is compatible with (7.1.1) when the derivative

(∂Sk/∂Nk) does not depend on Nk. However, it is easy to see that at sok =
const the entropy Sk of a particular component does not meet this require-
ment. It really follows from (7.1.1):
∂Sk(Nk)/∂Nk = sk(Nk) – Rµ, (7.2.5)

i.e. is a function of Nk. Hence the Gibbs’ assumption that sok = constant and
the same before and after mixing is groundless.
One more contradiction revealed by B. Kedrov (1969) is that the varia-
tion of ideal gas mixture entropy depends on the process path. Let a vessel
A contain a mixture of two moles H2 (hydrogen) and two moles Cl2 (chlo-
rine). The vessel is kept in darkness so that a chemical reaction within it is
practically “inhibited”. Let us separate the mixture with a partition into two
equal parts and initiate in one of them by exposure to light an isobaric-
isothermal chemical reaction resulting in generating two HCl molecules. As
a result of the chemical reaction, the entropy of this part of the system will
vary by some value ΔSx. Let us now remove the partition between the first
half of a system containing the H2+Cl2 mixture and the second half contain-
ing the HCl gas. Owing to the fact that the gases are different in both halves
of the vessel, according to the Gibbs’ mixing theory the system entropy will
rise by some value ΔSmix. Now let us expose the both halves of the vessel to
light – the reaction will develop further with one more HCl mole generated
and the entropy further varied by ΔSx. The total entropy variation in the
three said processes resulted in two HCl moles generated is equal to 2ΔSx
+ΔSmix. However, the same two HCl moles could be obtained by exposing
the mixture to light as a whole, i.e. without its separation followed by mix-
ing. In this case the gas mixture entropy would evidently vary by only a
value of 2ΔSx. Since the initial and final state of the system and the heat ef-
fect of the reaction are the same in both variants, a contradiction is present.
Further contradiction is revealed when using the mixing process entropy
to calculate the exergy (capability for technical work) losses of the Ex sys-
tems at the mixing of gases. According to the Gibbs’ mixing theory which

141
does not involve any parameters characterizing differences between gases,
the exergy loss ΔExmix at the mixing of substances featuring ideal gas state
(i.e. complying with the Clapeyron equation) is defined by exclusively the
mixing process entropy ΔSmix and the environmental temperature T0 and
does not depend on the chemical nature of the substances being mixed:

ΔExсм = ToΔSсм . (7.2.6)

Let us consider, however, a fuel cell, to which electrodes, e.g., oxygen

and hydrogen are fed separately under a minor pressure (so that their state
would not differ from ideal gas). The chemical affinity in reversible fuel
cells is known to be realized in the form of electric current work which is
theoretically equal to the chemical affinity value. Now let us mix oxygen
and hydrogen partly or completely before feeding them to the fuel cell elec-
trodes, i.e. let us feed not pure gases, but some oxy-hydrogen mixture. The
similar experiments have repeatedly been conducted and are known to have
led to drop of the voltage developed by the fuel cell down to total disap-
pearance of current in the external circuit. Hence the actual loss of fuel cell
capability for work depends on the nature of gases being mixed (their
chemical affinity) and reaches the 100% value when the reaction becomes
thermodynamically irreversible. This example is even more remarkable be-
cause allows to distinguish the losses at the mixing and chemical transfor-
mation. From this example it follows that the major losses arise not during
mixing of gases, but in subsequent chemical reaction which due to this be-
comes thermodynamically irreversible. Therefore it would be more correct
to refer in this case not to losses at the mixing, but rather losses due to mix-
ing.
Inapplicability of the Gibbs-obtained result shows up when as well es-
timating the capability for work of whatever substance which concentration
differs from its environmental concentration1). Classic thermodynamics ac-
cording to the Gibbs’ theory gives the following equation for the exergy of
working medium in flow Exр (Szargut J., Petela R.,1968):

Exр = H – Ho –To(S – So) + RсTo Σk ln рk/рok, (7.2.7)

where H, Ho и S, So – enthalpy and entropy of a working medium (gas) at

initial state and at equilibrium with the environment, respectively; рk , рok –
partial pressures of the ith substance in initial mixture and in the environ-
ment, respectively.
The last term of this expression defines the so-called “chemical” (more
exactly, concentration) exergy caused by the difference between partial
pressures of the kth substances in the system pk and the environment pok.

1)
This allows in-principle constructing an engine using this concentration difference

142
This may supposedly be realized with the help of semipermeable mem-
branes which allow isothermally expanding the gas from the pressure pk to
pok in an expanding machine with heat obtained from the environment and
useful external work done.
As follows from this expression, the unit mass exergy for an ideal gas
does not depend on its chemical nature and tends to infinity as its environ-
mental concentration is decreasing. Inapplicability of such a conclusion is
evident.
Thus estimating the Gibbs’ approach to the gas mixing theory it has to
be admitted that this gives no answer to not only the most important ques-
tion about criteria of difference or identity of gases being mixed, but either
about the theoretical value of work to be done to separate the mixture. The
experiment shows that the less the difference between the mix components,
the more the work on gas separation, even to say nothing of the test hard-
ware imperfection. In particular, when producing nuclear fuel by separating
a gas mixture containing 99,3% U238F6 and 0,7% U235F6, it is theoretically
required (allowing for mixing process entropy) 0.023kWh of energy per 1kg
of the second component. However, the actual energy consumption amounts
to 1.2⋅106 kWh, i.e. approximately fifty million times as much (J. Ackeret,
1959). Thus the Gibbs’ mixing process entropy can not serve as a basis for
even approximate estimation of the theoretical mix separation work.

7.3. Entropy Reference Point Shift in Mixing Process

as Entropy “Jump” Reason

Far from all investigators of the Gibbs’ paradox have related this with
the change to investigation of open system in the Gibbs’ concept. In fact,
considering the ideal gas mixture entropy or ideal gas set entropy from the
positions of “pre-Gibbs” thermodynamics of closed systems as a composi-
tion function, i.e. S = S(T,p,Nk), the entropy at the mixing will remain un-
varied since the temperature, pressure and mole numbers Nk (masses Mk) of
all system components remain unvaried at that. In other words, for the dif-
fusion process in its intrinsic meaning as the concentration equalization in a
closed system the entropy remains unvaried despite the irreversibility of this
process.
Let us consider now the isobaric-isothermal mixing process in the
Gibbs’ concept as a composition variation in each of the open subsystems
due to the exchange of the kth substances among them (i.e. a mass transfer
among them). In this case, because additional degrees of freedom appear
(related to the kth substances exchange), the exact differential of the system
entropy becomes:

dS = (∂S/∂T)dT + (∂S/∂р)dр + Σk (∂S/∂Nk)dNk . (7.3.1)

143
 The first and the second partial derivatives in this expression are defined
at constant composition and mass of the whole system (Mk = const, ΣkMk =
const) and may be found from the joint equation of the 1st and the 2nd laws
of thermodynamics for closed systems. In particular, for the derivative of
entropy with respect to temperature from (7.1.1) via caloric state equation
(7.2.2) the following forms may be found:

(∂S/∂T)V,N = Cν /T ; (∂S/∂Р)T,N = – MRс/Р. (7.3.2)

As for the derivative (∂S/∂Nk)Т,V, this can not be defined only based on
the laws of thermodynamics for closed systems. To define this, relationship
(6.5.1) should be applied, which gives:

(∂S/∂Nk)Т,V = sk , (7.3.3)

where sk – partial molar entropy of the kth component, i.e. the value charac-
terizing the entropy S rise in an open system when one mole of the kth sub-
stance enters in it at isobaric-isothermal conditions, whereas the mole num-
ber of other, jth substances, does not vary (k ≠ j).
Thus for open systems the exact differential of the entropy S = S(T,p,Nk)
is:

dS = (Cν /T)dT + (MR/V)dV + Σk skdNk. (7.3.4)

Integrating this expression from an initial arbitrary state with the en-
tropy S0 assuming Cv constant and allowing for the relationship dV/V =dυ/υ
evident at N = const gives:

S = CνlnT + NRµln υ + Σk∫skdNk + So . (7.3.5)

This expression differs from (7.1.1) offered by Gibbs by the third term
appeared on the right-hand side and describing the entropy variation at gas
mixing. This is what Gibbs neglected when having integrated equation
(7.3.4) with respect to only the variables T and V. As a result, he defined the
entropy to an accuracy of some function of state So(Nk), i.e. obtained not the
entropy constant, but some function of composition as the sum of the two
last terms on the right-hand side of expression (7.3.5). This value necessar-
ily varies at the mixing in course of diffusion of the kth gases.
It may be easily shown that at the mixing to the Gibbs’ concept both the
mix entropy and its reference point, i.e. the So(Nk) value, vary simultane-
ously and equally.

144
 Let us consider the same gas set system which Kedrov used in his
imaginary experiment when mixing H2 and Cl2. Let us assume after Gibbs
that sok = 0, So = 0. Let us further transfer the system via an arbitrary quasi-
static (e.g. isochoric) process to a state with a temperature of T1 and volume
V1. The entropy of the system will then rise by a value ΔS0-1 =
ΣkNkRµlnT1/To. Now let us remove the partition and provide the isobaric-
isothermal mixing where the system volume remains unvaried (V1=V2),
whereas the entropy according to the Gibbs’ theory rises by some value
ΔSmix and becomes equal to S2 = S1 +ΔSmix. Then let us cool the mixture ob-
tained down to a state with a temperature of T3 = T0. The entropy will sub-
sequently decrease by a value of ΔS2-3 = -ΔS0-1. The system has again re-
turned to the state with the same temperature and volume, however, now the
entropy in this state (which we adopt as initial) is equal to S3 = S2 + ΔS2-3 =
So + ΔSmix. Thus the gas mixture entropy value at the reference point pa-
rameters has varied by the exact mixing process entropy value! In other
words, in a diffusion process as the subject of the Gibbs’ concept not the en-
tropy itself experiences a jump, but its reference point! Other result could
hardly be expected since the Gibbs’ mixing process entropy does not de-
pend on temperature and hence is the same both in its current state and an
arbitrary reference point. As a matter of fact, applying expression (7.1.1) to
an arbitrary reference point for entropy any unbiased investigator would ar-
rive at a conclusion that this reference point has as well experienced the
same jump. Thus the entropy jump, should it really take place, equally re-
lates to also the entropy reference point since this jump depends on only the
ratio of mixed gas volumes before and after mixing. However, to justify the
fallacy, it should be noted that in the days of Gibbs the problem of the en-
tropy reference point selection and the entropy magnitude definition, which
has eventually led to the third law of thermodynamics, did not yet exist.
This is the circumstance that, in our view, engendered the Gibbs’ paradox. It
should seem so that the after-investigators could not have omitted the fact.
They knew the third law of thermodynamics. It is this law that defines the
reference point for entropy of all condensed substances. As a matter of fact,
according to the third law of thermodynamics “as temperature is approach-
ing the absolute zero, the entropy of any equilibrium system in isothermal
processes ceases depending on whatever thermodynamic state parameters
and to the limit T = 0 adopts a constant value, the same for all systems,
which may be assumed as zero” (I. Bazarov, 1976). Therefore such a shift
of the entropy reference point comes into antagonism with the third law of
thermodynamics which reads that the entropy of whatever equilibrium sys-
tem at the absolute zero temperature adopts a constant value, the same for
all systems, which may be assumed as zero. Thus the entropy reference
point jump as ensuing from the mixing process to the Gibbs’ concept leads
to a conflict with the third law of thermodynamics. This fact, which is, as

145
far we know, beyond other investigators’ comments, exactly reveals the
paralogism of the Gibbs’ paradox.
From the above-mentioned it becomes clear why some of the investiga-
tors, based on imaginary experiments, came to the necessity of calculating
the mixture entropy from the mixture total volume, whereas the others –
from the total pressure and temperature of the mixture. The fact is that both
of these standpoints are equally valid and applicable since neither of them
leads to a jump of entropy at the mixing of non-interacting gases.
Hence from the positions of thermokinetics as well the Gibb’s conclu-
sion of a step rise of entropy at the mixing of ideal gases appears as an erro-
neous statement caused by the arbitrariness in choosing the reference point
of entropy for open systems, i.e. by the violation of the third law of thermo-
dynamics. This result shows that in thermodynamics the Gibbs’ paradox
does not take place whatever meaning is read into it. As for statistic and in-
formational entropies, the jump is not something paradoxical here since the
number of possible permutation of particles depends on whether the gases
are considered identical or distinguishable.

7.4. Thermodynamic Theory of Mixing Processes

The gas mixing theory must give a solution to two problems, viz. the
definition of the useful (free) energy of a particular mixture and the work
needed for its separation. The answer the Gibbs’ theory gives solves neither
of them. This is a challenge to approach the issue from the positions of
thermokinetics. Thermokinetics considers the mixing as an irreversible
process of equalizing the concentrations of the components all over the sys-
tem volume while keeping their number unvaried for the system as a whole.
This corresponds to the strict import of the word “diffusion” (from the Latin
“diffusio” meaning spreading). Such a redistribution of components in-
volves the variation of the moment of the kth substances’ distribution Zk. As
any irreversible process, this may be maintained from outside, e.g. by feed-
ing the kth substances across the system borders. However, this is independ-
ent of the transfer of these substances across the system borders (diffusion
across the borders), i.e. of the diffusion to the Gibbs’ concept. The latter, for
the avoidance of mishmash, we have named above the selective mass trans-
fer of a system. The considered process involves a variation of the mass of
the whole system М = ΣkMk or the mole number in it N = ΣkNk. Unlike this,
the diffusion leaves M and N unvaried. In fact, the mixture component dis-
tribution may be equalized also in the absence of a substance transfer across
the system borders as it occurs, e.g. under the influence of electric or cen-
trifugal fields. Thus the mixing is a specific qualitatively distinguishable ir-
reversible process irreducible to other processes. This can not be ap-
proached by the “adaptation” to whatever other process.

146
 The diffusion coordinate for the kth independent substance is the mo-
ment of its distribution Zk, while the motive force Xk – the component po-
tential concentration gradient which is defined by the uniqueness conditions
of the mixing process and, according to paragraph 6.1, above means the
negative gradient of diffusive, osmotic, etc. potential. The diffusion process
is irreversible and involves thermal and bulk effects resulting in internal
sources of entropy and volume appeared. To find these sources, let us repre-
sent the entropy S and the mixture volume V in terms of their partial molar
values sk and υk, respectively:

S = ΣkNksk; V = ΣkNkυk , (7.4.1)

and considering (6.1.3) obtain:

ΔSmix = ΣkNk (sk – sko); ΔVmix = ΣkNk (υk – υko). (7.4.2)

The value (sk – sko) in this equation characterizes the variation of entropy
of the kth pure substance sko in the mixing process, while the associated heat

qk∗= T(sk – sko) (7.4.3)

– thermal effects arising at the mixing of interacting components1). These

effects are caused by the available partial molar entropy (defined by an ac-
tual increment in mixture entropy sk at isobaric-isothermal input of the unit
kth substance) and by the entropy sko superinduced from outside by one mole
of the pure kth substance. This means that the above value belongs to the
thermodynamic function of mixing. Similarly the difference ΣkNk(υk – υko)
characterizes the bulk effects arising at the mixing of interacting compo-
nents, while the value
wk∗= р(υk – υko) (7.4.4)

– cubic strain work involved in these effects. For non-interacting substances

the thermal and the bulk effects (sk – sko; υk – υko) at input of the kth sub-
stance are absent (qk∗= wk∗= 0), which confirms the above conclusion that
the ideal gas mixing process entropy is absent.

1)
In the theory of irreversible processes (TIP) the value qk∗ is introduced as one of the trans-
port factors Lkj, while its interpretation as the energy transported by one mole of the kth sub-
stance in the absence of temperature gradient becomes possible just “a posteriori” (by ex-
perimental results)

147
 Thus the thermal and the bulk effects qk∗ and wk∗ described by expres-
sions (7.4.3) and (7.4.4) may serve as a measure of mixing process irre-
versibility. These relationships allow answering all questions raised above
in the thermodynamic mixing theory. According to them the theoretical
work on mixture separation depends on the nature of gases being separated
and for non-interacting gases becomes zero along with the difference sk –
sko. This circumstance leads to a necessity to distinguish the vector and the
scalar stages of an energy dissipation process. Let us elucidate this by ex-
ample of a fuel cell realizing the chemical affinity of reagents in the form of
electric work. If the reagents, before fed to the fuel cell electrodes, are al-
lowed to be completely mixed, the fuel cell emf is known to fall down to
zero. The mixing process has the vector nature according to the tensor order
of its coordinates Zk. Though this stage is irreversible and involves thermal
and bulk mixing effect, it just a little changes the value of reagents’ chemi-
cal affinity. However, this stage results in that the subsequent homogeneous
chemical reaction becomes thermodynamically irreversible with the charges
not separated and useful work done, but with a heat released in amount
equivalent to this work. Thus the capability for work is lost as a result of the
spatial homogeneity spontaneously set in for the chemically reacting sys-
tem, but not the subsequent chemical reaction itself. The chemical energy
appeared to have already been dissipated by the beginning of this reaction
and there was nothing for it but to pass into the random energy (heat). In
other words, the chemical energy of the reacting mixture appeared to have
already been less ordered than the initial energy of the spatially separated
reagents. It is referred to the fact that instead of a macro-heterogeneous sys-
tem we have got the micro-heterogeneous one differing in just the structure
(configuration) of the molecules and atoms comprising it. However, this is
not yet the heat energy of reaction products! Thus we come to a necessity to
distinguish between the macro-physical stage of a dissipation process in-
volving the disappearance of system spatial heterogeneity (it is expressed
mathematically by “scalarization” of the process, i.e. loosing its vector na-
ture) and its micro-physical stage associated with scaling heterogeneity
down to a mere random form of energy obtained. Using equation of dis-
placement vectors’ balance (2.6.2) the first stage may feature the value

Wd = ΣiFi·duRj, (7.4.5)

which may be named the “dissipation micro-work”.

The second, micro-physical, stage of the dissipation process reflects the
destruction of a chemically reacting system. This stage as well involves the
internal “disggregation work” (R. Clausius) done. However, this work fea-
tures already the scalar character and must be described as the “dissipation
micro-work”. This is the work that defines the decline of free energy of a

148
chemically reacting mixture. In a general case this category of work should
include not only the scalar chemical reaction work, but as well any work as-
sociated with further “disordering” of the substance (rearrangement of its
molecular, crystal, cluster, etc. structure).
Hence the standard affinity of a homogeneous chemical reaction is
partly consumed to prepare the reaction mixture (vector stage of the process
not obeying the stoichiometric proportions), and then – to run the reaction
itself (V. Etkin, 1991). This may be the reason why the standard affinity of a
reaction can not play the part of its thermodynamic force (G. Gladyshev,
1991). Anyway, the possibility to distinguish between the thermal effects of
mixing and subsequent chemical reaction may serve as the incentive to fur-
ther investigations.

Chapter 8.

DENIAL OF THE THEORY OF “HEAT DEATH”

OF THE UNIVERSE

Among the problems of classic thermodynamics reflecting the scientific

view of the world the conclusion that heat death of the Universe is inevita-
ble takes the high-end position. This problem has repeatedly been a hot
topic both for physicists and philosophers. In this chapter we are consider-
ing this issue from more general positions of thermokinetics and making an
attempt to expose the sources of this mistake which has led to the concept of
heat death of the Universe contradicting the today’s vision of the world.

8.1. Reasons of Appearing Concept of “Heat Death”

of the Universe

The authorship of the heat death theory belongs to R. Clausius, the

founder of thermodynamics, who came to such a conclusion based on the
law of entropy rise he himself formulated. The concept of entropy as itself
was introduced by him due to a necessity to find the reversible heat ex-
change coordinate, i.e. a physical value necessarily varying at this process
and remaining constant in its absence. Why shouldn’t this value have kept a
low profile as just one of the system energy independent arguments! How-
ever, R. Clausius noticed that, unlike other state coordinates (mass, volume,
charge, momentum), entropy did not remain constant in isolated systems
and spontaneously rose as such systems were approaching equilibrium. That
unusual property of entropy as the only known then inconstant value re-
149
strained Clausius for some time from publishing the results of his investiga-
tion. However, the logic of the thermodynamics development and the inten-
tion to use that property of entropy to find the criteria of evolution and equi-
librium for thermodynamic systems had such an outcome that Clausius not
only raised the entropy rise law to the rank of the second law of thermody-
namics, but also spread it over the entire Universe. Such an extrapolation of
the entropy rise law was best of all manifested in his pithy phrase, “energy
of the Universe is constant, Entropy of the Universe is rising”. Clausius’
contemporaries immediately traced far reaching consequences from that
conclusion –from a disposition of Providence in the “Creation of the Uni-
verse” and to the inevitability of its “heat death”. To a good few of investi-
gators such an extrapolation of the entropy rise law seemed extremely un-
convincing yet in that time (J.M. Gelfer, 1979). However, despite the hot
arguments over the issue, the status of the entropy rise law has not practi-
cally changed. In this connection it is of interest to clarify how the entropy
has converted from an independent parameter of equilibrium system into the
function of non-equilibrium state of the Universe as a whole.
One on these reasons, strange as it may appear, is the absence of other
entropy-independent criteria of thermodynamic systems’ evolution. The fi-
nite-rate real processes are known to be a sequence of non-equilibrium
states. The relaxation processes inevitable in this case cause spontaneous
variations of the state parameters. That fact was first reflected in entropy
balance equation (4.3.2) and became a ground to declare the entropy rise
law. Now we are able to show that the similar principle may be as well set
up regarding other state coordinates of a poly-variant system.
As shown in Chapter 9 the efficiency of any cyclic engine converting
energy of whatever source demonstrates the same behavior as the heat en-
gine. In this case the R. Clausius’ reasoning to substantiate the entropy rise
law (Chapter 4) may be as well extended to non-heat engines. Let us con-
sider an arbitrary cycle of a non-heat engine by substituting the coordinated
T and S in Fig. 4.1 for Ψi and Θi. Following Clausius, let us divide this cycle
with a series of vertical lines into a number of elementary reversible cycles
similar to Carnot cycle. In each of these cycles the energy Еi′ is supplied at a
constant value of the potential ψi', while the same-kind energy Еi" is re-
moved at a constant value of the potential ψi". Then according to (6.2.1) the
elementary cycle efficiency will be defined by the relationship:
ηi = Wj /Ui΄ = 1 – ψi˝/ψi΄. (8.1.1)

Let us accept after Clausius that the efficiency ηi of any irreversible en-
gine is less than that of a reversible one (at the same values of potentials
characterizing the source and the receiver of the ith kind energy – ψi' and ψi",
respectively). Then the sign of equality in relationship (8.1.1) may be sub-
stituted for the sign of inequality:

150
 ηi ≡ 1 – đUi˝/đUi΄ < 1 – ψi˝/ ψi΄ . (8.1.2)

Allowing for the signs Ui΄ < 0; Ui˝ > 0, based on (8.1.2) gives đUi΄/ψi΄ +
+ đUi˝/ ψi˝ > 0, i.e.

dΘi > đUi/ψi. (8.1.3)

Thus entropy is by no means the only parameter spontaneously varying

due to irreversibility. This is the way the following parameters behave: the
masses of the kth substances Mk (due to phase transitions and chemical reac-
tions), their momentum Мkvk (due to relative motion dying out), the value V
(at extension into void), the polarization vectors P and magnetization vector
M (due to relaxation), etc. Therefore A.A. Guhman (1947) is right telling
that entropy has become the evolution criterion on account of a number of
historical reasons. A number of other mistakes hereafter described have as
well contributed to this.

8.2. Discrimination between Thermodynamic and

Statistical-Informational Entropies

One can hardly find one more concept in the world literature, which
would give raise to the same amount of disputes, loose talks and insinua-
tions as entropy. Tens of books and hundreds of articles have been dedi-
cated to entropy. This concept has long outstepped the borders of physics
and penetrated into the inmost of human mind. The statistical, informa-
tional, mathematical, linguistic, etc. entropies have appeared. Truth and er-
rors have been so deeply intertwined that it is practically impossible to date
to get to the historical, epistemological and pragmatic roots of this concept.
Besides, no alternative to entropy made the job a blind alley. Only introduc-
ing the concept of ynergy as an analog of the Gibbs’ free energy – more
universal, physically more evident and more feasible – engendered hope to
once and for all loose the fetters of entropy misreading.
As shown in Chapter 3, entropy is just one of the independent state
coordinates, which serves as an extensive measure of random momentum in
a system. This value relates to the Helmholtz bound energy TS in the same
manner as the ordered momentum Мv with its work-kinetic energy v2 (dou-
ble kinetic energy). The random momentum value varies in only the thermal
process1), whereas the system energy – in any processes running in the sys-

1)
Remember that the thermal process is understood, after K.M. Putilov, as a process involv-
ing the internal thermal energy variation (see Chapter 3).

151
tem. The internal thermal energy Uт or the bound energy TS varies due to
the heat Q supplied from outside or the dissipation heat Qd released in the
system. However, the Qd releasing and the associated entropy rising is just a
part of the state variations caused by the energy dissipation. In particular,
when metals are cutting or hard material crushing, a part of the work herein
done is consumed for the destruction of the materials and for their surface
(potential) energy variation. As a result, the heat output ratio (the heat re-
leased related to the work done) in such processes appears to be less than
unit. Hence, the dissipation of energy (reduction of its free (reversible) part)
may exceed the amount of the heat released and is not therefore defined by
only the entropy rise. The issue of entropy as one of the parameters rising
with energy dissipation should seem to be closed here. However, in 1911
Boltzmann, who did not wish to make up his mind to the inevitable “heat
death of the Universe”, proposed other substantiation of this principle. That
proceeded from the assumption that the entropy rise in irreversible proc-
esses reflected the tendency of nature toward a more probable state. In case
of ideal gas with non-interacting particles this corresponds to such their ve-
locity distribution, which is realized by the most of means (i.e. most prob-
able). Then the relationship between the entropy S and the said “thermody-
namic probability of state” Ŵ has the form:

S = kБ ln Ŵ , (8.2.1)

where kB – constant subsequently named after Boltzmann.

According to this expression the entropy of thermodynamic systems is
proportional to the logarithm of probability of their state Ŵ. When having
derived expression (8.3.1), the main assumption was that the most probable
distribution of particles was at the same time equilibrium one. That was
based on the fact that both said values (entropy S and thermodynamic prob-
ability of state Ŵ) were additive and reached an extreme in equilibrium
state. Since the maximum Ŵ is associated with the “molecular chaos” state,
the entropy in the Boltzmann’s concept has been construed as a measure of
the system state disorder. As a result, it has been converted from the re-
versible heat exchange coordinate into a global measure of “chaos”. The
development of this concept has led to the entropy in a number of up-to-date
theories being attributed to not only the Universe, but even to objects not at
all having the heat form of motion, including the physical vacuum.
In this connection the question is quite natural, to what extent the
Boltzmann’s principle is sound as stating the proportionality between the
entropy and the state probability logarithm. These two values being corre-
lated does not mean at all these are related via a unique dependence of the
(8.3.1) type. Entropy is after all not the only value spontaneously varying in
one direction. Many of the independent coordinates Θi vary in the same way
in an isolated system, as well as such functions of state as the Helmholtz F

152
and the Gibbs G energies, which much more profoundly reflect the variation
of a state than the bound energy TS. Besides, these, unlike entropy, vary in
the same direction with the Boltzmann “collision integral”. These are pa-
rameters that should seem to be related to the probability of state rather than
the entropy as one of the arguments of these characteristic functions.
Broadly speaking, there are a lot of questions to be asked of L. Boltzmann
(P. Chambadal, 1967). All of them evidence that entropy has become a
probability measure just on subjective accounts.
We will not discuss here the question to what extent the phenomenol-
ogical (i.e. experience-based) thermodynamics may be “substantiated” by
the methods of statistical mechanics proceeding from a number of hypothe-
ses “containing much vagueness” (R. Kubo, 1968). We will not either ex-
plore how close to the primary principles of thermodynamics the “Boltz-
mann principle” lies, which identifies the thermodynamic entropy with the
state probability logarithm, as well as to what extent those assumptions are
correct the Boltzmann statistical entropy is based on. Let us just note the
differences in their physical meaning and behavior. To start with, the en-
tropy in thermodynamics is a carrier of the heat form of motion, i.e. the
value that can be transferred across the system borders in course of the heat
transfer or mass transfer between the system and the environment. This fact
is reflected in thermodynamics of non-equilibrium processes as the concept
of an “entropy flow” similar to the substance flow, charge flow, etc. It is ab-
solutely senseless to speak of a “state probability” transfer across the system
borders. Furthermore, as the math modeling of mixing processes shows, the
tendency of the particle distribution probability toward rising appears in this
process yet at a number of molecules equal to or exceeding three at an arbi-
trary small interaction between them, i.e. at the conditions quite irrelevant in
application to a thermodynamic system. Therefore, the fact itself the most
probable state has been reached is not yet a sufficient evidence of thermo-
dynamic equilibrium. In other words, equilibrium and chaos are distinguish-
able notions. A special part is assigned here to metastable states which do
not correspond to the probability maximum being an equilibrium state vari-
ety, though.
Different behavior of the thermodynamic and statistical entropies may
be additionally instantiated by also the spontaneous ice crystallization in su-
percooled liquid or by the precipitation in oversaturated solution involving
its structure ordering (i.e. Boltzmann and Gibbs entropies decrease) and, at
the same time, temperature rise and thermodynamic entropy increase. Scien-
tists from Fourier University and ILL (Institut Laue-Langevin, Grenoblois,
France) have recently obtained a substance that solidifies when heated and
melts when cooled, i.e. behaves exactly opposite to the statistical entropy.
At last, when placing a system in the field of external potential forces, its
state as well becomes more ordered, whereas the thermodynamic entropy
remains unvaried at that (due to the adiabatic action of these forces). The

153
difference between the thermodynamic and statistical entropies shows also
at estimation of their values by the temperature of the relict radiation filling
the Universe. The statistical temperature of this radiation found by the aver-
age velocity of cosmic particles’ motion exceeds 2,000 K, whereas the
thermodynamic temperature found from its spectral characteristics is below
3 K. Thus quite enough experimental facts have been accumulated up to
date evidencing that the thermodynamic entropy and the statistical entropy
are far from being the same notion.
Even further in its physical meaning the informational entropy stands,
which was introduced by Shannon (1949), as many investigators suppose,
just “due to a careless application of this term”. He found that the volume of
information obtained from measurements on a system was related to the as-
sociated state probability variations of the system via the same relationship
(up to a sign) as for the statistical entropy. This formal likeness between the
thermodynamic entropy expression and the information volume decrease
led him (and after Brillouin (1955) – also other investigators) to their
groundless identification.
In fact, the notion of information is ambiguous and even now is still in
the making. One of its early definition is semantic and means a new knowl-
edge obtained from outside. This definition is rather subjective as the vol-
ume of information in the same message is different for people with differ-
ent knowledge. Fisher endued this notion with quite other meaning utterly
excluding the semantic (notional) matter from consideration. The informa-
tion by Fisher is associated with the math expectation of indeterminacy
evaluation and mathematically expressed by a negative logarithm of prob-
ability of an experimental outcome (the higher the expectation of whatever
experimental result, the less the new information that can be taken from this
result; and vice versa – the higher the entropy S of the information source,
the more the information can be obtained from it). The information by
Shannon is another kind of information construed as the probability to ob-
tain reliable information via some communication channel allowing for in-
evitable interference. The fourth kind of information is the information by
Brillouin named also “structural” or bound information. This is construed as
the difference between the entropies S of a system in its current and equilib-
rium states, i.e. the “deficit” of the entropy as compared with its maximum
value in equilibrium state. It is this structural information that is associated
with the notion of informational entropy as a measure of indeterminacy of
our knowledge about the system. Note that unlike the previous definitions
characterizing the information as the function of a process (cognition, ex-
perimental data acquisition, information transfer, etc.) the structural infor-
mation stands closer to the function-of-state concept thermodynamics oper-
ates. Actually, the higher the system organization level (i.e. the further the
system stands from equilibrium), the greater the future increment of its en-
tropy as it is approaching equilibrium (i.e. the entropy deficit). In this sense

154
the structural information transfer to the system is someway or other associ-
ated with its ordering, i.e. with doing work on it. In this case the informa-
tional entropy is just an ineffectual substitute of the free energy or the useful
work concepts in thermodynamics.
Thus the application range of the term “information” is very wide –
from a measure of some system elements’ degree of order to the contents of
the signals a creature perceives from the outside world. The informational
entropy is generally associated with the information acquisition process and
is not the state parameter unlike the thermodynamic entropy. As a result, the
investigators, who examined the case more closely, have come to a conclu-
sion that those two entropies, despite their affinity, are evidently distin-
guishable, and their identification may proceed just from lack of under-
standing. Anyway, using the same term (entropy) for different values just
misleads (I.P. Bazarov, 1991).

8.3. Correction of Entropy–Dissipation and Entropy–Irreversibility

Relations

Irreversibility has long become the “stumbling block” for many of

physicists and philosophers. Some of them consider it to be a result of the
interaction of a great number of reversible elementary processes, some as-
cribe it to the irreproducibility of boundary and initial conditions, some – to
the non-commutativity of measurement procedure and impossibility because
of it to return to initial state, some – to the disturbed symmetry of physical
laws at a time sign variation, others – to the statistical nature of time, etc.
Meantime, R. Clausius and W. Thomson, the founders of thermodynamics,
construed the irreversibility just as a result the system lost its capability for
work (efficiency reduction in heat engines). This irreversibility ensues from
the dissipation of energy with its ordered forms changing to heat. Any dissi-
pative process of such kind is irreversible insofar as the “dissipated” heat
can not be completely converted into work. We will refer to such irreversi-
bility as thermodynamic one. These are just the irreversible processes the
entropy rise law applies to in thermodynamics.
The irreversibility associated with the “path branching” is another kind
of irreversibility (K. Denbigh, 1989). Since according to the TIP, when
some non-static process is running, all available in the system thermody-
namic forces of the same tensor order are overcome, then depending on
these forces’ nature (mechanical, electrical, chemical, surface, magnetic,
etc.) the energy is converted into not only heat, but also other its forms. In
this case even in the absence of the dissipation it is impossible to return to
the process start-up by just reversing the sign of one of these processes – the
reversal of the sign and value of all thermodynamic forces being overcome
in the direct process will be needed. It is generally the same difficult as me-

155
tastasis eradication in advanced cancer case. Because of this the poly-
variant systems subjected to the action of many forces can develop irre-
versibly (i.e. in one direction) never returning to exactly initial state. This is
how the Universe is developing according to the up-to-date conceptualiza-
tion.
Furthermore, the irreversibility may be a result of the limit transition to
infinite number of particles due to impossibility to reconstruct their initial
distribution. Lastly, the irreversibility may appear in an infinite-size system,
as the Universe, because a “signal” does not return to the system or returns
with a delay.
Thus the current concept of irreversibility is much wider than its inter-
pretation by Clausius and Thomson. That philosophical amplitude and
sounding was added to the irreversibility concept by M. Planck, who con-
strued it as the “impossibility to return entire nature to the state it had been
in to the moment the process started up” (M. Planck, 1935). Various aspects
of the irreversibility issue run into one in this his statement. From a scien-
tific standpoint all processes are irreversible which follow the cause-effect
chain since effect can not engender cause. From a statistic-mechanical
standpoint all processes are irreversible which increase the state probability.
From the positions of the theory of information all processes are irreversible
which involve reducing the determinacy of our knowledge of a system state.
In thermodynamics all processes are irreversible which lead to the conver-
sion of ordered forms of energy into heat. The scope of these concepts is
very ample. Therefore it should be distinguished a thermodynamic irre-
versibility due to the thermodynamic entropy rise, a statistical irreversibility
due to the statistical entropy rise, and an informational entropy due to the
informational entropy rise. This is a medley of these aspects of irreversibil-
ity and associated concepts of entropy, where the errors are rooted, which
have led to the absolutization of the entropy rise law and to the conclusion
of the inevitable “heat death of the Universe”.
So, the interpretation of entropy as antipode of “organization”, “order”,
“complexity”, etc., has not only distorted the true relation of this concept
with the irreversibility and dissipation, but also led to an incredible mish-
mash of notions. This mishmash has caused a number of paralogisms in
thermodynamics. Some of them are considered in this book.

8.4. Inconsistence of Entropy Rise Principle

That was not done, however. Furthermore, in the TIP the specific en-
tropy was as well considered according to the local equilibrium hypothesis
as a function of the same number of the specific thermostatic variables u, υk,
ck, т.е. s = s(u, υk, ck). Meantime, the Universe as a system possesses the in-
comparably more degrees of freedom including that gravitational, radiant,

156
chemical, intra-atomic, intra-nuclear, etc. So the under-determinacy of the
Universe as a system is evident with all consequences ensuing. The correct
proof of the entropy rise law should have been evidently rooted in the
knowledge of the entropy as a function of all independent parameters of its
state. However, such equations for the Universe in whole were unknown to
Clausius and have remained so until the present. At these conditions the ex-
trapolation of the conclusions Clausius made from the simplest thermo-
mechanical systems to the whole Universe can not be estimated other than
an evident violation of the methodological principles of thermodynamics.
Later we will see that the entropy being substituted for more representative
functions of state of non-equilibrium systems not only eliminates the con-
clusion of the inevitable heat death of the Universe, but also allows finding
criteria more adequate to the character of the processes observed in it.
Furthermore, as shown in Chapter 6, the internal wells are in principle
attributed to not only entropy, but also to many purely “thermostatic” vari-
ables (in particular, the masses Mk of the kth substances, which can appear or
disappear in chemical reactions, the volume V, which can spontaneously in-
crease when a system expanding to an underpressure zone (and according to
the special theory of relativity – at also relativistic acceleration of the sys-
tem). They are also attributed to the parameters characterizing material
structural defects, e.g. the number of dislocations, the size of crystal grains
and nuclei of a new phase, etc., as well as to all relaxation process coordi-
nates Zi without exception. Therefore, from an up-to-date standpoint,
Boltzmann did not have any reasons to consider the entropy proper as a
state probability measure. In fact, according to (5.3.3) the entropy rises in
only dissipative processes and, besides, in only those of them, where or-
dered forms of energy are converted into heat energy. In other words, the
dissipation heat release and the associated entropy rise are just a part of the
state variations associated with the energy dissipation and, more impor-
tantly, with irreversibility. This circumstance once again shows off the con-
clusion that the entropy rise can not serve as a measure of not only “ any
and all” irreversibility, but either the “dissipation at all”.
It is impossible within the framework of this book to adduce all argu-
ments to prove that the absolutization of the entropy rise law can not be jus-
tified by considerations of the physical character. Therefore let us just refer
to a conclusion one of the famous investigators of the thermodynamic prin-
ciples, K. Putilov (1971), made, “In classic and the last investigations on
thermodynamics we do not find a statistics-independent, perfectly rigorous
proof of the thermodynamic inequalities (except, may be, the reasoning
Planck developed 1)). As for the inequalities just thermodynamically derived

1)
This case pertains to the process of thermal relaxation of two subsystems with different
temperatures, where ynergy of a thermally heterogeneous system changes to anergy in strict
compliance with energodynamics.

157
from unattainability of the perpetual motion of the second kind or from
other sweeping enough statements of the second law of thermodynamics,
those appeared to have often been so non-rigorous that many authors were
disposed to see an incurable logical drawback in this part of thermodynam-
ics. This explains why a number of solid manuals deny the possibility to
prove the entropy rise theorem as purely thermodynamic and statistics-
independent”.
According to the above-mentioned it was the statistical-informational
interpretation of entropy as antipode of “organization”, “order”, “complex-
ity”, etc. that made it a measure of “any and all” irreversibility. Therefore
the question is quite natural what purely thermodynamic arguments substan-
tiate the so sweeping generalizations of the entropy rise law. As we are now
starting to guess, there are no and, in substance, can not be general proofs of
this law. This is especially evident from the positions of thermokinetics, ac-
cording to which entropy varies only with the variation of the “bound” (or
internal heat) energy of a system. It does not matter in this case why this en-
ergy varies – either due to the system heat exchange or because other or-
dered forms of energy spontaneously are converted into it. Dissipative func-
tion expression (5.3.3) in the theory of irreversible processes clearly reveals
the failure of attempts to prove the thermodynamic entropy rise in irreversi-
ble processes having nothing to do with appearance of internal heat sources
(wells). In fact, any dissipative processes are irreversible. However, by no
means any irreversible process is dissipative. In particular, mixing of iso-
topes or isomers of the same-kind atoms at the same temperature and pres-
sure is an irreversible process. However, this process does not involve heat
release and loss of capability to do work since the system of non-interacting
gases having just two degrees of freedom yet before mixing was in com-
plete (thermal and mechanical) equilibrium. For interacting gases, as shown
in the previous chapter, the mixing process involves the “mixing heat” re-
lease and appearance of internal heat wells. Such wells appear in all proc-
esses associated with converting ordered forms of energy into anergy. This
is what happens, e.g., at mixing of gases having different temperatures and
thus – a certain inergy.
As for the reasoning based on the “matter-of-course” affirmation that
the efficiency of any irreversible machine is less that the efficiency of a re-
versible one, it is invalid for even heat engines. Just imagine a heat engine,
for which the heat source are tanks containing various isotopes, isomers or
isobars of the same ideal gas at the same temperature. At their isothermal
mixing the temperature of the heat source and subsequently the efficiency of
the heat engine will not change, though the mixing process itself was irre-
versible in the system heat source–working medium–heat receiver. The effi-
ciency will not change either at the isothermal throttling of the gas being the

158
heat source, though this process is irreversible as well. These examples once
again confirm that the entropy internal wells appear in only the cases when
the dissipative work (against the dissipative forces) WD = WR – WIR in-
volves releasing the dissipation heat Qd. Should this conversion be absent,
the entropy wells will not appear. However, the equality Wd = Qd can also
be violated as it was in the said processes of metal cutting or material crush-
ing. If so, the change of a part of inergy during dissipation to the strain, sur-
face and other non-heat forms of energy may be revealed experimentally by,
e.g., the difference between the isothermal dissolution heats of the initial
and the strained materials2). There are also other processes, where some
forms of the ordered energy are spontaneously converted into other similar
forms at a negligible dissipation. The entropy rise does not reflect at all the
nature of such processes. Therefore the remark is absolutely true that the
question regarding the physical grounds of the monotonic entropy rise re-
mains … still open” (L. Landau, E. Livshits, 1973).

8.5. Inadmissibility of Entropy Rise Principle as Extrapolated

to the Universe

The Boltzmanns’ probabilistic interpretation of the second law of ther-

modynamic first looked like a lifebuoy among the attempts to impose an in-
evitable destiny on the Universe. According to the Boltzmann’s interpreta-
tion of the second law of thermodynamics, the tendency of the Universe to-
ward thermal equilibrium reflects just the most probable, though not at all
unconditional direction of its evolution. There are large-scale fluctuations
possible in various zones of the infinite Universe, i.e. spontaneous devia-
tions from equilibrium involving the local decrease of entropy. In this case
the entropy of the Universe rises just in the mean, while the world we are
living in is a gigantic fluctuation.
However now, with experimental facts accumulated, such a concept of
the Universe seems more and more metaphysical. It is noted on this account
in one of the most solid study guides on thermodynamics (M.P. Vukalovich,
I.I. Novikov, 1968), “In spite of the generally progressive character of the L.
Boltzmann’s ideas, it is nevertheless necessary to indicate a known meta-
physical nature of his fluctuation hypothesis. The drawback of this hypothe-
sis lies in the fact that the probability of such a gigantic fluctuation is too
low to be realized. Its metaphysical nature can be seen from the following.
According to this hypothesis the entire development of the Universe is re-

2)
This is the way, e.g., the difference between heat capacities of a strained and unstrained
spring is revealed

159
duced to random deviations (fluctuations) from the state of thermodynamic
equilibrium, where the Universe is staying. This is not so actually. The de-
velopment of the Universe is a continuous complex process that could be
instantiated by the formation of new stellar systems… In the widest sense
the up-to-date picture of the world does not allow considering the develop-
ment of the Universe as its transition to more and more probable states”.
In this context a question arises about the propriety of extrapolation of
whatever statements of thermodynamics to the systems which state equa-
tions or, at least, the degrees of freedom are unknown. In fact, thermody-
namics gets knowledge of the properties of an object of investigation from
the state and transfer equations taken from outside as uniqueness conditions
of a kind. These equations reflect the relationship between the parameters of
various degrees of freedom of a system under investigation. In the days of
Clausius and Thomson, when the simplest thermo-mechanical systems, such
as ideal or real gases, were considered, it was quite natural to adopt for an
evolution criterion the entropy S as a quantitative measure of heat motion
since the bound energy rise was the only consequence of approaching equi-
librium in such systems. However, in poly-variant systems, such as the Uni-
verse, which, besides heat, possesses also other forms of energy (kinetic,
gravitational, chemical, radiant, intra-atomic, intra-nuclear, etc.), the multi-
various manifestation of irreversibility reduced just to the entropy rise
looks, to say the least, fancy. Looking back one can not help being in a puz-
zle how the entropy has transformed from the argument of such a character-
istic state function as system energy into a global function of state of the
whole Universe. In fact, the energy in the thermo-mechanical systems the
Clausius’ theory referred to was considered as a function of two independ-
ent variables, viz. the coordinates of heat exchange and the cubic strain U =
U(S,V) processes. In this case the system entropy S as an inverse function
takes the form S = S(U,V) and for the isolated Universe (U,V = const)
should remain unvaried. Hence, to substantiate the entropy rise from a for-
mal mathematical standpoint, the entropy should have been considered as a
function of larger number of variables, at least a part of which could vary
spontaneously (had internal wells). Such are, in particular, the spatial het-
erogeneity parameters Zi.
According to the above-mentioned it was the statistical-informational
interpretation of entropy as antipode of “organization”, “order”, “complex-
ity”, etc. that made it a measure of “any and all” irreversibility. Therefore
the question is quite natural what purely thermodynamic arguments substan-
tiate the so sweeping generalizations of the entropy rise law. As we are now
starting to guess, there are no and, in substance, can not be general proofs of
this law. This is especially evident from the positions of thermokinetics, ac-
cording to which entropy varies only with the variation of the “bound” (or
internal heat) energy of a system. It does not matter in this case why this en-
ergy varies – either due to the system heat exchange or because other or-

160
dered forms of energy spontaneously are converted into it. Dissipative func-
tion expression (5.3.3) in the theory of irreversible processes clearly reveals
the failure of attempts to prove the thermodynamic entropy rise in irreversi-
ble processes having nothing to do with appearance of internal heat sources
(wells). In fact, any dissipative processes are irreversible. However, by no
means any irreversible process is dissipative. In particular, mixing of iso-
topes or isomers of the same-kind atoms at the same temperature and pres-
sure is an irreversible process. However, this process does not involve heat
release and loss of capability to do work since the system of non-interacting
gases having just two degrees of freedom yet before mixing was in com-
plete (thermal and mechanical) equilibrium. For interacting gases, as shown
in the previous chapter, the mixing process involves the “mixing heat” re-
lease and appearance of internal heat wells. Such wells appear in all proc-
esses associated with converting ordered forms of energy into anergy. This
is what happens, e.g., at mixing of gases having different temperatures and
thus – a certain inergy.
As for the reasoning based on the “matter-of-course” affirmation that
the efficiency of any irreversible machine is less that the efficiency of a re-
versible one, it is invalid for even heat engines. Just imagine a heat engine,
for which the heat source are tanks containing various isotopes, isomers or
isobars of the same ideal gas at the same temperature. At their isothermal
mixing the temperature of the heat source and subsequently the efficiency of
the heat engine will not change, though the mixing process itself was irre-
versible in the system heat source–working medium–heat receiver. The effi-
ciency will not change either at the isothermal throttling of the gas being the
heat source, though this process is irreversible as well. These examples once
again confirm that the entropy internal wells appear in only the cases when
the dissipative work (against the dissipative forces) WD = WR – WIR in-
volves releasing the dissipation heat Qd. Should this conversion be absent,
the entropy wells will not appear. However, the equality Wd = Qd can also
be violated as it was in the said processes of metal cutting or material crush-
ing. If so, the change of a part of inergy during dissipation to the strain, sur-
face and other non-heat forms of energy may be revealed experimentally by,
e.g., the difference between the isothermal dissolution heats of the initial
and the strained materials2). There are also other processes, where some
forms of the ordered energy are spontaneously converted into other similar
forms at a negligible dissipation. The entropy rise does not reflect at all the
nature of such processes. Therefore the remark is absolutely true that the

2)
This is the way, e.g., the difference between heat capacities of a strained and unstrained
spring is revealed

161
question regarding the physical grounds of the monotonic entropy rise re-
mains … still open” (L. Landau, E. Livshits, 1973).

8.6. Processes in the Universe as Reciprocal Antipodes

According to the deductive methodology (from the general to the spe-

cial) the Universe should be construed as an entire set of interacting (rela-
tively moving) bodies and particles. Such a system is by definition isolated
and closed, i.e. none of external forces act on it, and all processes are
therein spontaneous. Strictly speaking, these are systems of such a kind the
laws of conservation of energy, mass, charge, momentum and angular mo-
mentum are stated to apply to in the form excluding exchange between
these systems and the environment. All forces acting in a closed system are
internal (having not a resultant) and, due to equality between action and
counteraction, appear and disappear in pairs only. Let us now introduce, by
analogy with gas pressure and for convenience, the specific internal force of
interaction between the body masses composing the Universe f (N/m2), i.e.
the force per unit closed surface area f (m2). Then the resultant F of these
forces will be defined as F = ∫ f·df, where df – vector element of system
closed surface. For a closed system by definition this resultant F = 0, there-
fore according to Gauss theorem:

F = ∫ f·n df = ∫divf dV = 0. (8.6.1)

From (8.6.1) the resultant being equal to zero at f ≠ 0 is possible in two

cases: (1) divergence of the force f is equal to zero everywhere (no gravita-
tion) and (2) divergence of the force f has opposite sign for various bulk
elements of a system under consideration. Let us assume, for example, divf
< 0 in one of the bulk elements. Due to equality between action and coun-
teraction this means that any couple of forces acting on the bulk element is
directed toward each other causing its compression (approaching particles).
This is what is interpreted as the attraction between particles. Then accord-
ing to (8.6.2), in another domain of the Universe, divf > 0, i.e. the forces
predominate, which cause particle-to-particle distance to grow. This evi-
dently does not mean that the gravitational forces are replaced in such an
element by the repulsion forces (anti-gravitation). It is just the sign of the
stress caused by particle-to-particle gravitational interaction forces that has
changed, i.e. the compression stress has been replaced by the tensile stress
in this domain of the Universe. Thus processes having opposite directivity
are inevitable in the spatially heterogeneous Universe. In other words, the
Universe is unstable in terms of internal processes continuously running
there and leading to mass concentration in certain domains of the Universe,
stellar cluster formation in these domains, accretion, collapse, thermonu-

162
clear reactions initiated, “supernovas” produced, explosion followed by
dissipation of the matter and its accumulation in other domains of the Uni-
verse with the same revolution of the events recurring therein. Recognition
of this dichotomy gives a new evolution idea to the Universe principally
differing from the Universe models accepted to date.
The extensive and intensive parameters of spatial heterogeneity intro-
duced in Chapter 1,2 and pertaining to the Universe as the moments of dis-
tribution of all its inherent extensive properties Zi and associated forces Хi
are of fundamental importance to comprehend the peculiarities the evolution
of the Universe features. These introduced made it possible to distinguish
within the internal energy of a nonequilibrium system an ordered (converti-
ble) part – the system inergy Е = ΣiXi·Zi. This component generalizes the
notion of Gibbs and Helmholtz free energy to open and spatially heteroge-
neous media. The ordered energy is expressed here directly in terms of the
spatial heterogeneity parameters, which allows its finding for not only a sys-
tem in whole, but also separately for each form of energy and degree of
freedom pertaining to the system. This provides investigators with rather
“physic”, flexible, non-entropy criteria of the Universe evolution, which are
much more informative than the entropy. Unlike the entropy capable to re-
flect just the approach of the Universe as a whole to equilibrium, the or-
dered energy can be found for each of the Universe-inherent degrees of
freedom or space domain dV, which makes it possible to reflect both the ap-
proach of this domain or degree of freedom to equilibrium

dЕi = –Хi·dZi < 0, (8.6.2)

and their moving away from equilibrium

dЕi = –Хi·dZi > 0 . (8.6.3)

The availability of such criteria immediately reveals the naturally exist-

ing tendency to state ordering, which manifests itself in the fact that sponta-
neous ordering processes in some domains of a nonequilibrium polyvariant
system at any hierarchic level of the Universe run due to work done on them
from other its domains or degrees of freedom, as well as due to decrease of
their ordered energy. Such processes running in the mutually opposite direc-
tions are termed as coupled in nonequilibrium thermodynamics. As already
noted, such anti-dissipative processes in the Universe involve redistribution
of mass over its volume (including accretion processes), which results in
collapse of associated cosmic objects, thermonuclear reactions initiated
therein, their temperature rise, “supernova” outburst and explosion, stellar
material scattering and subsequent “recycling” of the events in other do-
mains of the Universe. Neither of the evolution criteria based on the entropy
rise principle can predict the irreversible (one-sided) processes of such a

163
kind existing, since the entropy or its “deficiency” (deviation from maxi-
mum) is just a reflection of how far a system in whole stands from equilib-
rium.
Such an approach makes it possible to state there are anti-relaxation
processes running in the Universe (in the direction away from equilibrium)
along with relaxation processes, whereupon the Universe develops omitting
the equilibrium state. The direction of the processes running in separate
domains of the Universe is determined by the resultant of all internal forces
acting there, each of the component forces being defined as a derivative of
the ordered energy of a particular kind with respect to the adequate moment
of distribution. This means that the relationship between the forces of dif-
ferent kinds with the resultant of the same value may be quite different,
which is what predetermines the process path in the space of the same vari-
ables that define the state of the system. This circumstance that K. Denbigh
called the “path branching” is caused by many factors including the process
sequence and the rate of each particular process and its reaction force. Proc-
esses of such a kind, even if dissipativeless, are irreversible at least because
the sign reversal for any of the acting forces provokes the process to run
along now another path. Therefore with this force reversed the inverse proc-
ess runs through other states, while the system does not return to its initial
state. This is what we mean when mention the process irreversibility in any
part of the Universe.
For this reason the investigators’ striving for using the entropic criteria
of evolution in analysis of self-organization processes is a certainly good-
for-nothing attempt. It is difficult to find a parameter to estimate the system
order degree, which would be so much lame as entropy. This becomes even
more evident when comparing the entropic criteria with the ordered energy
(inergy) Е as a measure of order of the system as a whole and each of its
degrees of freedom separately Еi. These are functions of the parameters Zr
and Zi not depending on the entropy S, which makes them a new and an ex-
tremely flexible tool of the thermokinetic analysis. First of all, it is worth
noticing their simple and clear physical meaning as values characterizing
the capability of a system to do useful (ordered) work (internal and external,
respectively). It is extremely significant in this respect that the inergy of any
degree of freedom Еi can be substantially found for a current state of a non-
equilibrium system by known fields of temperatures, pressures, concentra-
tions, etc, which significantly facilitates the prediction of their behavior in
various processes without whatever calculations.
These criteria allow the same easy formulation of the process spontane-
ity criteria. According to the axiom of equilibrium self-non-disturbance the
processes of system deviation from equilibrium can not be spontaneous.
This means that any spontaneous process is always directed toward estab-
lishing equilibrium of a particular kind. This evident statement allows dis-
tinguishing natural processes from the “induced” (forced) ones. Such are, in

164
particular, the so-called “coupled” chemical reactions, when some of them
running toward equilibrium engender reactions of the opposite nature. This
is quite significant in order to perceive the reasons of oscillation process
generation and the character of state cyclic variations for systems under in-
vestigation including the “circulation” of substance and energy in nature (E.
Lightfoot, 1977).
An unconditional advantage of inergy over entropy consists also in the
utter determinacy of inergy magnitude, whereas only entropy variations
when approaching equilibrium can be substantially found. Knowing the
inergy magnitude allows introducing a specific criterion of local degree of
order for the system as the relation between its inergy E and energy U:

ηε ≡ E/U (8.6.4)

Thus the inergy expressed exclusively in terms of the non-equilibrium

parameters provides investigators with a very convenient, clear and univer-
sal criterion of evolution, equilibrium and stability of poly-variant systems,
which leaves the entropy potentialities far behind.
It is significant that criteria (8.6.2)…( 8.6.3) do not impose a line of be-
havior on the system since the sign of the dЕi variation is defined in ther-
mokinetics by exclusively experimental method as one of the system prop-
erties the uniqueness conditions of a problem are based on. In particular,
this criteria allow for the system development omitting equilibrium. In par-
ticular, thermokinetics well allows the development of the Universe with
non-periodic continuous-wave mode of its particular zones, which is quite
natural since the reaction forces (responses) from other zones lag given the
infinity of the Universe.
The absolute definition of magnitude of inergy is its unconditional ad-
vantage over entropy which can be substantially defined as only its varia-
tions when approaching equilibrium. The availability of such criteria cardi-
nally changes the results of such analysis. The application of entropy as a
tool to analyze the evolution and self-organization problems can be now ex-
plained by solely investigators’ ignorance or by their reluctance to bring
thermodynamics to conformity with the scope of the problems to be solved.
The consideration initiated here brings us to comprehension that it is
presently sensible to mention the evolution of not the Universe as a whole,
but rather of its particular domain, for which observational astronomy al-
lows to formulate uniqueness conditions. Only then it becomes clear what
processes predominate in this particular part of the Universe – whether en-
ergy degradation (ΣidЕi< 0) or energy concentration (ΣjdЕj > 0) – and what
processes of energy transmutation cause a character of energy evolution. No
theory can “dictate” a particular evolution scenario to the Universe. Theory
just confirms that the isolated Universe develops in whole omitting the equi-
librium state.

165
 Chapter 9
NEGATIVE ABSOLUTE TEMPERATURE CONCEPT INVALIDITY

The concept of negative spin temperature was originally introduced into

theoretical physics as some elegant notion that allowed “laying a bridge” in-
between nuclear magnetism and thermodynamics. Expediency of such a no-
tion was estimated thru the experimental data interpretability. However, it
was soon discovered that such an extrapolation of the thermodynamic no-
tions and methods led to a conflict with the law of excluded perpetual mo-
tion of the second kind and to a necessity of its “inversion”. This chapter
aims at demonstrating the fallacy of such notions and outlining ways of
eliminating the methodological errors that have led to those.

9.1. Negative Absolute Temperature Notion Origin

The notion of negative absolute temperature was introduced in the late

XX century after subsystems had been discovered, wherein thru the mag-
netic field inversion or high-frequency pulse the energy-level “population
inversion” had been successfully created. This is construed as a state
wherein the prevailing number of energy carriers (nuclear spins, in this
case) are, unlike the usual state, on a higher energy level (I. Bazarov, 1976).
Some of these subsystems are rather self-sufficient in the effect that the
“particles” constituting the system of nuclear spins quite quickly come to
equilibrium between themselves and, on the contrary, slowly enough – with
the remainder part of the system (crystal lattice). Investigations have shown
that the subsystems of such kind may be distinguished in the composition of
many bodies, in particular, in a number of crystals. Such is, e.g., the nuclear
magnet system of lithium ions in the lithium fluoride crystals (LiF).
The notion of temperature was then applied to describe the population-
inversion spin systems wherein temperature had to be endued with negative
value. The reason for that was as well the said statistical interpretation of
entropy and absolute temperature. If the statistical state probability is ac-
cepted identical to the thermodynamic entropy on the ground that both these
values are additive and reach a maximum in the equilibrium state (Boltz-
mann law), then comparing the derivative (∂Us/∂Ss) for the statistically de-
fined internal energy Us and the entropy Ss with the well-known definition
of the thermodynamic temperature

T ≡ (∂U/∂S)V , (9.1.1)

gives that the distribution parameter β in the Boltzmann equation for the
population ni of some ith energy level εi
166
 ni = n0 exp (–βεi ) (9.1.2)

is related with the absolute temperature T by a simple relationship β =

1/kbT, где kb – Boltzmann constant. From this it follows that the population
nо of the “zero” energy level εо relates to the population ni of the ith energy
level εi as

ni /n0 = exp [(ε0 – εi)/ kbT]. (9.1.3)

If T > 0, then the population of higher energy levels according to (9.1.3)

decreases by exponential curve. However, if an equilibrium state with the
population inversion is obtained, when the majority of particles are on the
upper energy level, this state should be endued with negative absolute tem-
perature T<0 statistically defined. Thus from the statistical-mechanical posi-
tions the application of the concept of temperature (both positive and nega-
tive) to spin subsystems sounds not less reasonable than, say, the notion of
electron temperature in plasma or magnetic temperature in magnetic materi-
als (N. Ramsey, 1956).
It should be noted that the existence of systems with the population in-
version is now a fixed fact. The population inversion in such installation as
lasers is created by “uploading” them with the microwave radiation generat-
ing a stationary non-equilibrium state of the system. However, the popula-
tion inversion is not yet enough to operate with negative absolute tempera-
ture. It is important the system to be in equilibrium at the population inver-
sion. This needs, firstly, the particles constituting the subsystem to attain in-
ternal equilibrium between themselves much sooner than with the environ-
ment or the remainder part of the system. Secondly, the energy spectrum of
such subsystems must be restricted from above so that attaining the popula-
tion inversion would not demand infinite energy to be supplied. The first
subsystem meeting these requirements was the above-mentioned nuclear
spin system of lithium ions in the lithium fluoride crystals (LiF). If the LiF
crystals are placed in the magnetic field, the nuclear magnets can in princi-
ple occupy four different energy levels including those in the field direction
(lower energy level), across the field and anti-parallel the field (upper en-
ergy level). Should now the external field direction be quickly changed (as
it was in the experiments made by E. Pursel and R. Pound, 1951), the nu-
clear magnets will not be able to follow this, and the major part of them will
appear in the upper energy level, i.e. the population inversion will occur.
After minor and quickly collapsing oscillations the nuclear magnets will
come to mutual equilibrium. This will occur for a time t2 much inferior to
the time t1 of attaining equilibrium between the spin subsystem and LiF
crystal lattice (5 thru 30 minutes), which allows concluding of a certain
“self-sufficiency” in the spin subsystem behavior.

167
 It was also experimentally found (E. Pursel and R. Pound, 1951) that if
the LiF crystal was withdrawn from a magnetic field and placed in the weak
geomagnetic field and then in several seconds returned to that initial, its fi-
nal magnetization would appear not much lower than the initial one. E.
Pursel and R. Pound repeated the same experiment with a subsystem which
initial magnetization was opposite to the field. In this case, after the speci-
men had been placed in the geomagnetic field and then returned to the
strong magnetic field, not only its magnetization value was recovered, but
also the magnetization orientation opposite to the field. The most amusing
here was the fact that the specimen having been in a field much weaker that
the local field of the test stand did not lead to the magnetization totally de-
stroyed. The last circumstance substantiated the two men’s conclusion of
two types of interaction having existed: a spin-spin interaction that led to an
internal equilibrium quickly reached in the system of nuclear magnets, and a
spin-lattice interaction of an unknown nature. The latter was likened to heat
exchange. The crystal lattice was taken there to be a thermostat, while the
spin system demagnetization – its cooling. With such an interpretation one
had to accept that the states with a negative absolute temperature extended
the thermodynamic temperature scale beyond the zone of Т = ∞ and (what is
much more important!) to draw a conclusion regarding an “inversion” of the
law of excluded perpetual motion of the second kind in such systems (N.
Ramsey, 1956).

9.2. “Inversion” of the Second Law in Negative Absolute

Temperature Range

The said “inversion” means the possibility of complete conversion of

heat into work in such systems and, on the contrary, the impossibility of
complete conversion of work into heat (M. Zemansky, 1968; I. Bazarov,
1991). As a matter of fact, let us imagine the Carnot cycle realized at the
temperatures of hot and cold bodies Th and Tc, respectively, less than abso-
lute zero. The thermal efficiency of the reverse Carnot engine will then have
the form:

ηtК = 1 – Qc/Qh = 1 – Тc /Тh . (9.2.1)

By Ramsey, a body is considered as hot in the range of T < 0 providing

its temperature is higher (i.e. negative temperature modulus is lower). At
that Tc/Th > 1, i.e. the thermal efficiency will be negative, while the modulus
│Qc│ >│Qh│. This surprising result, by Ramsey’s correct remark, means
that the Carnot cycle work done within this temperature range will be posi-
tive providing heat Qc is taken from the “cold” body (well), whereas the
hotter body is a heat receiver (sink). In this case the cycle will as well run

168
clockwise (since dS = đQc /Тc < 0 at Тc < 0). According to the law of conser-
vation of energy to do positive work, the heat amount │Qc│ must exceed
│Qh│. Since thru a heat contact between the well and the sink the entire
heat Qh given to the “hot” body may be returned thru heat transfer to the
“cold” body, the work will be done in this continuous cycle sequence due to
the heat of the solely “cold” body without any residual changes in other
bodies. Along with this Ramsey also drew a conclusion that heat could not
completely convert into work. Thus all main provisions of the law of ex-
cluded perpetual motion of the second kind suffered “inversion”. The most
amazing thing is that the conclusion of the second law violation was
drawn…from the same second law proper! The fact that heat can com-
pletely convert into work actually means that expression (9.1.4) is inappli-
cable in the range of T < 0. In this case, however, all conclusions based on
this expression become invalid as well. Despite this “vicious circle” the
statement confirming the “inversion” of law of excluded perpetual motion
of the second kind has got into educational books (M. Vukalovich, I. No-
vikov, 1968) and is being reproduced even in the best of them (I. Bazarov,
1991). The latter, e.g., writes, “Perpetual motion of the second kind, i.e. the
device that would completely convert heat of whatever body into work
(without partial heat transfer to other bodies), is impossible; this statement
excluding the inversion for usual systems with T > 0 and admitting the in-
version for unusual systems with T< 0”. This fact promotes critical attitude
toward the experimental data interpretation and its compliance with the
methodological principles of thermokinetics. In this regard the violation of
the process distinguishability law immediately becomes evident. This viola-
tion is that the experimentally discovered specific spin-lattice relaxation
qualitatively distinguishable and irreducible to other processes has been in-
terpreted as heat exchange. This leads to “under-determinacy” of the system
since instead of specific parameters of the spin-lattice relaxation to be intro-
duced the parameters of thermal degree of freedom were used. Meantime, in
case of the spin-lattice interaction the matter of interest is not a heat ex-
change (i.e. an internal heat exchange between bodies spatially separated),
but the energy redistribution by mechanical degrees of freedom of the same
atoms in the LiF crystal lattice. The fact there is a certain connection be-
tween the heat form of motion and the spin orientation does not give ground
to attribute this form to the spin system, the more so because cooling of
condensed media down to the absolute zero does not lead to disappearance
of the intrinsic rotational moment of nuclei. The existence of this interaction
leads to non-conformance between the experimental conditions and the
temperature concept definition in thermodynamics (9.1.1). In fact, accord-
ing to (9.1.3) negative values of the thermodynamic temperature may be
reached only when the system is converted, by reversible heat exchange,
into a state featuring a higher internal energy U and a lower entropy S.
Meantime, both known ways to reach the population inversion in the nu-

169
clear spin system (the external magnetic field inversion and the radio-
frequency pulse action) do not comply with these conditions. In the first
case the external magnetic field direction changes so fast, as Pursel stressed,
that the nuclear spins have no time enough to change their orientation.
Hence, the internal state of the system (including its entropy S) remained
unvaried – just the external potential (Seemann) energy of spins in magnetic
field entering in the Humiltonian function of the system along with the en-
ergy of spin-spin interaction changed. The system internal energy U, which
by definition does not depend on the state of the system as a whole in exter-
nal fields, remained unvaried in that case. Otherwise (if U changed), the
condition would be violated that in expression (9.1.1) the coordinates of all
kinds of work should be constant, but not just of volume. This refers to also
the other way of the population inversion reachable with the help of the
high-frequency (180-degree) pulse. This action can by no means be referred
to heat exchange since it has also a directional nature and corresponds to the
adiabatic process of external work done on the system.
Another remark concerns the question to what extent it is justified to at-
tribute the entropy S to the nuclear spin system as its state coordinate. In
thermodynamics the necessary condition of entropy existing in a system is
known to be the availability of other states in the vicinity of the arbitrary
state of this system, which are unattainable from it adiabatically. This
statement known as the “adiabatic unattainability axiom” means the ac-
knowledgment of the evident fact that a thermal interaction leads to such
state variations which can not be attained by any other quasi-static way (I.
Bazarov, 1991). Meantime, as the same Ramsey’s experiments showed, the
LiF crystals being cooled down to the liquid helium temperature, produces
the same effect as the adiabatic demagnetization of the specimen. The
“adiabatic unattainability” being absent in this case excludes the “mathe-
matically most rigorous and coherent” system of the entropy existence sub-
stantiation (I. Bazarov, 1991) in application to spin systems. Thus we have
one more example of discrepancy between the thermodynamic entropy and
the statistical-mechanical entropy.
The most illustrative in this respect were quite complicated and delicate
“mix” experiments on the two opposite-polarized spin systems (7Li and
19F) of the LiF crystals (A. Abragam, W. Proctor, 1959). Those experi-
ments confirmed (to acceptable accuracy) the law of angular momentum
conservation at the spin-spin interaction and showed the mix “temperature”
defined by the expression:

T = (ΣiСi/Ti)/ΣiСi , (9.2.2)

where Ti – temperature of whatever part of the spin system; Сi – weight fac-

tor the experimenters named the “spin heat capacity”. As follows from
(9.2.2), the “spin heat capacity” Сi is referred to a value inverse to absolute

170
temperature. Thus the mixing of spin systems with different “temperatures”
does not at all obey the traditional conservation laws. On the contrary, the
weighted average in (9.2.2) refers not to a temperature, but to its inverse
value meaning in this case the nuclear magnetization Zм and featuring the
quite other degree of freedom of the spin system. This fact once again con-
firms the necessity of additional state variables to be introduced with a new
degree of freedom revealed.
Anyway, it is not without reason that a number of investigators suppose
the concept of spin temperature (positive or negative) lacks the physical
meaning of thermodynamic temperatures and often misleads (A. Abragam,
W. Proctor, 1959). Therefore the above interpretation of the said experimen-
tal results looks like rather a “some statement of laws of spin-temperature
game” (M. Goltzman, 1972).

9.3. Non-Thermal Nature of Spin-Spin Interaction

As already noted above, the same experiments considered from the posi-
tions of thermokinetics drive to a conclusion the spin-spin and spin-lattice
interactions experimentally revealed are irreducible to heat exchange, but
belong to a specific interaction class named orientation in Chapter 2. This
interaction is caused by an orientation component of the system energy Е =
Е(φi), i.e. by its part dependable on relative position of the bodies. Different
positions of a body in space and its different spatial orientations are known
to be non-equivalent from a mechanical standpoint (L. Landau, E. Livshits,
1973). Unfortunately, the orientation component of the system energy Е =
Е(φi) has been hitherto studied with insufficient care. This may be attributed
to the fact that to solve many application problems, the laws of body motion
have been more conveniently reduced to the laws of motion of particular
material points which spatial orientation did not matter anymore. In that
case it was enough to consider the so-called central fields which potential
energy U(R) depended on only the distance between the bodies (on the ra-
dius vector of center of their inertia R). However, for rotating bodies and
those with non-spherical symmetry the energy is defined by not only their
position, but also orientation, i.e. U = U(Ri,φi). This relates to the potential
energy of particle interaction, which depends on the relative orientation of
particle spins. This is the energy efficiency of a state with a certain relative
spin orientation that predefines the nature of a number of chemical trans-
formations (in particular, ortho- or parahydrogen formation) and explains
ferromagnetism and anti-ferromagnetism). So in molecules with the cova-
lent chemical bond (e.g. in the hydrogen molecules) that state is more en-
ergy-efficient where the spins of valence electrons in the bonding atoms are
anti-parallel. In ferromagnetic, on the contrary, that state has lower energy
where the spins of electrons in vacant shells of the adjacent atoms (and their

171
magnetic moments) are parallel, which causes spontaneous magnetization.
Therefore, when describing a number of macro-physical properties of sub-
stances, the processes of nuclear particle spin orientation (reorientation)
running there must be taken into consideration. In thermokinetics these
processes are described by the last sum in equation of energy conservation
(2.4.5). Their specific character is in this case that the ordered orientation of
a nuclear spin subsystem is transferred to other one (including that with op-
posite orientation) and a single “weighted average” orientation of the two
spin subsystems is set up. The specificity of such interaction is as well ac-
knowledged by quantum mechanics, according to which the main part in the
spin-spin equilibrium set-up belongs to some specific interaction attributed
to the so-called exchanged forces. These forces become perceptible only
when the mean particle-to-particle distance becomes comparable with the de
Broglie wave-length. Their effect is manifested even in the case when the
direct power (electric, magnetic) interaction between particles may be ne-
glected. The nature of these forces is not yet discovered, however, this is not
that important from the positions of thermokinetics since for their definition
the available relationship U = U(Ri,φi) is sufficient. Now let us clarify how
the spin systems will be described with the orientation degree of freedom
taken into consideration.

9.4. Description of Spin Systems from the Positions

of Thermokinetics

Nuclear particles, when rotating, are known to undergo precession, i.e. a

motion when the axes of their rotation make a spatially oriented angle φs
with the external field vector H. From a thermodynamic standpoint this
means one more degree of freedom to be allowed for in the thermodynamic
equation. This degree of freedom depends on not only the total intrinsic an-
gular momentum of the nuclei in the substances under investigation Js, but
also on its spatial orientation (angle φs). Depending on this angle, the spin
projection Is of the ith elementary particles on a chosen spatial direction (e.g.
on the external magnetic field H direction) varies from – Is до + Is (which
corresponds to φs= 1800 and φs= 00). Hence the total angular momentum Js
is related to the spin Is thru the relationship:

Lс = Σ i hIс cos φс , (9.4.1 )

where h – Planck constant.

Along with this, nuclei, atoms and molecules of condensed substances
are known to possess some magnetic moment Zм primarily caused by the
orbital motion of electrons around nucleus and by their spins. Thus the in-
ternal energy E of condensed substances generally depends on not only their

172
temperature (or entropy S) and magnetic moment Zм, but on also the rela-
tive orientation of spins (angle φs). The parameters S, Zм and φs are in prin-
ciple independent so that the energy E of condensed substances as the func-
tion of their state looks like U = U(S, Zм, φs), while its exact differential is
expressed by the relationship:

dU ≡ TdS – НdZм – Мс⋅dφс , (9.4.2)

where T ≡ (∂U/∂S) – absolute temperature of the system; Н ≡ (∂U/∂Zм) –

external magnetic field intensity; Мs ≡ (∂U/∂J) – orientation moment. In
this expression the term Мs⋅dφs features the work associated with the orien-
tation polarization of the nuclear spin system (as the term НdM features the
work associated with the system magnetization). From this, based on the
exact differential properties, additional differential relationships follow in
the form:

(∂Zм/∂φс)Н = (∂Zм/∂Н)φ. (9.4.3)

According to this relationship the angle of orientation polarization φs of

a nuclear spin system varies under the influence of the external magnetic
field H inasmuch as the system magnetization Zм – due to the spin system
reorientation (angular momentum Js variation). However, the magnetic field
H being constant, the Zм variation can be caused by just the variation of nu-
clear magnetization (dipole magnetic moment of nuclei) Zn that is related to
the total angular momentum Js thru the so-called gyro-magnetic ratio γs:

Zя = γLс , (9.4.4)

Then (9.4.2) may be substituted for:

(∂φс/∂Н)L = γс . (9.4.5)

Thus the orientation polarization of a nuclear spin system under the in-
fluence of external magnetic field really takes place and is expressed by the
said gyro-magnetic ratio γs. The fact that both the left-hand side and the
right-hand side of equation (9.4.3) differ from zero validates the idea of al-
lowing for an additional degree of freedom associated with the spin system
orientation. Thus the acknowledgement the orientation processes and the as-
sociated additional degree of freedom of condensed substances actually ex-
ist results in compliance with experiment. Thereby thermokinetics elimi-
nates the paralogism of negative absolute temperatures confirming once
again the inviolability of the laws of thermodynamics.

173
 Chapter 10

PARALOGISMS OF RELATIVISTIC THERMODYNAMICS

In the years passed after the fundamental work of A. Einstein had ap-
peared (1905) with a formulation of the special theory of relativity (STR)
physicists endeavored to impart such a form to the classic laws that would
be invariant in all inertial frames of reference. M. Planck was the first who
declared that in thermodynamics (1907), while A. Einstein approved his
transformations. However, then an event occurred rare for physics – the
Planck’s transformations were discovered to have led to an absurd result
half a century later. The subsequent lively discussions not only failed in
reaching an univocal result, but showed up such a discord in definitions and
interpretations of the basic concepts of energy, heat and work that a neces-
sity appeared to completely shuffle up the fundamentals of thermodynam-
ics. That circumstance was in that time one of the incentives to develop
thermokinetics. This chapter is dedicated to the situation arisen, which will
be analyzed from the positions of thermokinetics.

10.1. Ambiguity of Relativistic Transformations of Thermodynamic

Values

According to the Lorentz – Poincaré – Einstein general theory of rela-

tivity all physical laws should be written in a form invariant with respect to
any inertial frame of reference. M. Planck first revised the laws of thermo-
dynamics from that standpoint (1907). He proceeded from the assumption
that the first and the second laws of thermodynamics must remain as well
valid for transformed values in an arbitrary frame of reference. Based on the
expressions known in mechanics for transformation of the energy U and ac-
celeration work đWw he came to a conclusion that the heat Q and absolute
temperature T should be transformed in accordance with the expression:

Q΄ = Qγ ; Т΄ = Тγ , (10.1.1)

where Q′, Т′ – heat and temperature in a reference frame moving relative to

an observer with a velocity of w; γ = (1 – w2/c2)½ -– Lorentz factor, с – ve-
locity of light in vacuum.
As for the entropy S, it must, by Planck, remain lorentz-invariant since
the uniform acceleration of all system parts refers to adiabatic processes.
Those relationships were doubtless for all until 1963 when H. Ott discov-
ered the absurdity of that result from a thermodynamic standpoint. In fact,
accelerating a heat source (well) with a temperature of Th up to a velocity of
174
w, using its heat Q′h in the relativistic Carnot engine (with a fast moving
heat well) and then decelerating it again to a velocity of w = 0, the result of
the said operations must exactly coincide with the work done by the classic
Carnot engine. This does not occur, though. To accelerate a heat well with
an own weight of M, the work Ww′ to be done corresponding to its kinetic
energy increment Еk = М′ - М, which, given the Einstein’s relativistic for-
mula М′ = М/γ, is equal to Ww′ = Мс2(1/γ – 1). After rejecting the heat Q′h
from the heat well its own weight will vary by the value ΔМ = Q′h/с2 and
when decelerating the heat well the work Ww˝ = (Мс2 - Q′h)(1/γ – 1) will be
restored. From this it follows that to accelerate the heat well and to deceler-
ate it, the work will be done as Ww′ – Ww" = Q′h(1/γ – 1). Subtracting this
work from the work Wc′ gives:

Wc = Qг(1 – Тх/Тг) – Qг (γ – 1/γ). (10.1.2)

This result does not correspond to the classic expression Wc = Qh(1 –

Тc/Тh), which evidences incorrectness of the Planck’s transformations.
Therefore Ott proposed other transformations for heat and work:

Q΄ = Q/γ ; Т΄ = Т/γ , (10.1.3)

which allows eliminating this discrepancy.

The difference between the Planck’s and Ott’s transformations is re-
ferred to the equivocal definition of the accelerating forces (Möller, 1970).
In fact, when deriving the relativistic equation for work, Planck used the
force as a momentum derivative:

Fw = d(Mwm/γ)/dt, (10.1.4)

which includes the system weight (mass) under the derivative sign. How-
ever, H Otts applied the following expression to the force:

Fw = Мd(wm/γ)/dt . (10.1.5)

The distinction between those two are that expression (10.1.4) contains
an additional “Planck’s force” appearing in calculations, which is necessary
to maintain the constant velocity of a body being heated and the associated
increment in the rest mass. This force has a number of unusual properties
(does not obey the conventional force transformation rules) and lacks dis-
tinct physical meaning. However, thanks to this force the law of action and
reaction equality coincides with the law of conservation of momentum. Be-
sides, the interpretation of electromagnetic phenomena is facilitated. The
distinction between expressions (10.1.4) and (10.1.5) naturally affects also

175
the relativistic transformation of heat which, failing an independent defini-
tion, is usually defined “by inverse balance” (i.e. as “what is not work”).
The H. Ott’s article passed unheeded when alive. However, H. Arzelies
came to the same conclusion (1966) apart from Ott. Unlike Ott, he treated as
wrong also the energy and momentum transformation formulae ensuing
from relativistic mechanics of elastic bodies. That time the work was
heeded and an avalanche of publication followed having led to a lively dis-
cussion at the international symposiums in Brussels (1968) and Pittsburg
(1969). Those discussions revealed indiscrete chaos in the definitions of
fundamental notions and concepts of thermodynamics, which gave H. Arze-
lies grounds to announce the current crisis of thermodynamics. The point is
not just in the absence of unity between the relativistic transformations of
energy, heat and work, but rather in investigators’ reluctance to apply to the
fundamentals of thermodynamics each time the necessity arises to general-
ize its methods to a more general class of systems. This shows, in particular,
in the works appeared, which authors endeavor to “reconcile” different
transformations. They propose “combined” expressions for the heat Q′ vol-
untarily going over into either (10.1.1) or (10.1.6). Some go so far as to de-
clare that the application of each of the formulae depends on the spatial po-
sition of thermometer (I. Bazarov, 1983). Meantime, this is a consequential
decision since the available contradictions directly threaten the status of
thermodynamics as a theory most logically consistent and mathematically
rigorous.

10.2. Non-Invariance of Expression for Efficiency of Relativistic

Carnot Cycle

The discordances arisen in relativistic thermodynamics touch not only

the methodological aspects of thermodynamics, but also the fundamental
consequences of the heat engine theory and the law of excluded perpetual
motion of the second kind. Following the M. Planck’s reasoning the relativ-
istic Carnot engine may be imagined as a cylinder with a gas under the pis-
ton. Its operation differs from that of the classic Carnot engine in just the
acceleration of working medium after its adiabatic compression and in the
deceleration after receiving the heat Q′h from a moving heat well. Let us
choose such a frame of reference where the heat sink (receiver) is at rest.
Let a gas-containing cylinder first be moving along with the heat well re-
ceiving from it a heat of Q′h at a temperature of Тh′. Then the gas-containing
cylinder is adiabatically decelerating to rest and the gas temperature be-
comes equal to Th. The gas in the cylinder is subsequently expanding to the
temperature Tc of the heat sink giving it some heat Qc at a temperature of Tc
and then adiabatically compressing again to the temperature Th. After that
the gas-containing cylinder is again accelerating and the cycle repeating.

176
Due to the entropy invariance evident equalities are observed for such a cy-
cle:

Q΄г /Тг΄ = Qг /Тг = Qх /Тх . (10.2.1)

The work Wc′ done in the relativistic Carnot cycle is equal to

Wц΄ = Q΄г – Qх = Qг(1 – Тх/Тг΄) = Qг(γ – Тх/Тг). (10.2.2)

Herefrom, given (10.1.1) and (10.1.2), the expression follows for the ther-
mal efficiency of the relativistic Carnot cycle (M. Planck, 1907):

ηtK ≡ Wц΄/Q΄г = 1 – Тх/Тгγ . (10.2.3)

Thus the Planck’s transformations did not allow for invariance of the
Carnot cycle efficiency ηtC which expression is one of the mathematical
formulation of the second law of thermodynamics. By Planck, the tempera-
ture of a moving heat well is always below the one measured in a stationary
frame of reference, and according to his transformations the relativistic Car-
not cycle efficiency (see 10.2.3) is always below the classic one. And what
is more, at certain γ this efficiency may appear even negative. On the con-
trary, by Ott, the temperature of a moving heat well is always higher, and
his Carnot engine has the efficiency higher than that of the classic engine:

ηtK(Отт) = 1 – Тхγ/Тг . (10.2.4)

From this it follows that even with the heat well and heat sink temperatures
being equal in the own frame of reference the Ott’s engine will do work. At
γ → 1 the efficiency of this engine will approach unit (1) irrespective of the
heat well and heat sink temperatures. The non-invariance of thermal effi-
ciency for the ideal Carnot cycle, which features one of the most fundamen-
tal laws of nature, viz. the law of excluded perpetual motion of the second
kind, looks like evident paralogism, especially granting the fact that this
conclusion has been made as proceeding from the invariance of all laws of
nature (including the first and the second laws of thermodynamics) in any
inertial frame of reference. That is why it looks reasonable to discuss this is-
sue from more general positions of thermokinetics.

10.3. Relativistic Carnot Engine as Compound Engine

The contradictions arisen can be partly settled with a more general defi-
nition of absolute efficiency of heat and non-heat engines (see 6.6.4) receiv-
ing energy from several sources of various nature:

177
 ηi = Wj /ΔUi′ = 1 – Ψi″/Ψi′.

For this let us first make sure that the relativistic Carnot engine is not
merely a heat engine. In fact, the working medium of such an engine re-
ceives, along with the heat Q′h, also the kinetic energy Еk = Q′h(1/γh – 1) re-
quired to maintain its velocity during the heat absorption and the associated
rest mass increasing ΔМ = Q′h/с2. Therefore such an engine is a combination
between a merely heat engine doing the work Wt owing to the heat well Q′h
and a mechanical engine doing the work Ww owing to the kinetic energy Еk
received from the heat well in the course of heat absorption. This work of
the heat sink generally moving will be defined by the expression:

Ww = ΔЕ к = ΔМ(с2/γh – с2/γc), (10.3.1)

where γc = (1 – vc2/c2)½; vc – relative velocity of heat sink.

Thus the efficiency of such a “compound” engine must be defined as:

η(t+w) = (Wt΄+ Ww)/(Q΄h + Е к) = ηt γh + ηw (1 – γ γh), (10.3.2)

where

ηt = Wг΄/Q′h; ηw = Ww /Е к = 1 – γh/γc (10.3.3)

– absolute efficiencies of, respectively, the Carnot merely heat engine

with the moving heat well and the “kinetic” engine doing useful work due to
deceleration of the working medium with a mass increased in the course of
heat absorption. It can be easily seen that the efficiency of such “com-
pound” engine (10.3.2) is “weighted mean” of ηt and ηw, i.e. takes an inter-
mediate value tending to ηt at γh → 1 and to ηw at γh → 0. This result meets
the correspondence principle since in the particular cases considered we
deal with the merely heat and merely mechanical engine. Thus one of the
reasons of the confusions arisen refers to the arbitrary extrapolation of the
heat engine concept to systems with fast-moving heat wells. However, the
contradictions relating to the relativistic transformation of thermodynamic
values remain unsolved.

10.4. Inapplicability of Relativity to Absolute Values

All conclusions following from relativistic mechanics relate to moving

material point or body with invariable rest energy U depending on exclu-
sively rest mass M and a known relation U(М) = Мс2 associated with it. Let
us consider how all these conclusions would change with the entire set of

178
interacting (relatively moving) material bodies considered. Such an object is
by definition an isolated and closed system with all its energy U being self-
energy (intrinsic). Intrinsic energy of a system is defined as that part of its
total energy which does not depend on position or movement of the system
relative to other bodies, but is conditioned by exclusively its internal proper-
ties. It follows from this definition itself that intrinsic energy of a system is
its Lorentz-invariant value independent on velocity the system is moving
with relative to whatever reference system. What argued in favor of the rela-
tion U′ = U/γ known from relativistic mechanics and usually taken for the
intrinsic energy relativistic conversion formula (R. Tolman, 1974)? Such a
conclusion appears to be usually drawn based on the following simple rea-
soning. Let two bodies with a mass M and self-energy U each move toward
each other with a velocity w. Then, at the moment their relative rest occurs,
their kinetic energy Еk = (М′ – М)с2 converts into self-energy (thermal at
elastic collision and potential at inelastic one). Then the inertial mass (М′ –
М) transforms into the rest mass ∆М, which results in increase in the self-
energy U(М):

U(М + ∆М) = U(М) + Мс2(1/γ –1) = U/γ . (10.4.1)

This relation characterizes, however, nothing but the dependence of

self-energy on rest mass which has changed here due to the conversion of
external energy into internal one. This not at all means that the self-energy
has become a velocity w function. In the similar way the rest mass will
change when external energy of a body transforms into its self-energy. In
other words, equation (10.4.1) expresses self-energy variation due to rest
mass variation whatever reasons would cause it. In this case any other bod-
ies’ relative velocity function different from the Lorentz formula might take
the place of the factor γ. So a logical error is present here as rooted in a sub-
stitution for argument of function or, more generally, for the philosophic
categories of essence and phenomenon.
The necessity in relativistic self-energy conversion is equally as often
argued for by body volume variation due to the Lorentz dimensional reduc-
tion in the direction of movement. At mechanical equilibrium with the envi-
ronment such a volume variation involves cubic strain work done, which al-
legedly causes the self-energy variation. However, such an “argumentation”
is inconsistent, too, since a reduction in dimensions in the direction of
movement can be easily neutralized by the adequate dimensional variation
in the transversal direction leaving the body volume V invariable. Further-
more, the said reduction in dimensions takes place in also vacuum, where
expansion work is not at all done.
The necessity in relativistic self-energy conversion is often argued for
by variation of system parameters due to system acceleration. In particular,
it seems to be rather natural that the temperature of a body decreases with

179
the velocity of the system approaching the velocity of light, because this
temperature is, like entropy, a measure for energy of chaotic motion parti-
cles are involved in, which gradually gives way to ordered motion of the
body as a whole. In the special theory of relativity (STR) transformations do
not depend on whether a system is moving relative to an observer or the ob-
server is moving relative the system. Therefore it is quite admissible to con-
sider a body as stationary (with its inherent chaotic motion), whereas the
reference system – as movable with relativistic velocity relative to the body.
In such a case the system status ordering will be nothing but apparent for
the moving observer, while using the Lorentz transformations will be aimed
at reducing measurement results to those observed in the inherent reference
system. The same problem is to be solved when the true nature of phenom-
ena observed is considered. E.g., the observer moving together with perma-
nent magnet relative to stationary conductor would, according to STR, at-
tribute the current generated therein to the action of merely magnetic forces.
On the contrary, should conductor move relative to magnet, he would con-
sider this phenomenon as nothing but electric (R. Feynman, 1977). Such
ambiguity would disappear if the investigator studied the entire set of mov-
ing bodies in inherent reference system considering it as a single whole.
Such a reference system (absolute) would make it possible to find the true
nature of acting forces.
It may be easily instantiated how far the STR conclusions differ from
thermokinetics. Assume some two domains of an isolated system with the
same mass m start relative motion. Then according to the relativistic relation
between mass and energy m΄ = m/γ and the law of mass conservation for iso-
lated system as a whole the increment of inertial mass by a value of 2(m΄ –
m) = 2m(1/γ – 1) will be provided at the expense of the equal decrement of the
system rest mass М – 2m΄ = М – 2m/γ. Equating these mass variations gives:

γ = 4m/(М + 2m). (10.4.2)

According to (10.4.2) at m > 0 it is impossible to reach the velocity v =

c (γ = 0) in whatever domain of the system. Meanwhile, from astrophysics
the total mass of photon flux moving with the velocity of light in the Uni-
verse is not at all equal to zero. This example evidences that STR based on
the Lorentz transformations is in principle inapplicable to isolated sys-
tems. The reason is clear: no domain of a system can acquire infinite mass
if provided at the expense of finite masses of other system domains. So if
considering not a material point (as in STR) as a system, but an isolated
system in whole and using the mass-on-velocity dependence ensuing from
STR, it is not so hard to discover STR nonconformance with experiment.
With such an approach it is equally as easy to make sure of inadmissi-
bility to apply relativistic transformations for also any state parameters be-
ing arguments of self-energy U as a state function for isolated system. Let

180
us consider for this purpose some isolated system, which separate domains
are moving with relativistic velocities relative to stationary center of grav-
ity. As shown above (Chapters 1 and 2), the state of such a system is char-
acterized by a certain set of variables Θi and Ri or Zi, i.e. U = U(Θi ,Ri).
Then according to the general definition of generalized potential Ψi ≡
(∂U/∂Θi) any of them stays invariable at any relative motion of the system
domains since their displacement influences the variables Ri only. Hence
neither the mean temperature of the system nor its mean pressure vary in
relative motion of domains of the isolated system. This fact confirms the
conclusion made in Chapter 3 of the necessity to measure any of potentials
in absolute scale which zero corresponds to complete degeneracy of this
kind of interaction. Apropos the necessity to find absolute scales of poten-
tials was first realized yet in fluid mechanics as applied to pressure. With
classic mechanics a necessity appeared to know absolute temperature T and
with changing to open system – also entropy S. Now we are in a position to
show that this requirement concerns not entropy only, but any parameters
Θi. To do so, it is enough to consider the generalized expression for ordered
work đWiе = Хi·dZi. Since Хi ≡ –∇ψi , dZi = ΘidRi and (dRi·∇)ψi = dψi, then
đWiе = – Θi dψi, so that the unique definition of energy conversion quantita-
tive measure requires to measure any parameter Θi in absolute scale inde-
pendent on the nature of measuring substance and values of other system
parameters. Thus the necessity to measure the generalized potentials ψi and
generalized coordinates Θi of any processes ensues from the law of energy
conservation (for energy transfer and energy conversion, respectively). This
relates to the system energy U itself, too, as the function of these variables.
This fact entails the invariance of heat Q and work W, which is quite
natural, since each of them separately features the variation of self-energy
as an invariant value. The invariance of specific self-energy, entropy and
volume results as per (10.4.1) in the invariance of thermal efficiency of
Carnot cycle.
However, thermokinetics allows advancing even further. Considering
the entire set of interacting (relatively moving) bodies as a single whole al-
lows finding an absolute scale for also velocity. Absolute velocity in such a
system is defined relative to center of inertia, which position can be changed
by no means according to the law of momentum conservation. While meas-
uring such a velocity is problematic, the fact of its availability as itself is
significant.
Being a deductive discipline, thermokinetics considers rotating systems
in the most general case. For rotational motion a primary reference system
is known to be available – this is the center-of-inertia system (L. Landau, E.
Livshits, 1973). Therefore the Poincaré-Lorentz-Einstein relativity principle
requiring the invariance of physical laws stated relative to any inertial refer-
ence system is evidently inapplicable. In this respect thermodynamics may
be considered as an “absoluteness theory”. In thermodynamics such refer-

181
ence systems are primary, wherein the laws of physical phenomena look es-
pecially simple. Such a standpoint evidently contradicts STR.
It hardly needs proving that absolute values differ from relative ones
just because they do not depend on the state of any outside bodies including
the movable status of the observer. This fact draws the final line under the
issue of relativistic transformations of thermodynamic values and supports
the idea the academician I. Tamm (1956) expressed, “none can certainly
foresee what further development of physics would be, but one thing, as it
seems to me, could be certainly affirmed – the ideas of Einstein, his analysis
of the concepts of space and time and space-time relations interacting with
the matter existing in space and time could undergo fundamental changes in
the future”.

Conclusions to Part 3

When considered from more general positions of thermokinetics, a

number of thermodynamic generalizations not supported with any correc-
tion of thermodynamic concepts and laws reveal paralogisms in each of the
applications of thermodynamic jeopardizing its status as a theory logically
immaculate and immutable in its deductions. These paralogisms are caused
by the fact that the classic thermodynamic method based on the equilibrium
and reversibility concepts has been drawn out of the strict framework of its
applicability. Additional non-equilibrium state variables having been denied
in the investigation of spatially heterogeneous media and systems with addi-
tional degrees of freedom, heavy consequences followed. Those manifested
themselves in reducing the idea of excluded perpetual motion to exclusive-
ness of hot well properties, in acknowledging the “inversion” of the law of
excluded perpetual motion in the systems with negative absolute tempera-
tures, in non-invariance of the relativistic Carnot cycle efficiency, in deny-
ing a thermodynamic solution to the problem of thermodynamic inequali-
ties, in enduing negentropy with the properties of system-degree-of-order
measure, in distorting the relation of thermodynamic entropy with dissipa-
tion and irreversibility, in extrapolating the law of entropy rise to the entire
Universe, in applying the theory of relativity to absolute values, etc. Espe-
cially heavy consequences appeared as a result of departure from the meth-
odology of thermodynamics for open systems, which showed in inadequacy
of the material equilibrium criteria, in substituting the conditions of TIP sta-
tionarity for those of material equilibrium, in loosing the free energy proper-
ties as potential, in denying the applicability of the law of excluded perpet-
ual motion to such systems, etc. Elimination of these contradictions con-
firms the inviolability of thermodynamics within the framework of its laws’
validity and opens up new vistas in further extension of the applicability of
thermodynamic methods.

182
 Part 4

FURTHER GENERALIZING OF HEAT ENGINE THEORY

Chapter 11

THEORY OF SIMILARITY OF ENERGY CONVERSION PROCESSES

Classic thermodynamics, having time excluded from its equations, did

not considered the processes of energy and its carriers transfer. Therefore
the issue regarding unity of energy transfer and conversion processes could
not arise in its depths. N. Umov was the first (1873) who paid attention to
that relation and stated the energy conservation law (2.3.6) for mechanical
processes rather as a consequence of energy transfer across the system bor-
ders than as a result of energy disappearing in some points of space and ap-
pearing in others. It remained unanswered, however, what was the “transfer
substrate” in that case – either the material energy carriers attributed to the
extensive values Θi category or the intensive values ψi as analogs of me-
chanical stresses. Thermokinetics gives an answer to this question. Accord-
ing to (2.3.10) the flux Jui of any ith form of energy is expresses as the prod-
uct of the intensive measure of motion – the generalized potential ψi – and
the energy carrier flux Ji that is defined, in its turn, as the product of the ex-
tensive measure of motion Θi of this particular kind being transferred and
the velocity vi of its transfer in space. This flux density ji is accordingly ex-
pressed as the product of the density ρi =ρΘi of this extensive value and its
transfer velocity vi. The transfer concepts extraneous to classic thermody-
namics find thereby their realization in thermokinetics
Using the thermokinetic body of mathematics applicable for the investi-
gation of processes with any irreversibility degree allows posing the issue of
establishing general behavior of non-static energy conversion processes and
thus enabling a synthesis of the theory of irreversible processes (TIP) with
engineering thermodynamics dealing with the energy conversion process in
heat engines.

183
 This chapter highlights the new results in the theory of heat and non-
heat, cyclic and non-cyclic machines, direct action machines and heat
pumps obtained when applying the thermokinetic methods to them. One
more step will thereby be done toward the creation of the unified theory of
real processes.

11.1. Unity of energy Transfer and Conversion Processes

Classic thermodynamics, having time excluded from its equations, did

not considered the processes of energy and its carriers transfer. Therefore
the issue regarding unity of energy transfer and conversion processes could
not arise in its depths. N. Umov was the first (1873) who paid attention to
that relation and stated the energy conservation law (2.3.6) for mechanical
processes rather as a consequence of energy transfer across the system bor-
ders than as a result of energy disappearing in some points of space and ap-
pearing in others. It remained unanswered, however, what was the “transfer
substrate” in that case – either the material energy carriers attributed to the
extensive values Θi category or the intensive values ψi as analogs of me-
chanical stresses. Thermokinetics gives an answer to this question. Accord-
ing to (2.3.10) the flux Jui of any ith form of energy is expresses as the prod-
uct of the intensive measure of motion – the generalized potential ψi – and
the energy carrier flux Ji that is defined, in its turn, as the product of the ex-
tensive measure of motion Θi of this particular kind being transferred and
the velocity vi of its transfer in space. This flux density ji is accordingly ex-
pressed as the product of the density ρi =ρΘi of this extensive value and its
transfer velocity vi. The transfer concepts extraneous to classic thermody-
namics find thereby their realization in thermokinetics.
Furthermore, thermokinetics allows easily distinguishing between the
energy transfer processes (within the same form) and the energy conversion
processes (from some form into another). With this purpose in mind let us
consider a system with an arbitrary volume of V wherein a steady process of
converting the ith energy form into the jth one is being realized. As shown in
Chapter 2, this process is described by the second sum in the equation of
energy conversion in the form (2.4.7), which terms characterize the capacity
of energy conversion processes in the system.
The terms of this sum may be endued with a more conventional form if
taking into consideration that the redistribution processes correspond to the
expansion of ∇⋅(ψijie) into the (2.4.6) given that ∇⋅jie = 0. For the system as
a whole this condition becomes ∫∇⋅jie dV = 0. Changing in this expression
from the space integral to the integral ∫jie⋅n df = 0 taken through the closed
surface f of the system gives that at jie ≠ 0 the said condition indicates the
parts f' and f" existing on this surface with the opposite sign of the jie⋅n

184
product, i.e. the flows ji' and ji" existing as entering the system and leaving
it, respectively. In this case

∫ji⋅n df = ∫ji΄⋅n df ΄ + ∫ji˝⋅n df ˝ = Ji˝ – Ji΄ = 0, (11.1.1)

where Ji΄ = ∫ji΄⋅n df ΄ ≤ 0 and Ji˝ = ∫ji˝⋅n df ˝ ≥ 0 – flows (flow rates) of energy
carrier at the system input and output, respectively.
Thus in balance equation (2.4.3) the external component of the flow Jiе
caused by the useful external work We done may be distinguished from the
relaxation component Jir by the conservation of the energy carrier flow rates
Ji΄ and Ji˝ at the system input and output. Representing in (2.3.7) the bulk
element dV covered by the flow jie as a product of its cross section vector
element df normal to the flow and the normal element dn (dV = df·dn) and
also considering that in a steady flow (dn·∇)ψi = dψi one may derive instead
of the addend in (2.3.7 ):

Ni = ∫(∫ jie·df)dψi = – ∫ Jiеdψi = Jiе(ψi΄ – ψi˝), (11.1.2)

where Jiе = – ∫ jie·df = Ji˝ = – | Ji΄| – external “transit” flow of energy carrier
crossing the system and remaining unvaried in value; ψi', ψi" – input and
output potentials of energy carrier, respectively.
Thus the useful component is associated with the “transit” flow of en-
ergy carrier Jiе which crosses the system with its value invariable. This
statement may be instantiated by an electric machine where the input cur-
rent is known to be equal to the output one. This is equally evident for a cy-
clic heat engine, too (Fig. 3-1), since in any cycle the entropy rise with the
heat supply ∆S1-2 is numerically equal to the entropy decline ∆S2-1 with the
heat rejected to heat sink. Therefore, at steady conditions the entropy flow
rates Jsе = ∆S/∆t at the system input and output are as well equal. This is the
circumstance that was foundational for S. Carnot to compare heat engine
with a mill-wheel (hydraulic machine).
As follows from the aforementioned, the condition necessary to provide
the conversion of some ith energy form consists in the availability within the
energy converting system an adequate energy carrier flow Jiе crossing the
areas with a various value of the generalized potential Ψi. For heat engines
this is the entropy flow Jsе from heat well with a temperature of Т′ to heat
sink with a temperature of Т″; for electric machines – the electric current Jее
associated with the electric potential differential Δϕ; for concentration ele-
ments – the substance flow Jkе crossing the borders of a system with differ-
ent values of chemical potential; for mechanical engines – the gas or liquid
flow between the zones with different values of enthalpy or pressure, etc.
This distinguishing feature of the energy conversion processes, viz. transfer
of a corresponding thermostatic value Θi in the field of the Xi forces, is

185
characteristic for micro-heterogeneous systems, too. In fact, a flow proper
of some value across the borders of a system implies there are “flow-out”
and “flow-into” zones in the system, i.e. subsystems reversing their proper-
ties in the course of energy conversion. These may be electrons and “holes”
in semiconductors, positive and negative ions in electrolytes, opposite
charges or poles of electric and magnetic dipoles, bulk elements shifting in
opposite directions in bodies under deformation, etc.
This means that any system taking part in energy conversion process
(either a source of technically fit energy, or the converting device proper, or
the work object, or whatever else) should always be considered as an ex-
tended (heterogeneous) system, which different parts reverse their state.
Such state variations are, in the broadest sense, that “compensation” for the
energy conversion process, which necessity R. Clausius advocated.
Expression (13.1.2) is valid for both the ith (being converted) and the jth
(having been converted) forms of energy. This means that any energy con-
verting device (both cyclic and non-cyclic) is a system “transfused” with the
energy-carrier flows of both “being-converted” Ji and “having-been-
converted” Jj forms of energy. Such a “two-flow” system in electrical engi-
neering is represented as a “four-terminal” network in electrical engineering
and radio-engineering, i.e. a device with two pairs of input and output ter-
minals. For clearness, an electrical “equivalent circuit” to an energy con-
verting device is shown in Fig. 11.1.

Rii΄ Ji Jj R΄jj
Energy Source

Work Object
Xiе

Xjе
Xi

Xj
Rij

Rji

Rii˝ Ji Jj Rjj˝

Converting Device

Fig.11.1. Electrical Equivalent Circuit to energy Converting Device

The energy converting device depicted on this figure (central part) looks
like being “transfused” with two different-kind flows of the energy carriers
Ji and Jj. It is emphasized thereby that this device interacts with not just
one, but two objects – the source of technically fit energy (inergy) and the
work object. The principal distinction of a four-terminal network from the
heat engine concept adopted in thermostatics as a set of heat well, working
medium and heat sink is that the energy converting device (working me-
dium analog) in the four-terminal network contacts two “extended” systems,
each being non-equilibrium and including energy carrier wells and sinks

186
(similar to the extended system). The function of the energy converting de-
vice in such a circuit is using the spatial heterogeneity in some material ob-
jects (inergy source) in order to create an artificial heterogeneity in other
ones (work object).
The functional division of the material objects participating in the en-
ergy conversion process stresses the inseparable unity of the energy transfer
and conversion processes and provides the necessary generality in posing
the issue on the thermodynamic perfection of the energy conversion process
in heat and non-heat, cyclic and non-cyclic engines.
Such an approach is extremely important since the wells and the sinks
of whatever source of energy can not be distinguished in nature with the
same easiness as the heat wells and heat sinks. It is very difficult to distin-
guish the energy wells and sinks in such media as, e.g. homogeneous
chemically reacting mixture, polarized or magnetized media, dissociated
gases, deformed elastic bodies, etc. Nevertheless, those are known to be
able to do useful work like the discontinuous systems. Sensing them as a
non-equilibrium single whole promotes the formation of a correct world
outlook and the comprehension that all man-used energy sources, without
exception, ultimately belong to only the environment and are caused by en-
vironmental spatial heterogeneity. This implies not just a possibility, but
also an inevitability of using the environmental energy. The apparent con-
tradiction of this conclusion with classic thermodynamics (its second law)
that excludes the usage of the environmental heat is caused by just the fact
that classic thermodynamics interprets the environment simplistically as
something homogeneous. It is this environment that embodies the Ostvald-
proposed concept of perpetual motion of the second kind as a heat engine
using the practically inexhaustible heat well of the world ocean. Now, when
power installations exist using the temperature difference between the sur-
face and subsurface oceanic layers, i.e. the part of this indeed inexhaustible
heat well, the narrowness of equilibrium thermodynamics in posing prob-
lems and drawing conclusions is undeniable.

11.2. Universal Index of energy Converters’ Perfection

The necessity to have both wells and sinks of entropy, charge, sub-
stance, momentum, etc. ensuing from (13.1.2) once again emphasizes the
flexibility of the law of excluded perpetual motion of the second kind. This
statement results in a possibility for each of the energy converting devices
to be estimated by the process power Ni related to the energy flow at the
inlet to the energy converter Jе΄:

ηi = Ni /Jе′ = Ni = 1 – ψ̄i̋ / ψ̄΄i, (11.2.1)

187
where ψ̄΄i, ψ̄i̋ – mean-integral values of the generalized potential at the de-
vice inlet and outlet.
This index refers to the so-called “absolute efficiencies” class. Accord-
ing to this expression the absolute efficiency characterizes the convertibility
degree of the energy supplied from some energy well. This degree is de-
fined by exclusively the relation between the mean values of a correspond-
ing generalized potential and, providing this relation is constant, does not
depend on the working medium properties, design and other characteristics
of the energy converting system. Since energy wells with a potential of ψ̄΄i =
∞ or energy sinks with a potential of ψ̄i̋ = 0 are physically unfeasible, the
100% absolute efficiency is unattainable in even reversible processes. This
verdict generalizes the Carnot theorem known for ideal heat engines con-
firming once again the conclusion that the laws of heat conversion stated by
thermodynamics are just a particular case of the universal laws of nature
valid for any forms of energy.
Let us now show that expression (11.2.1) is valid for cyclic heat en-
gines, too. To do so, let us interrelate the so-called “mean-thermodynamic”
temperatures Т̄1 and Т̄2 in the heat absorption and rejection processes with
the mean-integral temperatures Тs′ and Тs″ of entropy flows at the device
inlet and outlet. If ΔS1 and ΔS2 characterize the variations of the working
medium entropy with the heats absorption and rejection Q1 and Q2 in some
cycle, while Δt1 and Δt2 – duration of these processes, then the total flows of
heat Jq' and Jq" and entropy Js' and Js" will be defined as:
Jq΄ = Q1 /Δt1 ; Jq˝ = Q2 /Δt2 ; (11.2.2)
Js΄ = ΔS1 /Δt1 ; Js˝ = ΔS2 /Δt2 . (11.2.3)

Considering the relationship between the flows of heat and entropy as Jq =

Тs Js gives:

Т̄1 = Q1 /ΔS1 = Jq΄/Js΄ =Тs΄ , (11.2.4)

Т̄2 = Q2 /ΔS2 = Jq˝/Js˝ = Тs˝. (11.2.5)

The identity of the temperatures averaged over the time of heat supply
to the working medium in a heat exchanger to the temperatures averaged
over the heat exchange surface allows deriving the thermal efficiency of an
arbitrary cycle in terms of the heat flow parameters:

ηt = 1 – Q2 / Q1 = 1 – Т2 /Т1 = 1 – Т̄s˝/Т̄s΄ . (11.2.6)

Then

Js΄ = Jq΄/Тs΄ = Js˝ = Jq˝/ Тs˝ . (11.2.7)

188
 Thus marking the border of the energy converting device along the sur-
face of heat exchange with the heat well and heat sink the operation of the
cyclic heat engine may be reduced to the operation of a direct-heat-
conversion machine with the same theoretical power Ne :

Ne = Jq΄ηt = Js΄ (Тs΄ – Тs˝) = Js΄ΔТs, (11.2.8)

where ΔТs = Тs΄ – Тs˝ – motive force of the heat transfer and conversion
process defined by the energy well and sink parameters.
The unified expression of absolute efficiencies for heat and non-heat,
cyclic and non-cyclic engines (see 11.2.1) shows that this value depends on
neither a kind of the energy being converted and design of the energy con-
verting device, nor on its working medium nature or parameters. Therefore
the value ηi , as well as ηt, strictly speaking should not at all be termed as
“engine efficiency”. This value is defined by exclusively the parameters of
the “hot” and “cold” energy well on the border with the working medium,
i.e. by the energy flow parameters, but not the properties of the engine as it-
self. In other words, the “absolute efficiencies”, including the “thermal”,
“internal”, “effective”, “electrical”, “enthalpy”, etc. efficiencies widely used
in analysis of heat power units characterize rather the capabilities provided
by nature due to its intrinsic spatial heterogeneity in dispensing of whatever
intensive property. These efficiencies characterize just the maximal “con-
vertibility degree” of a particular energy form, which is attained in the same
machine with the quasi-static nature of the processes involved therein (when
all the processes are “reversible”). According to the Carnot theorem this ef-
ficiency does not depend anymore on the working medium properties within
the energy converting device, i.e. relates to some abstract (idealized) ma-
chine naturally not occurring. Since all reversible processes are infinitely
slow running, the power of such a machine is infinitesimal. Therefore in
thermokine- tics dealing with real processes and machines we will be avoid-
ing to apply the term “efficiency” to the value ηi terming it rather “energy
convertibility degree”.
The so-called “relative” efficiencies have other meaning. In classic
thermodynamics this notion is introduced in order to estimate the perfection
of particular processes in heat power engines, e.g. the working medium ex-
pansion processes. Such efficiencies mean the work Wi actually done in the
process related to its theoretically possible value Wit (A. Andrushchenko,
1985). These are the efficiencies (relative) all other disciplines operate ex-
cept thermodynamics. For non-cyclic machines doing one kind of the useful
work these efficiencies actually characterize the perfection degree of a ma-
chine (i.e. the losses therein). The fact that thermodynamics operates abso-
lute efficiencies, whereas the other disciplines operate relative ones, has
caused confusion time after time. Quite often lamentations have been raised
for a low efficiency of heat power stations and up-to-60%-of-fuel heat

189
“wasted” to the environment, whereas the efficiency of fuel cells, electric
and many other machines reaches 90% and above. But for all that it has
never occurred to the “complainants” that only several percents of the Q2
heat given away in excess of a certain minimum of Q2t can be referred to the
so-called “wasted heat loss” since the heat Q2t rejected according to (13.1.3)
is not a loss, but the condition necessary to provide the energy conversion
process as itself.
To eliminate this misconception and provide a methodologically unified
approach to the converters of any forms of energy, the concept of relative
efficiency should be generalized to all kinds of the energy converting de-
vices. To this end let us consider the power ratio between the converter inlet
Nj and outlet Ni :
ηN = Nj /Ni = Xj Jjе/Xi Jiе ≤ 1 . (11.2.9)

This expression is quite close by implication to the notion of relative in-

ternal efficiency of heat engines in classic thermodynamics and goes over
into it after multiplication of the numerator and denominator in (11.2.1) by
an arbitrary time span ∆t. In this case the (11.2.1) numerator characterizes
the useful work actually done by the machine Wi = XjJjе∆t, while the de-
nominator – its theoretically possible value Wit = XiJiе∆t. We will be terming
the (11.2.1) ratio a “power” efficiency of the energy converting device. It is
this expression that ensues from the law of energy conservation in the form
of (2.4.9) for a particular case of a single-target machine intended for the
conversion if the ith form of energy into the jth one.
Let us show now that it is this efficiency that characterizes the thermo-
dynamic perfection of a machine, i.e. the degree the machine implements
those potentials the well of the technically adequate (ordered) energy pro-
vides for it. In fact, according to the principle of heat and work equivalency
(see 2.1.1) the ordered work We done by a cyclic machine related to the al-
gebraic sum of the heats đQ it receives at all stages of the cyclic process is
always equal to unity. Relating this work to only that part of the heat Q1
which is taken from the “hot” well is, from a mathematical standpoint, an
implicit technique, moreover, such one that is not always feasible for even
the heat engines (in particular, for the processes where temperature rise co-
habits with heat rejection). This even more concerns the non-heat engines
where the energy wells and sinks occupies the same space (such are, in par-
ticular, polarized, magnetized, etc. bodies). At such conditions it is advis-
able to consider only the actual power of the machine Nj related to that theo-
retically possible; such a ratio being construed as the power efficiency (see
11.2.1).
Another advantage of the power efficiency lies in its applicability to the
analysis of multi-functional machines implementing a simultaneous conver-
sion of several kinds of energy. Further advantage of the power efficiency is
its applicability to the same extent to both the direct cycles and the reverse

190
ones, for which the relative efficiency loses its meaning because the power
consumed in this process is more than theoretically required. Ultimately the
power efficiency application makes machines more compatible in the de-
gree of perfection of the processes running therein since it allows for losses
of all kinds – both at the energy transfer from the well to the converting de-
vice and at the energy conversion in the device itself.
To show the non-triviality of this concept, let us compare the ηN with
the so-called “exergy efficiency of technical systems for energy and sub-
stance conversion” ηex (Brodiansky, 1991). The latter characterizes the ratio
of outlet exergy flow to inlet exergy flow in the machine. The concept of
exergy as a measure of system working capability relative to the environ-
ment is described in Chapter 2 above. For an “extended system” (as includ-
ing wells and sinks of heat, substance, charge, etc.) the exergy becomes its
state function. The exergy flows are equidimensional with power, which
links the exergy and power efficiencies. However, the exergy efficiency is
not always associated with the conversion of some kind of energy into other
one due to the so-called “transit” exergy flows existing. Such is, e.g. the ex-
ergy of flow of some substance crossing the reactor and not participating in
the chemical transformations therein. Therefore the exergy flow ratio does
not yet characterize the effectiveness of the energy conversion in a machine.
Besides, the exergy efficiency depends on the exergy value itself as it is
commonly construed, i.e. on the reference point choice for some form of
exergy or other. This makes the exergy efficiency ambiguous, which re-
stricts the applicability of this working capability function in the efficiency
analysis for power and process installations.
Representing the power efficiency in terms of the flows Jiе and Jjе refer-
ring exclusively to the energy conversion processes eliminates the short-
comings of the exergy efficiency. The further consideration will confirm the
power efficiency is the concept that not only corresponds to its physical
meaning, but also provides the most objective assessment of the thermody-
namic effectiveness of any energy converters.

11.3. Kinetic Equations of energy Conversion Processes.

Antisymmetry of Relationships Therein

As shown above, any device converting some ith form of energy into the
th
j one is a two-flow system, i.e. features energy carrier flows of both the
energy being converted Ji and the energy having been converted Jj. The in-
terrelation between flows Ji and Ji is a distinctive feature of the energy con-
version processes. One can make sure such an interrelation exists if com-
pare, based on the law of conservation of energy, the expressions of the en-
ergy being converted and having been converted:

191
 Ni = Niд + Niе = Njд + Njе. (11.3.1)

This relationship is an equation of balance between the power input to

the machine and power output from the machine. According to this equation
the power input Ni is consumed to overcome both the subscript-similar dis-
sipation forces Xi and the “alien” forces Xj, i.e. is not only converted into
other form of the useful energy Niе, but also partly dissipates on the input re-
sistors (this loss is usually designated as Nid). In its turn, the power Nj con-
verted into the jth form is also partly lost in the secondary equivalent circuit
(Fig. 11.1) thereby reducing the output useful power Njе necessary to over-
come the useful forces Xjе. Since by the law of conservation of energy Niе =
XiJiе = Njе = XjJjе, the flows Jiе and Jjе appears to be interrelated. This interre-
lation shows itself in the fact that each of the flows Ji and Jj appears to be
dependant on both the active forces Xi and Xj acting in the system, i.e. the
kinetic equations of the energy conversion process should be written a priori
in the form of Onsager’s phenomenological laws (4.1.6):

Ji = Lij Xi + Lij Xj , (11.3.2)

Jj = Lji Xi + Ljj Xj . (11.3.3)

The number of the phenomenological coefficients in these equations

may be reduced by applying the reciprocity relationships to them. However,
it is advisable to preliminarily make sure the reciprocity relationships (see
4.5.3) for the energy useful conversion processes bear invariably anti-
symmetrical character. To this end let us take into account the difference in
sign for the work done to the energy converting device (this work is nega-
tive according to the rule of signs in thermodynamics, i.e. đWi = – XidZi < 0)
and the work done by this device to the work object đWj = XjdZj > 0. With
the sign properly considered the exact differential of system exergy may be
expressed as

dЕ = – XidZi + XjdZj . (11.3.4)

Repeating the same reasoning as described in paragraph 4.3 when deriv-

ing the reciprocity relationships, based on the properties of the inergy Е ex-
act differential, one comes to the anti-symmetrical reciprocity relationships:

(∂Ji /∂Xj) = – Lij = (∂Jj/∂Xi) = Lji . (11.3.5)

This proof of the so-called Cazimir’s reciprocity relationships differs

from those earlier proposed (V. Etkin, 1993) by a particular simplicity and

192
obviousness. The opposite sign of the works đWi and đWj taken into account
results in kinetic equations (11.3.2)–(11.3.3) taking the form:

Ji = Lij Xi – Lji Xj , (11.3.6)

Jj = Lji Xi – Ljj Xj . (11.3.7)

The kinetic equations of such a character (with their terms having dif-
ferent signs) better correspond to the concept of phenomenological (experi-
ence-based) laws than initial equations (11.3.2)–(11.3.3). The primary en-
ergy carrier flow Ji (e.g. the current in the primary winding of a transformer)
is commonly known to decrease as the forces Xj being overcome increase
(with approach to “no-load” operation) or, on the contrary, to increase as
these forces decrease (with approach to “short circuit” operation). Similarly,
the secondary energy carrier flow (e.g. the current in the secondary winding
of a transformer) is commonly known to increase as the supply voltage Xi
increases and to decrease as the secondary circuit resistance and the Xj de-
crease.
The proof of the Cazimir’s reciprocity relationships adduced here for
reversible part of the energy conversion processes is based on the law of en-
ergy conservation and, therefore, valid whenever linear phenomenological
laws (11.3.2) and (11.3.3) remain valid. This circumstance throws fresh
light on the origin of the Cazimir’s reciprocity relationships exposing the
underneath meaning of the requirements for different parity of forces with
respect to time inversion. In fact, for the dissipation forces not changing
their signs with time inversion (i.e. for the so-called “α-type forces”) the
Onsager’s symmetry conditions, as shown above, are valid. Whenever a part
of these forces have the reversible character (i.e. refer to the “β-type
forces”), the Onsager’s reciprocity relationships give place to anti-symmetry
conditions (11.3.5). At the same time the consideration endeavored here
shows that the applicability of the Cazimir’s relationships is not actually re-
stricted to the different-parity forces (α and β-type) case. In fact, let us as-
sume dealing with energy conversion processes of purely dissipative charac-
ter. Such are, in particular, thermal conductivity, electric conductivity, dif-
fusion and viscous friction described by Fourier’s, Ohm’s, Fick’s and New-
ton’s laws and resulting in only the substance and energy transfer. In this
case all terms of kinetic equation (4.4.1) describing vector phenomena have
the same sign Хi·Ji > 0 defined by their contribution to dissipative function
(5.3.3). In this case reciprocity relationships (4.5.4) defining value and sign
of the phenomenological coefficients Lij in linear kinetic equations (4.4.1)
give invariably positive values of the phenomenological coefficients Lij > 0
in these equations and result in Onsager’s reciprocity relationships Lij = Lji.
So for linear transfer processes of a purely dissipative character the ma-
trix of phenomenological coefficients is always symmetrical. However, if in

193
a transfer process useful (reversible) energy conversions occur, i.e. work is
done against whatever forces other than dissipation ones, the reciprocity re-
lationships acquire other character. In this case phenomena of the “ascend-
ing diffusion” type (transfer of components toward their concentration in-
crease), system ordering, etc. are observed. These processes lead to gradi-
ents or differences of temperature, pressure, concentration, electric poten-
tial, etc., i.e. to deviation of the system from the internal equilibrium state
for some of its degrees of freedom, whereas the system in whole is tending
toward equilibrium. As a matter of fact, this is the nature of all the so-called
“superposition effects” the theory of irreversible processes deals with. As
we will make sure hereafter, the effects of such a kind bear anti-dissipative
character. The processes of useful conversion of the ith form of energy into
the jth one in various machines apply to these effects, too. This substantially
extends the applicability of the anti-symmetrical reciprocity relationships
(V. Etkin, 1993).
Let us now elucidate what character the relationships between the phe-
nomenological coefficients Lij acquire in a more general case, when these
coefficients are some functions of the related thermodynamic forces Xj, so
that kinetic equations (11.3.5) and (11.3.7) are non-linear. Applying to them
generalized reciprocity relationships (11.3.5) gives for this case:

(∂Jj /∂Xj ) = Lij(Xj) + ΣjXj(∂Lij(Xj)/∂Xj ) , (11.3.8)

so the generalized reciprocity relationships become:

Lij(Xj) + Σj Xj(∂Lij /∂Xj ) = Lji(Xj) + Σj Xj(∂Lji /∂Xi)... (11.3.9)

From this it follows that the Onsager-Cazimir’s symmetry conditions Lij

= ± Lji apply to also those cases when the diagonal phenomenological coef-
ficients

Lii = Lij(Xj), Ljj = Ljj(Xi),

i.e. are some functions of the forces associated with them, whereas only mi-
nor (non-diagonal) terms of the phenomenological laws characterizing the
superposition effects are linear. In fact, if the non-diagonal coefficients Lij ,
Lji are constant, then all expressions within brackets in relationships (11.3.9)
become zero. In other words, to meet the symmetry conditions or non-
symmetry ones, it is enough the non-diagonal terms in equations (11.3.6)
and (11.3.7) responsible for the superposition effects to be linear. We have
demonstrably instantiated this by the phenomena of cross diffusion and fil-
tration of ideal gases thru the finite-thickness membrane (see Chapter 11).

194
 It is a matter of interest to confirm the anti-symmetrical reciprocity rela-
tionships by specific examples of the useful energy conversion processes.
As one of the examples, the thermoelectric generator (TEG) may be consid-
ered which operation is based on the Seebeck and Peltier thermoelectric ef-
fects. If the ends of different-type conductors A and B are connected and a
current source or a load switched into the break of one of them, then in the
formed thermo-circuit along with the Thomson’s homogeneous effect (elec-
tric potential gradient generated along the dead conductor with heterogene-
ous temperature field) the so-called heterogeneous effects will be observed.
One of them arises with the hot and cold junctions of the thermocouple
maintained at different temperatures T1 and T2 < T1. It consists in the gen-
eration of the electromotive force Xe = -Δϕ proportional to the temperature
difference Xq = – ΔТ = T1 – T2 and reaching the maximum value in the ab-
sence of the electric current Je (Haase, 1967):

(Δϕ/ΔТ)cт = εAB/T (Je = 0), (11.3.10)

where εAB – the so-called thermoelectric force in the thermo-circuit of con-

ductors A and B depending on their nature and the mean temperature of the
thermo-element T.
Another effect arises with the current Je passed thru the junctions and
consists in absorbing or releasing by the junction a Peltier heat which flow
Jq is proportional to the current Je and the junction temperature T:

Jq = εAB Je T . (11.3.11)

With the junctions of the thermo-circuit maintained at different tem-

peratures, the electric current Je generated will lead to the heat absorption Jq'
= εABJeT2 in junction А and the heat release Jq˝ = εAB JeT2 in junction В.
Then, in the absence of side processes, the thermo-element becomes similar
to a cyclic heat engine with working medium being the free electrons circu-
lating in the thermo-circuit. The dependence of the thermal conductivity λ
and the electric conductivity σе of the thermo-circuit branches on the fields
of temperatures along the conductors (and, thus, on the electric fields asso-
ciated) makes the TEG phenomenological laws, strictly speaking, non-
linear. However, the said non-linearity concentrates in their diagonal terms
so that the TEG phenomenological laws become (Haase, 1967):

Jq = Lqq (Xq) Xq – Lqe Xe , (11.3.12)

Jе = Leq Xq – Lee (Xe) Xe , (11.3.13)

where Lqq(Xq), Lee(Xe) – coefficients inverse to, respectively, thermal and

electrical impedance of the thermo-circuit branches considered as a function

195
of their mean-integral temperature; Lqe, Leq – cross phenomenological coef-
ficients considered as constant. Here in equations (13.3.6) and (13.3.7) the
circumstance is taken into account that with the voltage drop Xe at the TEG
output the current Je on the load and Peltier heat flow rise.
To make sure the anti-symmetry conditions Lqe = –Leq for TEG are met
also in the case the coefficients Lqq and Lee depend on the associated forces,
let us apply the differentiation process corresponding to the left side of
equation (11.3.5) to equation (11.2.12). Allowing for constancy of Lqq(Xq) at
Xq = сonst gives:

(∂Jq /∂Xe ) = – Lqe . (11.3.14)

The same with the (13.3.5) right side gives:

(∂Jе /∂Xq ) = Leq . (11.3.15)

Thus, Cazimir’s reciprocity relationships Leq = – Lqe for TEG remain

valid also with non-linear character of the main terms in equations (11.3.12)
and (11.3.14).
A significant consequence from the anti-symmetry of phenomenological
coefficients matrix for the useful energy conversion processes is the absence
of limitations for the value of the cross phenomenological coefficients,
which ensue in TIP from the condition of positive determinacy of their ma-
trix (4.1.9). In fact, at Lij = – L ji condition (4.1.9) is met with any values of
the coefficients Lij and Lji. As will be shown a little bit hereafter, this cir-
cumstance is of key importance since removes the limitations for efficiency,
which is unusual to energy converting systems.

11.4. Similarity Criteria for energy Converting Systems

The unity of phenomenological laws (11.3.6)-(11.3.7) describing differ-

ent forms of energy conversion processes and their independence on ma-
chine hardware implementing such processes allows laying a foundation to
the similarity theory of linear energy converting systems. The math model
of such a system includes, along with equations (11.3.6)-(11.3.7), the
uniqueness conditions of processes under investigation. These conditions
include:
a) conductivity Lij or resistance Rij coefficients (i,j = 1,2) characterizing
the transport properties of the system (analog of the thermo-physical proper-
ties in the heat and mass transfer theory);
b) boundary conditions defined in this case by the values of motive
forces at the current conditions Xi, Xj, or by the flow values at these condi-
tions Ji, Jj;

196
 c) initial conditions defined by these force or flow values in “no-load”
operation of the machine (at Jj = 0) Xjо, Jiо, Jjо and in “short circuit” opera-
tion (at Xj = 0) Jik, Jjk. These uniqueness conditions allow forming a number
of dimensionless criteria of energy converting systems similarity. Equations
(11.3.6)-(11.3.7) are more practical to be expressed in terms of the resis-
tance coefficients Rij in the form:

Xi = Rii Jj – Rij Jj ; (11.4.1)

Xj = Rji Jj – Rjj Jj . (11.4.2)

Let us consider these equations jointly with the uniqueness conditions.

Assuming Jjo = 0, Xi = Xio, Xi = Xio they can be derived as:

Xio /Xio = Rjj /Rii ; Jj = Xjo /Rji ; Jj =( Xj – Xjo )/Rjj . (11.4.3)

Similarly for “short circuit” conditions (Xj = 0, Jj = Jjk, Jj = Jjk) equation

(11.4.2) becomes:
Jjk/Jik = Rji /Rjj . (11.4.4)

Substituting equations (11.4.1) and (11.4.2) in power efficiency expres-

sion (11.2.11) and making a number of transformations using relationships
(11.4.3) and (11.4.4) gives:

ηN = (Xj/Xj o) /[1 + Rii Rjj /Rji2(1 – Xj/Xj o)]. (11.4.5)

This equation. n allows reducing the uniqueness conditions characteriz-

ing the transport properties of the energy machine to a one dimensionless
complex:

Ф = Rji2/Rii Rjj . (11.4.6)

This complex is similar to the relationship between active and reactive

resistances known in radio engineering as tuned-circuit Q-factor and coin-
cides (accurate up to a temperature factor) with the “torque-to-weight ratio”
Ф introduced by A. Yoffe as a generalized characteristic of thermoelectric
generators (TEG). Therefore it was termed as a good quality criterion of
machine (V. Etkin, 1990). Its value ranges from zero to infinity (0 < Ф < ∞)
increasing with decrease of the “active” resistances Rii and Rjj (from the dis-
sipation forces) and increase of the “reactive” resistances Rji (from the use-
ful forces). Like the thermal resistances these ones depend on the transport
properties of the system (cross-sections and lengths of the heat, substance,
charge, etc. transfer lines, properties of the materials applied, transport fac-
tors, etc.), i.e., ultimately, on the design perfection of the machine indirectly

197
depending on also its manufacturing expenditures. Based on this, the above
criterion could also be termed as a criterion of design perfection of machine.
Other dimensionless criterion can be composed of boundary conditions
defined by the values of the forces Xj, Xjo or flows Jj, Jjk:

B = Jj /Jjk = 1 – Xj /Xjo. (11.4.7)

This criterion depends on exclusively the load to the machine (“load

current”) and ranges from zero at no-load conditions (Jj = 0) up to unity at
short circuit conditions (Xj = 0). Therefore it was termed as a load criterion
of machine (V. Etkin, 1990, 1991).
Using these criteria expression (1.4.5) can be formed as a criterion
equation of energy conversion process:

ηN = (1 – B)/(1 + 1/BФ) . (11.4.8)

This equation determines the dependence of the thermodynamic crite-

rion of machine perfection upon machine load B at invariable parameters of
the technically fit energy source. Thereby one more step has been made on
the way to approach the thermodynamic assessment of technical system ef-
ficiency to reality.

11.5. Universal Load Characteristics of energy Converters

Like the theory of heat exchange processes similarity, criterion equation

of energy conversion processes (11.4.8) allows extending the results of effi-
ciency investigation from some technical systems to other ones (insuffi-
ciently explored). To do so, it is advisable to build based on criterion equa-
tion (11.4.8) a generalized load characteristic of linear energy converting
systems. Such a characteristic is plotted on Fig.11.2. Solid lines show here
the power efficiency of the machine ηN plotted against the load criterion B
for different values of the Q-criterion Ф, while dotted lines – the output
power Nj plotted against the load. As follows from the figure, in the absence
of energy losses (Ф = ∞) and with quasi-static character of its conversion (B
→ 0) the machine efficiency reaches, as should be expected, unity. How-
ever, in all other cases the power efficiency becomes zero twice: in “no-
load” (B = 0, Jj = 0) and “short circuit” (B = 1, Xj = 0) operations. This en-
sues from taking into account, along with the irreversible power interchange

198
and friction, also various “leaks” 1)of power occurring in also “no-load” op-
eration of the machine, in particular, its auxiliary power. With the loads
growing and the operation withdrawing from “no-load” conditions the effi-
ciency rises and reaches maximum at a definite load. The efficiency maxi-
mums then lie on the same line connecting the operating conditions with
“zero” output power (В = 0, ηN = 1 и В = 0,5 , ηN = 0). The efficiency
maximums appeared are explained by varying relationship between the rates
of useful and dissipative energy conversions in the system and show that all
types of energy converters have the most economical loads usually adopted
as nominal.
To find the nominal loads Bн, let us set the derivative of (11.4.8) with
respect to B equal to zero. Some transformations give:

Bн = ( 1 + Ф – 1)/Ф. (11.5.1)

According to this expression the power efficiency maximums lie on the

same line connecting the points with the load B = 0.5 and the efficiency ηN
=1. The max efficiency value depends in this case on exclusively the Q-
criterion value. In fact, substituting (11.5.1) into criterion equation (11.4.8)
gives after simple transformations:
ηN max = ( 1 + Ф – 1)/( 1 + Ф + 1) . (11.5.2)

As can be seen, with the Ф -factor of the machine rising (by additional investment
1.0 ηN 1.0
operation of high-efficiency power Φ=∞ N/Nmax 0.9
0.9
machines on partial loads is inad-
Energy (power) efficiency

0.8 50 0.8
visable. To clarify how the load 0.7 0.7

Relative power
impacts on the output power of a 0.6 20 0.6
machine, let us substitute (11.4.8) 0.5 10 0.5
into the relationship Nj = NiηN : 0.4 5 0.4
0.3 3 0.3
Nj = Xjo JjkB(1 – B) . (11.5.3) 0.2 0.2
1
0.1 0.5
0.1
B
Since at B = 0.5 the power Nj is 0 0.2 0.4 0.6 0.8 1.0
maximal (Nj = Njmax), the value Relative load
XjоJjk = 4 Nj , and instead of (11.5.3) Fig.11.2. Universal Load
one can write Characteristics of Linear Systems

Nj /Nj max = 4 B(1 – B) . (11.5.4)

1)
This loss can not be taken into account by traditionally introducing the constant relative
efficiencies of work-doing processes since this loss is equal to zero in “no-load” operation.

199
 This plot is also shown on Fig.11.2 with dotted line. From this it follows
that at the efficiency ηN = 1 the output power of any heat engine is equal to
zero due to the absence of heat exchange between the heat well and working
medium and reaches maximum at a relative load of B = 0.5. The efficiency
of the engine corresponding to these loads can be found by substituting this
load value into (11.4.8):

η(N = max) = 1/(2 + 4/Ф) . (11.5.5)

According to this expression the power efficiency of linear energy con-

verters at max power conditions does not exceed 50 % for all forms of en-
ergy. For all that the difference in efficiencies of machines with different Ф
-factors is leveling off with approaching the max-output power-operating
conditions and may become practically indistinguishable. This fact reveals
lack of prospects in pursuing high efficiencies for power machines intended
for peak-load or power-augmentation operations. Fig.11.2 also shows that
with the Ф-factor of a machine rising, the dependence of its efficiency on
load grows more perceptible, while the operating conditions of max effi-
ciency and max output power are more and more drifting apart.
The indirect confirmation of the anti-symmetry of reciprocity relation-
ships is another significant conclusion from the load characteristics. In fact,
if the Onsager’s symmetry conditions were valid for energy converting ma-
chines, then they would engender known limitations for the relationship of
diagonal and non-diagonal phenomenological coefficients Lij2/LiiLjj ≤ 1 or
Ф = Rji2/Rii Rjj ≤ 1 as based on the positive determinacy of phenomenological
coefficients matrix (4.1.9). In accordance with Fig.14.2 this means that the
exergy efficiency of any energy converters can not exceed ~17.5 % as a
limit. Meantime a lot of machines are known with this efficiency exceeding
the limit. The said contradiction is eliminated if Lij = – Lji since in this case
condition (4.1.9) means just the requirement these coefficients to be positive
(Lii Ljj ≥ 0).
The universal load characteristics are very convenient when comparing
machines of different types for choosing the most challenging of them.
Since such an assessment depends on the load of a machine ranging consid-
erably in operation, the comparison needs the load characteristics to be
known otherwise being insufficient. Let us show this by example of a hy-
drazine-oxygen electrochemical current generator (ECG). When operating
this fuel cell, a no-load voltage (EMF) was obtained as Xjo = 1.6 V which
decreases with load growth and becomes Xj = 0.85 V at a power of Nj = 2
kW. Fuel utilization factor and oxidant utilization factor for this unit are ηf
= 0.8 and ηox = 0.9, respectively (B. Nesterov and others, 1980). To find the
ECG power efficiency, let us first calculate the fuel element relative load B
= 1 – Xj/Хjо = 1 – 0.85/1.6 = 0.47. For ideal ECG (with the Ф = ∞) under
such a load according to Fig.13.2 ηN does nor exceed 0.51. Allowing for the
200
mean factor of reagents utilization as ηr = 0.85 gives the effective efficiency
for this generator ηe = ηN ηr = 0.51·0.85 = 0.43. Detailed calculations with
an exergy flow balance analysis give for this ECG the exergy efficiency ηex
= 0.41 (V. Etkin, 1991), i.e. quite close to ηe. At the same time such a
proximate analysis reveals the importance of considering the loads to ma-
chines when investigating how far they are challenging. Should in the ex-
ample considered the ECG effectiveness be estimated thru the max power
efficiency ηN ≅ 0.85, the situation would be cardinally outweighed on its
side.

11.6. Similarity of Load Characteristics for Real Machines

It is a matter of interest to confirm the similarity of load characteristics

for a number of different-type energy converters. As one of the examples,
let us consider a space-based propulsion system (SBP). For spacecrafts
moving in the absence of air drag and gravity the useful work means the ac-
celeration dv/dt of a spacecraft with the mass M under the thrust R. This
thrust is defined by a known expression (Favorsky and others, 1970):

R = Gm w + Fс (pc – po )n = Gm wэ = Rу g Gm , (11.6.1)

where Gm – gas flow thru jet nozzle, kg/s; w, weq - true and equivalent gas
outflow velocity, respectively; pn , po – pressure of combustion products at
the nozzle outlet and outside in the environment, respectively, N/m2 ; Rs, g –
specific thrust (specific impulse) of the jet engine and gravitational accelera-
tion, respectively.
In the general case of flights in the atmosphere or gravity field the part
Njd of the converted power Nj works against the environmental resistance
forces. Here the useful effect is deemed as the movement of a flying vehicle
and the Nj value defined as the scalar product of the thrust R by the distance
covered per unit time under this thrust. For spacecrafts taking off from the
orbit of the Earth with the starting orbital velocity vо this distance is obvi-
ously equal to the velocity of spacecraft movement relative to the starting
coordinate system vк = v – vо since at v = vо no work is done:

Nj = R⋅(v – vо ) . (11.6.2)

This expression is applicable also for the engines correcting the satellite
orbit and for the flying vehicles taking off from the Earth’s surface (vо = 0).
Taking into account that the vectors R and v are anti-parallel (so that the
problem may be considered as one-dimensional) the output power Nj may
be represented as the product of the scalar thermodynamic force Xj =⏐R⏐=

201
R as a motive force or acceleration of the vehicle by the scalar flow Jj = ⏐v
– vо⏐= v – vо as a generalized velocity of the movement process.
To find the flows and forces of the energy form being converted, let us
apply to the first law of thermodynamics for unit mass of moving gases.
Under this law, if the processes in combustors and nozzles are running at
adiabatic conditions, the specific useful work of the vehicle movement rela-
tive to the starting coordinate system is defined by the difference in enthalpy
of the combustion products deceleration between the combustor inlet hc* =
hc + vс2/2 and the nozzle outlet hк* = hк + vк2/2. According to this the
power input to the engine is defined as:

Ni = Gm (hc∗ – hк ∗) . (11.6.3)

This expression may be also represented in the form of the product of

the scalar gas flow Ji = Gm as a generalized rate of the process by the scalar
thermodynamic force Xi = –Δh*. The difference of enthalpies Δh = hк – hс
is known to define the theoretical velocity wt of gas outflow from the nozzle
(wt2/2= hк – hс ) which relates to the true gas outflow velocity w thru the
relative internal efficiency ηoi = w2/wt2. Taking into account that for space
engines the equivalent gas outflow velocity from nozzles weq is substan-
tially equal to the true outflow velocity w, while the absolute outflow veloc-
ity for working medium at the nozzle outlet in the starting coordinate sys-
tem vc is equal to the difference between w and the velocity of the flying
vehicle in this coordinate system v, the value Xi may be represented as:

hк* – hс* = w·(v – vо ) + (1/ηoi – 1) w2 /2 . (11.6.4)

The augend on the right side of this equation defines the useful compo-
nent of thermodynamic force Xi е = w·(v – vо) associated with the vehicle ac-
celeration, while the addend – the dissipative component associated with the
initial energy dissipation in the SBP combustors and nozzles (including
losses caused by incomplete expansion, dissociation and ionization of work-
ing medium, as well as friction loss) Xir = (1/ηoi –1)w2/2. The useful force
component Xiе = w·(v – vо) is proportional to the flow Jj =|v - vо| thru the
proportionality factor –Rij = |w|, while the dissipative component – to
squared velocity (and, consequently, squared flow Gm = Ji). This causes
non-linearity of the relationship Xi = Xi(Ji). However, this non-linearity is
concentrated in the diagonal term that may be represented as Xir = RiiGm =
[(1/ηoi – 1)w2/2G]Gm, where the diagonal phenomenological resistance coef-
ficient Rii = (1/ηoi – 1)w2/2Gm = Rii(Gm), i.e. is a function of the flow associ-
ated.
Similarly, the thrust Xj = R, in the general case of flights in the atmos-
phere or gravity field, may be represented as a sum of the useful component

202
Xjе = Gw associated with the vehicle acceleration and proportional to the
flow Ji = Gm thru the proportionality factor Rji = w and the dissipative com-
ponent Xjr associated with the atmosphere resistance to be overcome and
depending on the flight velocity Jj = |v – vо|. This dependence is also non-
linear, but may be represented in the pseudo-linear form Rjj(Jj)Jj thru the
proportionality factor Rjj(Jj) depending on the velocity increment Δv.
Thus the SBP phenomenological laws may be expressed in informal
variable as:

–Δh* = Rii(Gm) Gm – w⋅(v – vо ) , (11.6.5)

R = w Gm – Rjj (v – vо ) . (11.6.6)

Here the negative sign of the second term in equation (13.6.5) allows
for the opposite direction of the force Xi and the flight velocity v – vо. In ac-
cordance with this equation at Δh* = const increasing the flight velocity
demands increasing the working medium flow, which is the reality. Simi-
larly the negative sign of the second term in equation (11.6.6) allows for the
fact that at the constant thrust and constant flow of working medium G in-
creasing the flight velocity v – vо demands increasing the propulsive jet ve-
locity w, which is quite natural, too. Such a character of the SBP phenome-
nological laws leads to the anti-symmetry of reciprocity relationships. In
fact, applying relationships (11.3.5) to equations (11.6.5) and (11.6.6) gives:

(∂Xi /∂ Jj ) = – Rij = (∂Xj /∂ Ji ) = Rji = w . (11.6.7)

The fact that the non-linear laws of mechanical energy forms conver-
sion in SBP are quite within the frames of thermokinetics substantially ex-
pands its applicability and allows extending its analytical methods to com-
plex mechanical systems, which seems to be quite important (Burdakov,
1985).
Now let us find the SBP power efficiency using the above-obtained ex-
pressions for input and output powers of the machine. Simplifying expres-
sion (11.6.4) as:
hк*– hс* = w2/2ηoi + v2/2 , (11.6.8)

gives that the SBP power efficiency ηN takes the form:

ηN = w⋅(v – vо )/( w2/2ηoi + v2/2) . (11.6.9)

At vo = 0 and ηoi = 1 this relationship goes over into a known expres-

sion for the propulsion (thrust) efficiency (Alemasov, 1962). This differs
from the SBP total efficiency by the velocity increment for the flight time
(v – vo) it features instead of the absolute flight velocity v in the SBP total

203
efficiency. Thereby it is taken into account that maintaining the flight veloc-
ity v = vo in the absence of the air drag does not demand any work. Expres-
sion (11.6.2) can be reshaped to a more general form by introducing the re-
lation B = vo /v as the SBP load criterion:

ηN = 2В(1 – В)/[В2(w/vo)2 /2ηoi + vo /w] . (11.6.10)

The above relationship is correspondingly plotted on Fig.11.3. It is easy

to see that its character completely corresponds to the universal load charac-
teristics on Fig.11.2 despite the non-linearity of SBP phenomenological
laws (11.6.5) and (11.6.6). Fig.11.3 shows that the machine efficiency as a
function of the relative load B (flight relative velocity vo/v) becomes zero
twice: at B = 0 and B = 1. The first
1.0
ηN case corresponds to near-the-light
0.9 speed flights when a further flight
1
0.8 speed increment becomes impossi-
0.7 2
3 ble (Jj = Jjk – just like at “short-
0.6
4 circuit” conditions). Such a situa-
0.5 5 tion is of theoretical interest for the
0.4 spacecrafts with the solar sails (Fa-
0.3 vorsky and others, 1970), which
0.2
thrust drops to zero as the flight
0.1
vо/v speed is approaching the light
0 0.2 0.4 0.6 0.8 1.0 speed.
The opposite case (B = 1) ap-
1, 2, 3, 4, 5 – Rs/v0 = 1.0; 0.5; 0.2; 0.15; 0.1 plies to any spacecrafts at the mo-
ment they are taking off, when the
Fig.11.3. Generalized Curves SBP
Efficiency – Fight Speed flight speed is equal to the starting
one (v = vo). With the flight speed
increasing the efficiency curve reaches its peak value which position de-
pends on the vo/w ratio and the perfection degree of the internal processes in
the engine.
As another example let us compare the individual characteristic of a
thermionic energy converter (TEC) built based on the universal load charac-
teristic (Fig.11.2) for a known Q-criterion Ф with the power and effective
efficiency of one of such converters plotted experimentally against its out-
put voltage Δϕ (M. Tribus, 1970). When investigating the TEC Q-criterion,
we will proceed from the TEC characteristics, according to which the TEC
effective efficiency ηe does not exceed 11% at the cathode and anode tem-
peratures 2,2000C and 300C, respectively. Proceeding from the fact that at
these temperatures the theoretical efficiency of ideal TEC ηtК ≈ 0.88 gives
the power efficiency ηN =12.5% corresponding to the above effective effi-
ciency. In principle, this is enough to choose from the curves on Fig.11.2
and to plot a TEC individual load curve. To be able to compare it with the

204
experimental curve of the effective efficiency ηe against the TEC output
voltage, it is necessary to convert it into a generalized form of ηN = ηN(В).
This may be easily done by relating the TEC output voltage Δϕ to the “no-
load” voltage (TEC EMF) Δϕ0 = 0.9 V since for the TEC B = 1– Δϕ/Δϕo.
The results of such a conversion are illustrated on Fig.11.4. Here the light
dots denote the experimental values of effective efficiency, while the black
dots – the experimental values of the TEC specific power. As follows from
the figure, in general, the character of the curves ηN = ηN(В) and N = (В)
well complies with the universal load characteristics.
The universal character of the load characteristics on Fig.11.2 may be
further instantiated by the experimentally-found load characteristics of vor-
tex-type gas energy separators based on the Ranque-Hilsch effect. The phe-
nomenon of homogeneous gas energy separation into “cold” and “hot”
flows known as the “vortex effect” or
ηN , % Nj,
Ranque-Hilsch effect was discovered in 18 W/m2
1933 by the French metallurgical engi- 0.8
neer J. Ranque and then more thor- 16
0.6
oughly investigated by R. Hilsch in
14
1946. The vortex tube (VT), where the 0.4
Ranque-Hilsch effect appears, is a 12
smooth cylindrical or conical tube with 0.2
one of the ends closed by a diaphragm 10
B
0.0
having a central hole and the other end 0.0 0.1 0.2 0.3 0.4 0.8
provided with a throttle having a size-
controllable annular slot positioned Fig.11.4. Load Characteristics of
Thermionic energy Converters
around the tube periphery. The tube is
equipped with tangential nozzles positioned mostly near the diaphragm.
Compressed thermally homogeneous gas is fed thru the nozzles to the tube
forming two vortices there drifting in the opposite directions inside the tube.
One of them, peripheral, rotates with about constant rotational velocity vτ =
ωr = const (where ω – angular velocity; r – vortex current radius) and drifts
toward the throttle leaving it notably heated if throttled down enough; while
the other vortex, central, rotates by the laws of rigid body vτ/r = const and
drifts toward the diaphragm leaving it cooled. The cooling effect rate may
reach (60-70)0C and more, which causes a keen interest in the vortex effect
from both theoreticians and empirics. Vast information is presently avail-
able on experimental investigation of this effect. Fig.11.5 demonstrates a
piece of the most complete data of full-scale tests made on the vortex tube
(Biruk, 1993). Here the dotted lines denote the separation factor ηT = 1 – Th
/Tc (where Th , Tc – absolute temperatures of the “hot” and “cold” flows, re-
spectively) plotted against the relative load of the machine defined by the
cold air share μ ≡ B for various ratios between the gas pressures upstream
and downstream of the nozzle π = p1/p2 =2, 5, 7. In that case the relative

205
load value ranged from zero (at “no-load”) to unity (“short circuit” analog).
The solid lines on the figure denote the vortex tube “cooling capacity” μηT
related to primary gas flow unit and measured by the product of cold air
share μ and the temperature separation factor ηT and plotted against the load
0.24 η В ≡ μ. As this value is nothing else
μηT 0.10
0.22
T
0.09 but the outlet gas flow exergy per
0.20 0.08 gas flow unit, the curves on
0.18 0.07 Fig.11.5, at π given, are nothing
0.16 0.06 else but the individual load charac-
0.14 0.05 teristics of the vortex tube plotted
0.12 0.04 to a certain scale. It can be easily
0.10 0.03 seen that the curves ηT = ηT (В) and
0.08 0.02
μηT = μηT (В) in their character sat-
0.06 0.01
0.04
μ≡B isfactorily reproduce the universal
0 0.2 0.4 0.6 0.8 1.0 load characteristics of linear energy
Fig.11.5. Load Characteristics of converting systems ηN = ηN(В) and
vortex-type gas energy separators N = N(В) despite the non-linearity
of gas expansion in the nozzles of
the vortex separators. It is significant that with increasing π (Ф -criterion
analog) the peak values of the separator economical efficiency ηT drift to-
ward the relative loads В = μ reducing, whereas the capacity peak values
ВηT fall at same value B – just as it follows from the theory of similarity of
energy conversion processes. Thus the non-linearity of the processes run-
ning in the vortex tube tells on mostly the locations of the said peak values.
In particular, the location of the capacity peak changes from B = 0.5 to B =
0.55…0.6. This allows using the deductions of thermokinetics for analysis
of real machines thus facilitating the choice of their optimal operating con-
ditions.
The fact the max efficiency and max power operating conditions exist is
also confirmed by the efficiency curves of electro-technical energy convert-
ers, characteristics of pumps and fans, velocity characteristics of internal
combustion engines, etc. All this evidences that thermokinetics correctly
mirrors the general mechanisms of non-static energy conversion processes
thus moving one step further toward the reality.

206
 Chapter 12
THEORY OF ENGINEERING SYSTEMS PRODUCTIVITY

Perceiving the rate and the productivity of real processes as one of their
efficiency criteria has revealed two more branches in thermodynamics of the
XXth century along with the theory of irreversible processes (TIP) and
called, respectively, “thermo-economics” and “finite-time thermodynamics”
(FTT). Of the two branches, the former originates in the M. Tribus’s and his
research fellows’ works (Tribus, Evans, Crellin, 1966) and is intended for
commensuration of savings in current fuel and materials due to machine ef-
ficiency upgrade with the investments associated. Introducing economics
elements in thermodynamics modifies the engineering system efficiency cri-
teria themselves since not the maximum of efficiency optimizes now the pa-
rameters of engineering systems, but the minimum of their planned costs.
This direction involving also the heat exchange theory along with thermo-
dynamics was further developed and widely applied in our country largely
due to the works by A. Andrushchenko and his school (1974).
The second branch assigned a top priority to define conditions to obtain
maximal economic efficiency considering restricted contact of working me-
dium in power plants with heat well and heat sink. This is the theory that
first and most generally framed the question of interrelation between power
(productivity) of engineering systems and their thermodynamic efficiency,
i.e. of essentially the extreme capabilities of real processes given their irre-
versibility. However, this theory in the available form is suitable for only
the machines operating under maximum power conditions. In this context
quite topical is the question how to construct a more general theory of engi-
neering systems productivity, which should comprise both of the said
branches as particular cases and relate to the classic theory of heat engines
the same way as dynamics to statics.

12.1. Synthesis of Thermokinetics with Thermoeconomics

The abovementioned independent development of the three branches of

present-day thermodynamics (theory of irreversible processes (TIP),
thermo-economics and finite-time thermodynamics (FTT)) is largely condi-
tioned by the difference in efficiency criteria for the real (non-static) energy
transfer and conversion processes theses branches investigate. These criteria
are as follows: for TIP – minimum of entropy generation in a system σs (De
Groot, Mazur, 1964); for thermo-economics – minimum of machine
planned costs Зр = min (A. Andrushchenko, 1963; M. Tribus, 1970); for
FTT – maximum of machine output power Nj = max (Barrer, 1982;

207
Rudenko, Orlov, 1984). Creation of thermokinetics as a unitary thermody-
namic theory of real processes and the associated theory of similarity of
power and processing plants allows including load and productivity of such
machines in the number of the parameters being optimized. This facilitates
elaborating a unitary efficiency criterion for power and processing plants
and creating on its basis a theory of engineering systems productivity con-
sidering not only thermodynamic parameters, but also operation condition
and cost.
Let us adopt for an objective function when optimizing the operation
conditions of operating power and processing plants the maximal benefit П
defined as the difference between the sales income (VAT and excise tax not
included) and the production and sales expenses as included in the produc-
tion cost. Dividing, as usual, the production expenses into the variable de-
pending on production output Э and the conditionally constant Є and relat-
ing the variable expenses to the production output the objective function
may be represented as:

П = (ĝ – œ)Ə – Є = max, (12.1.1)

where ĝ means the finished product unit price, while œ – variable compo-
nent of unit cost.
Let us consider this expression together with the generalized relation-
ships between power and productivity of machines offered by the theory of
their similarity. Let us consider, in N – machine mean power, τ – machine
operating hours; ĝ – power release price, $/kWh; ĝ* – value close to fuel
cost component in total power cost, $/kWh). Representing the fuel cost
component ĝ* as the product of the fuel price þf ($/kg) by the fuel rate bт
(kg/kWh) and expressing the latter in terms of its minimal (theoretical)
value bтo (corresponding to ηN =1) as bт = bтo/ηN the objective function
may become:

П = (ĝ – bт þf) N τ – Є = max . (12.1.2)

Let us now express this objective function in terms of the machine load
B using the generalized relationships relating the machine power and effi-
ciency ηN with load criterion (11.4.8):

П = [(ĝ – bт þf)(1 +1/BФ)/(1– B)] Nmax B(1–B) – Є. (12.1.3)

Maximizing this expression as the load function B by equating its de-

rivative to zero gives after some transformations:

Bopt = 0,5(1 – ηNœ) . (12.1.4)

208
 where œ = bт þf /ĝ –fuel component share in the total electric power price.
Substituting (12.1.4) into (11.5.4) gives an expression for the machine
power economically most advantageous:

Nopt /Nmax = 1 – (ηN œ )2 (11.1.5)

According to this expres-

0.5 N /N 1.0
опт max sion, the less the fuel component of
0.4 0.8 power price (and more generally –
variable costs component in product

Relative Power
price) is and the less economical the
Relative Load

0.3 0.6
machine is, the closer the nominal
0.2 0.4 power approaches its peak value.
The relationships (12.1.4) and
0.1 0.2 (12.1.5) are more demonstrably plot-
ηN œ ted on Fig. 12.1. The complex ηN œ
0.0 0.0
0 0.2 0.4 0.6 0.8 1.0 values varying from zero to unit are
Fig. 12.1. Operating Conditions Economically
plotted here on the abscissa, while
Most. Advantageous for Power Plants the machine relative loads and pow-
ers economically most advantageous
are plotted on the ordinate. As fol-
lows from the figure, with œ → 0 or ηN → 0 the optimal load Nopt → Nmax.
This result is characteristic for machines on renewable energy sources (hy-
dropower stations, tidal, wave, wind, solar and geothermal power stations)
and complies with the efficiency criteria of finite-time thermodynamics. On
the contrary, with cost factors neglected (ст ≈ 0) quasi-static processes (В ≈
0) appear to be most advantageous. This result complies with the thermody-
namic approach where the machine efficiency is the only criterion. Ma-
chines with a considerable variable costs share occupy an intermediate posi-
tion. E.g., for thermal power stations with a power efficiency of ηN ≈ 0.45
and a fuel component in product price of œ ≈ 0.6 the optimal load Bopt is
equal to 0.36 and close to the nominal one corresponding to the utmost eco-
nomic efficiency of a machine with the Ф-factor Ф ≈ 4.5 (Fig. 11.2). This
result complies with the efficiency criterion of thermokinetics since for a
machine with a specified productivity (operating to a specified load curve)
the peak of benefit corresponds to operating conditions with minimal spe-
cific fuel rates bт = min. Thereby thermokinetics as if “lays a bridge” be-
tween classic thermodynamics, thermo-economics and finite-time thermo-
dynamics and allows to put further questions of optimal operating condi-
tions definition for power and processing plants of various types.

209
 12.2. Nominal Operating Conditions for Power Units

Including load in the number of factors to be optimized enabled deter-

mination that maximum of efficiency for power and processing plants of
various kinds was far from being always corresponding to their peak power.
In particular, for a big group of thermal power plants operating to a speci-
fied load curve Bopt corresponds to nominal operating conditions (with
minimal fuel rate). This demands the loads to be optimally distributed be-
tween the power units of a thermal power station. The solution to this rather
laborious problem is considerably facilitated due to the available universal
load curves for energy converting systems and due to the possibility to plot
on their basis individual load curves for various types of power units despite
a restricted scope of information. To do this, the data of full-scale test on all
power units or aggregates should be presently available. To obtain such
data, considerable time and costs are needed; therefore such tests are ex-
tremely seldom with the results often becoming obsolete yet in the period
between overhauls. Besides, the calculations of fuel rate increment per unit
power increase for any operating conditions of a machine the existing meth-
ods of load distribution are based on are so complicated and laborious that
their application is far from being always paid back. Therefore in practice a
known principle has to be followed, which reads that more economical ma-
chines should be loaded first of all. Nevertheless, this statement is far from
being always true. Let us assume there are two power units on a thermal
power station with the same installed power and good quality factors Ф =
4.7 and 3.0. Then according to criterion equation (12.4.8) the relative load
of the first power unit being increased from 0.2 to 0.3 will increase its
power efficiency ηN from 0.387 to 0.409, i.e. by 0.022, whereas for the less
economical unit with Ф = 3.0 this efficiency will increase from 0.30 to
0.331, i.e. by 0.031. Hence, the second power unit being loaded will provide
a higher both absolute and relative increase in economic efficiency of the
entire thermal power station and a lower increase in its fuel rate. In other
words, everything depends on the working point position on the universal
load curves of a machine. Fig. 11.2 visually demonstrates this statement.
E.g., the load of a machine operating with maximal ηN being increased, its
economic efficiency may even decrease, whereas the same for another ma-
chine operating under loads considerably inferior to nominal will cause
steep increase of its economic efficiency. In this context the visualization of
operation conditions for power system units in the form of an array of work-
ing points on the universal curve of the Fig.11.2 type facilitates the solution
to the problem posed since allows, subject to reasonable accuracy of the
prognostic estimates, cutting the expenses to conduct relevant tests and cal-
culations.

210
 To illustrate plotting and using individual load curves, let us consider an
arbitrary condensing steam turbine plant. The individual load curve for such
a machine can be plotted in linear approximation as based on the informa-
tion about its operation at one of the operating conditions with subsequent
reconstruction of the efficiency and power curves configuration on the basis
of criterion relationships or universal curves. A working point may be found
on Fig.14.2 by several ways. Firstly, this may be done based on the records
of the power unit mean load for some time period and of equivalent fuel
consumptions for the same period, which allows calculating the equivalent
fuel rates bт followed by determining the machine exergic efficiency ηN =
3600/bт þf in terms of the exergy eт known. The relative load B can be herein
found according to (12.1.4) as the ratio of generator currents at the present /
short circuit operating conditions. E.g., for the ТФ-series 63MW generators
the rated current and short-circuit current are 7,210A and 24,000A, respec-
tively, i.e. their relative load at nominal conditions is 0.3. Ultimately, a
working point can be also found by known quality factor Ф of the steam
turbine plant, which can be estimated as based on its recorded cycle parame-
ters. Let the steam turbine plant be operating, for example, at mean thermo-
dynamic temperatures of the heat well (fuel combustion products) and heat
sink (circulating water) equal to Th = 1,120K and Тc = 295K. If the mean
thermodynamic temperatures of heat application and heat abstraction in cy-
cle are in this case Т1 = 625K and Т2 = 305K, respectively, while the effec-
tive efficiency of the steam turbines ηoe = 0.87, then neglecting the feed-
water pumps operation the quality factor of the entire steam turbine plant
may be assessed in linear approximation by replacing the resistance rela-
tionship by the relationship between the adequate integral forces. As a result
of calculation the quality factor for the steam turbine plant under considera-
tion can be found according to (11.3.6) as Ф = 4.57. For this value the load
economically most advantageous is according to (11.5.1) Bn = 0.297, while
the peak value of the power efficiency ηN = 0.40. Thus rather simple calcu-
lations allow assessing the quality factor of an operating power plant and
plotting, as based on the quality factor, (in linear approximation) the total
load curve of this power plant without additional experiments. In that way
the similarity theory-based express methods channel the load optimal distri-
bution issue in the practical direction.

12.3. Optimal Overload Degree of Power and Processing Plants

We have considered above the case when the peak of economic effi-
ciency of power and processing plants was associated with their operation
under the relative loads either maximal or close to such. However, accord-
ing to Fig. 11.2, there is a rather wide spectrum of operating conditions in-
between those of maximal power and maximal economic efficiency usually

211
taken as nominal, which may be called overloading or boosting. Such oper-
ating conditions are quite characteristic for overland, air and space vehicles.
They occur in also stationary power plants intended for irregular load curve
operation. However, the choice of power augmentation optimum has not
been hitherto considered with due regard. In power engineering this is ex-
plained so that the nominal operating conditions of power units are close to
their limit power (according to the technical operation standards an electric
generator may be overloaded for several seconds to several minutes); in
other cases the interest in the problem is restricted because power units op-
erate according to an assigned load curve or because the thermodynamic
mathematical tool can not be applied to their analysis in finite-time format.
It can only add to the interest to consider these issues from the thermoki-
netic positions.
According to expression (12.1.4) the economically most advantageous
load Bopt at the utmost high fuel cost component œ = 1 is defined by exclu-
sively the efficiency of a machine and corresponds to minimal fuel rate.
Such loads are usually taken for the nominal Bn. For them the ratio Вopt/Вn is
evidently equal to unit (1). With œ decreasing at Ф = const (and, hence, at
Bn and ηex being invariable in the nominal operation) the optimal loads in-
crease, and the machine operation becomes economically justified for loads
exceeding those nominal, i.e. in the so-called augmentation operation. The
economically most advantageous augmentation level evaluated by the
Вopt/Вn ratio can be found from (12.1.4) and, for a higher degree of general-
ity, expressed as a function of the quality factor Ф. Such a relationship is il-
lustrated on Fig.12.2. According to the generalized graphs plotted therein
the economically justified overload degree for power and processing plants
increases with their quality factor. If, e.g., for steam turbine plants with the
quality factor Ф = 3...4 at the fuel cost component in the total power price œ
4.0 B /B
≈ 0.5 the economically most ad-
оpt n vantageous overload degree is
Ф = 50 1.25…1.35 (which means the
Optimal overload degree

3.0
power overload by only 4.1% at
2.5 ηN ≈ 0.4 and according to (12.1.4),
20
then the problem of using overload
2.0 operations is well actual for highly
10
economical processing plants.
3
1.5 Let us consider, for a specific
œ
example, an electrolysis plant to
1.0 extract hydrogen and oxygen from
0 0.2 0.4 0.6 0.8 1.0
water, which exergy analysis is
Fuel component in product price described by Nesterov in his work
(1980). The plant consists of an
Fig.12.2. Optimal Overload Degree for
Power Plants
electrolysis cells battery, gas-
liquid solution separation units,

212
electrolysis gas purification and dehydration units, heatexchangers and
pumps. The power supply to the plant includes the electric power Nel and
the flows of commercially suitable energy (exergy) with initial reagents Ехр.
The plant productivity in whole is defined by the amount of the hydrogen
and oxygen produced. Its energy (exergy) equivalent is construed as the
power Nj defined by the flows of commercially suitable energy of electroly-
sis products Ехpr and heat carriers Ехh. In the scheme under consideration
water is fed to the plant from outside, while gases are removed from the bat-
tery together with electrolyte. The plant power Nj = 1,500 W, battery volt-
age 1.95 V. The ambient air temperature and pressure are To = 298 K, po =
0.1013 MPa, respectively. To represent the operational conditions of such a
plant in terms of the similarity theory, let us consider the electrochemical
reactions running in the anode and cathode parts of the electrolysis ele-
ments:

4 OH- – 4 е = О2 + 2 H2O (Δϕ = +0.401 B), (12.3.1)

4 H2O + 4 е = 4 OH- (Δϕ = – 0.83 B), (12.3.2)

where Δϕ – potential difference across the proper electrode, e – designation

of electron as a system component.
As follows from (12.3.1) and (12.3.2), EMF of this reaction is equal to
1.23V, which corresponds to a battery power efficiency of ηN = 0.63 (a more
detailed calculation gives ηN = 0.621). In this case the voltage drop across
the battery internal resistance is 0.72 V, which corresponds to the relative
load В = 0.72/1.95 = 0.37. According to the universal load curves such pa-
rameters refer to a plant with the quality factor Ф ≈ 50. The optimal over-
load degree for this plant according to Fig.14-2 is 2.5…2.25 at œ ≈ 0.4 ÷
0.5. As a matter of fact, the plant operates with even higher overload. In-
deed, for a plant with Ф ≈ 50 according to the universal curves on Fig.13-2
the nominal operating conditions are associated with the relative load В ≈
0.12 and ηN = 0.77. Thus the plant under consideration actually operates
under a threefold overload relative to the most economically advantageous
operating conditions, which is economically justified subject to a moderate
price for electric power (œ ≈ 0.2). Hence, this plant has some reserves in
economic efficiency upgrade. Anyway, this example tells it is expedient to
modify the operating conditions of processing plants according to market-
determined prices for finish products and electric power. This fact taken into
consideration at market economy may become an extra incentive to main-
tain optimal operating conditions of equipment and to update the existing
power and processing plants in order to extend their overload capabilities
under the finished product deficiency.

213
 12.4. “Cruising Speed” Operation of Transport Vehicles

For transport vehicles of a wide spectrum the peak efficiency as per

condition (12.1.4) does not correspond any more to the peak power of their
propulsion systems. The point is that the power augmentation entails in-
creasing the fuel rate and the fuel load required to deliver a cargo in self-
sufficient sea or air freight, which results in the adequate reduction of the
net load being carried, i.e. in the drop of the vehicle capacity. The reduc-
tion in income from vehicle running thus caused can be accounted for in
(12.1.2) by adequately increasing the costs for transportation of fuel itself.
In this case in (12.1.2) along with the term bтþf characterizing the fuel
component in the variable costs an additional cost summand will appear
proportional to the transportation power inputs with a proportionality factor
equal to the unit fuel transportation cost þtr:

П = [ĝ – bт(þf + þtr)] Nτ – Є = max . (12.4.1)

Due to this the optimal load on the vehicles’ propulsion systems defined
by expression (12.1.4) will decrease a little as compared with the peak
power operating conditions and will be now associated with the cruising
speed of a ship or aircraft corresponding to the minimal fuel rate and, hence,
to the minimal mass of the vehicle itself. Minimum mass of a transport ve-
hicle at the given power of its propulsion system will evidently provide the
transportation of the maximal cargo, i.e. corresponds to the maximal capac-
ity of the vehicle. Therefore, the problem of transport vehicle mass minimi-
zation is one of the most important design and operational targets. This is-
sue is of utmost importance for space vehicles where the net mass of the
cargo being transported often defines the range and duration ability of the
vehicle, i.e. the feasibility of particular space programs.
Let us consider for example the total mass minimization problem for a
space vehicle proceeding from the only requirement to deliver a specified
net cargo while moving in a specified path (Etkin, 1996). Such an approach
assumes the definition of the so-called characteristic velocity of the space
vehicle v as the sum of its velocity increment magnitudes Δvi at particular
stretches of its path. For space vehicles starting off from the satellite orbit
this velocity is usually counted off from the orbital velocity and defined by
the Tsiolkovsky’s formula:

v = w ln(Mk /Mo) , (12.4.2)

where Mk , Mo – initial and final mass of the flying vehicle (with propulsions
on and off), respectively; w – jet stream velocity assumed constant.

214
 Since Mo = Mд + Мп = Мк – Mт, relationship (12.4.2) gives:

Mт = (Mд + Мп )[exp(v/w) – 1] . (12.4.3)

Then the total mass of the space vehicle is equal to:

Мк = (Mд + Мп ) exp(v/w) . (12.4.4)

Now let us express Mд in terms of the specific (related to the jet stream
kinetic energy) mass of the propulsion system γд =2Mд/Gwс2. Then:

Мк = (γдGwс2/2) exp(v/w) + Мп exp(v/w) . (12.4.5)

Equating the derivative of this expression with respect to the velocity w

to zero under the assumption of v, G, γд and Мп being constant gives:

2w – v – Мп v/w2Gγд = 0 . (12.4.6)

Introducing for convenience the net cargo share ρп = Мп /Мдopt as the ra-
tio of the cargo mass Мп to the optimal mass of the propulsion system (at w
= wopt) and considering that 2Мпv/ Gγд w2opt = ρп v gives finally:

(w/v)опт = (1 + ρп)/2 или R уопт

д = (1 + ρп)v/2g, (12.4.7)

where Rsp = w/g – the so-called specific jet thrust (g – gravitational accel-
eration).
According to this equation the optimal specific thrust value Rspopt at v,
G, γд and Мп = const does not depend on the flight duration, propulsion sys-
tem efficiency and specific gravity and is defined by exclusively the flight
path (v value) and the cargo net weight. That resulted from a more general
problem definition, which did not assume the constancy of the propulsion
system power N, full thrust R and the specified flight duration τ and de-
manded a minimal scope of initial data. The solution to the problem in the
form of (12.4.1) and (12.4.2) has the same general character and is valid re-
gardless the type of the jet engine, its design and size features. At the same
time such an approach to choosing Rspopt leaves enough freedom to choose
the propulsion system power Np = Gw2/2, working medium consumption G
and propulsion full thrust R, and, hence, the active propulsion time τ. It is
significant that the value Rspopt defined by (12.4.2) is considerably lower
than that found under constancy of the jet propulsion full thrust R and the
specified active flight duration τ and considerably closer to the optimal rela-
tionship between the jet outflow and flight velocities (w/v)opt, which is equal
to unit and corresponds to the peak value of the jet engine power efficiency.

215
 Fig.12.3 illustrates the relationship according to expression (14.4.5) be-
tween the spacecraft mass Mк, net cargo mass Мп and the velocity ratio w/v.
For a better generality the values Mк and Мп are related therein to the space-
craft optimal mass Мко calculated from equation (19.4.5) in the absence of
the net cargo and, thus, with the optimal ratio (w/v)opt = ½. As follows from
the figure, the minimal ratio
5 Mк/Мко (optimal Rsp = w/g) with
Mk/Mko
the net cargo share increasing
Cargo spaceship relative velocity
4 shifts, as should be expected, to-
ρп = 1.0 ward higher values of the outflow
ρп = 0.6
3
ρп = 0.2
velocity. However, the deviation
ρп = 0 from the optimal ratio w/v results,
2 at the conditions accepted, to a
steeper increase of the Mк/Мко
1 value than at the constant thrust of
propulsions. The spacecraft mass
w/v increases especially sharply with
0 1 2 3 4 5 the w/v ratio reaching some mini-
Jet stream relative velocity
mal value, which evidences that the
Fig.12.3. Spacecraft Specific Thrust and application of engines with a spe-
Net Cargo Optimization cific thrust below the said value is
unpractical. E.g., for the “round
trip” flights to Venus or Mars (when the minimal characteristic velocity is
equal to 3.7-3.8 km/c (Yavorsky and others, 1970)) with the net cargo mass
Мп = Мдopt the application of a liquid-propellant rocket engine on compo-
nents of the “oxygen-gasoline” or “nitric acid-kerosene” types (with a spe-
cific thrust of 260-300hp) is practically excluded. Engines of other types
have to be used for these purposes, in particular, the nuclear-fission rockets
with specific thrusts of 750-800hp on hydrogen operation. The more general
relationships obtained here between the spacecraft mass and velocity pa-
rameters may be useful for also determining the ranges of required specific
and full thrusts, propulsion system type and power, as well as a number of
other parameters of a space rocket engine for particular space flights.

12.5. Peak Power Reach Conditions for Heat Power Plant

Classic thermodynamics is known to have posed and solved the prob-

lem of determining the most economical thermodynamic cycles (with
maximal thermal efficiency or heat transformation ratio) within a specified
temperature range. These are known to be the ideal Carnot cycles. Classic
thermodynamics characterizes with a conclusion that the maximal efficiency
of heat engines does not depend on the properties of the working medium
(its state equations) and is defined by solely the temperature interval within

216
which the cycle is realized. However, the duration of real thermodynamic
processes is finite, while their efficiency depends on not only the perfection
degree of the heat-into-work conversion, but on also the intensity of the heat
exchange between the working medium and the heat wells. Therefore the
real cycle efficiency is always overestimated. This fact poses the problem of
revealing the limiting capabilities of real processes considering not only the
temperatures of heat wells, but also the heat transfer coefficients, duration
of a cycle in whole and its particular stages, cycle-averaged power, etc. Cur-
son and Ahlborn were the first who proposed such a problem definition
(1975) having considered the conditions of developing a Carnot cycle pro-
viding the peak cycle-averaged power N at temperatures of heat well and
heat sink of Th and Tc, respectively. In this definition the working medium
receives and releases heat at the constant temperatures Т1 < Тh and Т2 > Тc.
(Сurson, Ahlborn, 1975). The optimal values of these temperatures are to be
found here corresponding to the cycle peak power at a contact duration be-
tween the working medium and the heat well and heat sink of th and tc, re-
spectively, when the heat flows in-between them obey a linear law of the
type:

Jq' = Lh (Th – T1 ) ; Jq" = Lc (T2 – Tc ) , (12.5.1)

where Lh , Lc – constant heat transfer coefficients.

Settling the values of the parameters th and tc as satisfying the limita-
tions Jq'/T1+Jq"/T2 = 0 (ensuing from the entropy variation cyclic character)
the authors express the power N = (Jq''th + Jq"tc)/t averaged thru the cycle
duration th + tc= t as a function of above parameters and find the peak of this
function allowing for the said limitations. In this case the thermal efficiency
corresponding to the cycle peak power Nmax is defined by the relation:

ηt opt = 1 – Tc / Th . (12.5.2)

As can be seen, the optimal thermal efficiency of such a cycle is consid-

erably lower than the ideal Carnot cycle efficiency at the specified values
Lh, Lc of heat transfer coefficients and the peak power.
Many works dedicated to thermodynamics of finite-time processes are
characteristic by the fact that they consider limitations of various kinds, viz.
for working medium minimal and maximal volumes (Curson, 1975; Rubin,
1980), for piston acceleration in internal combustion engine (Band, Kafry,
1981), for heat amount Qh received from a heat well (Rubin, 1979; Band,
1982), etc. A great number of works in this field are dedicated to investiga-
tion of systems where the heat flow from a heat well to the working medium
is considered as given in the form of some time function Jq(t) (Salamon,
1981; Fairen, 1982). It is worth noticing that the first overview of works

217
dedicated to finite-time processes thermodynamics (Barrer, 1982) was pub-
lished in France as early as in 1980. It is mentioned therein that taking the
time factor into consideration when optimizing the cycles of heat engines al-
lows estimating the efficiency of real plants relating their parameters to the
limiting parameters of irreversible cycles having the same power, heat trans-
fer coefficients, duration, etc. Thus the finite-time thermodynamics reveals
the limiting capabilities of energy conversion irreversible processes in heat
engines at finite duration of the heat application and abstraction processes
(and, consequently, finite duration of the cycle), as well as under various
limitations to the plant parameters. Its methods are starting to be success-
fully applied to estimate the efficiency of not heat engines only, but also
processing plants, in particular, the photovoltaic cell-based plants and the
plants of water decomposition into oxygen and hydrogen (Dung, 1982). At-
tempts are known to optimize by this method the chemical efficiency of tu-
bular reactor (Ondrechem, 1980), as well as the output power of pulsating
laser (Keren, 1982) and the installation using the radiant energy (Adler,
1981; Мozurkewich, 1983). Thus the application sphere of finite- time
thermodynamics is rapidly widening (Linden, 1992). This is promoted by
not only the need to clarify the limiting capabilities of thermodynamic proc-
esses. Posing the problem in this theory changes the efficiency criteria for
heat engines and processing plants bringing them closer to reality. At the
same time it is worth noticing that finite-time thermodynamics considers
only the so-called externally irreversible cycles where the cycle-comprising
processes themselves are considered equilibrium (internally reversible).
Such inconsequence is explained by the fact that finite-time thermodynam-
ics imports the transfer equations involving the process rate from outside,
i.e. from the heat exchange theory. Process time and productivity do not
now, as before, enter in the equations of finite-time thermodynamics which
is thus incapable to describe the kinetics of the thermodynamic systems’
transfer from some non-equilibrium state into another (i.e. the motion of
such systems). Furthermore, this theory is based on solely the cycle method
and is therefore inapplicable to the analysis of the heat energy direct conver-
sion plants, as well as to the flow-type machines. Besides, it does not con-
sider the kinetics of energy-conversion process itself and assumes that the
duration of arbitrary cycle in finite-time thermodynamics is added of only
the times of heat application and abstraction. In the absence of unanimous
approaches to the energy dissipation in the energy transfer and conversion
processes the circle of problems solvable by the methods of finite-time
thermodynamics appears to be considerably limited. In this respect the syn-
thesis of thermodynamics and economics – “thermo-economics” – offers a
noticeable advantage. This direction in thermodynamics of the XXth century
is aimed to commensuration of savings in fuel and material due to plant ef-
ficiency upgrade and the investments associated. At the same time it should
be noted that the irreversibility of useful work processes in thermo-

218
economics is also considered not on the basis of their kinetics study, but by
introducing constant relative efficiencies of these processes, i.e. without the
relationship between power and cost effectiveness of engineering systems
taken into consideration. From here the problem ensues to directly introduce
the factor of real process productivity (useful power) into the equations of
thermodynamics as a thermodynamic variable along with the process time
and rate. This problem has been solved in thermokinetics that does not ex-
clude from the consideration the irreversible part of a real phenomenon (like
in Thomson’s pseudo-thermostatics) or its reversible part (like in Onsager’s
quasi-thermodynamics) and may be applied to investigation of processes
with any degree of irreversibility.
As shown above, in search of conditions to reach the peak cycle power
finite-time thermodynamics proceeds from the ideas usual for thermostatics
about the possibility to vary the durations of heat application Δt1 and ab-
straction Δt2 processes in cycles and to include them in the number of the
parameters being optimized (Curson, Ahlborn, 1975). In this case the opti-
mal duration of contact between the working medium and the heat well and
sink appears to be dependent on the heat transfer coefficients. However,
these ideas appear to be invalid when considering a steady-state energy
conversion process in a thermal power plant. As shown above, the availabil-
ity of a “transit” (crossing the system invariable) flow of corresponding en-
ergy carrier is the necessary condition of useful work done. For heat engines
this is expressed as the equality of total entropy flows entering and leaving
the engine Js' = – Js" (11.2.7). Comparing this condition with (11.2.6) and
taking into consideration that for any cycle ΔS1 = – ΔS2 drives to a conclu-
sion that the durations the working medium contacts with the heat well and
heat sink are equal to each other:

Δt1 = ΔS1 /Js΄ = ΔS2/Js˝ = Δt2 . (12.5.3 )

Thus, unlike finite-time thermodynamics, thermokinetics introduces the

additional condition Δt1 =Δt2. This provides such a result that the tempera-
tures of heat application T1 and abstraction T2 appear to be interrelated. To
reveal this relationship, let us consider the identity of the mean thermody-
namic heat application and abstraction temperatures T1 and T2 to the mean
integral temperatures of the heat flow entering and leaving the plant Т′ and
Т″, which is ascertained in the previous chapter. This allows expressing the
thermal efficiency of an arbitrary cycle in terms of the heat flow parameters:

ηt = 1 – Т̄ 2 /Т̄ 1= 1 – Т ˝/ Т΄ = 1 – Jq˝/Jq΄. (12.5.4 )

Representing on this basis the phenomenological laws of heat transfer

between the heat well and heat sink with temperatures of Th and Tc and the
working medium in the cycle as:

219
 Тh – Т1 = Rh Jq' ; Т2 – Тc = Rc Jq" . (12.5.5)

where Rh , Rc – overall thermal resistance to heat exchange from the heat

well and heat sink, respectively, easily gives that the temperature drops
from the heat well and heat sink are related with the thermal resistances thru
the relationship:

(Т2 – Тc )/(Тh – Т1 ) = (1– ηt ) Rc /Rh. (12.5.6)

Expanding the expression of ηt a relationship between the temperatures

of heat application and abstraction can be easily obtained (V. Etkin, 1991):

Т2 = Т1Тc /[Т1 (1 + Rc /Rh) –ТhRc /Rh ]. (12.5.7)

This relationship existing demands a correction of the conditions under

which the peak cycle power can be obtained. Let us consider for example
the problem of determining the optimal (at peak power conditions) heat ap-
plication temperature in cycles of a nuclear power plant (NPP) on saturated
steam (Calafati, 1963). Let the heat pick-up from a light-water nuclear reac-
tor Jq' be restricted by the temperature Tmax of the fuel elements. Then the
theoretical power of the plant Nj defined as the product of the heat flow Jq'
from the heat well by the thermal efficiency of the NPP cycle ηt = 1– Т2/Т1,
where Т1, Т2 – mean integral temperatures of heat application and abstrac-
tion, respectively, in the NPP cycle, will have a peak due to efficiency dete-
rioration as the heat flow from the heat well Jq' increases. Let us assume af-
ter the work mentioned that the temperature averaged over the heat ex-
change surface is identical to the mean integral temperature T1 of heat appli-
cation to the working medium (so that Jq' = (Тh –Т1)/Rh, where Тh, Rh – mean
integral temperature on the surface of the fuel elements and thermal resis-
tance to the heat transfer from the heat well to the working medium, respec-
tively). Let us also assume that the coefficient of heat release and tempera-
ture distribution variation over fuel element height kp = (Тh – Т1)/(Тmax – Т1)
does not depend on power. Then the NPP power will be:

Nj = Jq'ηt = kp (Tmax – Т1) (1 – Т2/Т1)/Rh . (12.5.8)

Considering this expression as a function of the initial cycle temperature

T1 and allowing for relation equation (12.5.7) after a number of transforma-
tions with constant Th and Tc gives the optimal heat application and abstrac-
tion temperatures in peak power cycles (Etkin, 1997):

Т1opt = ( Th Tc + ТhRc /Rh)/(1 + Rc /Rh) , (12.5.9)

220
 Т2opt = (Tc + Th Tc Rc /Rh)/(1 + Rc /Rh) . (12.5.10)

These expressions differ from those obtained in finite-time thermody-

namics in the degree of the Rc /Rh contagion. It may be easily found that this
ratio features the interrelation between the heat application and abstraction
temperatures in an arbitrary cycle. In fact, considering T2 as a function of T1
and differentiating (12.5.10) under the Rc/Rh constancy with ηtopt independ-
ent on this ratio gives:

∂Т2 /∂Т1 = – Rc/Rh (1– ηt ) . (12.5.11)

From this it follows that the relation degree between the heat application
and abstraction temperatures T1 and T2 is defined by the Rc/Rh ratio. To il-
lustrate how it is important to take into account the relation between initial
and final parameters for the power plant optimization, let us consider a spe-
cific example of determining the optimal heat application and abstraction
temperatures in NPP cycles with water-cooled reactors, for which, due to
the minor ст value, the peak power operating conditions are most economi-
cally advantageous (Calafati, 1963). However, unlike the traditional ap-
proach to determination of the NPP peak power cycle initial parameters
based on the usual assumption of cycle heat abstraction temperature T2 con-
stancy, we will proceed from expression (12.5.11). In the calculations fol-
lowing the author mentioned the maximal temperature of the fuel elements
made of metallic uranium is adopted as Tmax= 928 К (650оС) and Т2 =302 К,
which according to (12.5.2) gives the value Т1opt = TmaxT2 = (928⋅302)0,5 =
528 К. Like in the work mentioned, it is assumed in the calculations that the
coefficient of temperature distribution variation over fuel element height
with cosine heat release distribution kp =2/π does not depend on power,
which corresponds to the mean integral temperature of the fuel elements as
the heat well Тh = 782.8 К. The results of such calculations for various Rc/Rh
ratios are shown on Fig.12.4.
As follows from the figure, the optimal heat abstraction temperature
considerably exceeds its usually adopted level already at enough low values
of Rc/Rh ≈ 0.05...0.08 typical for NPP. As the Rc /Rh ratio increases, the gap
becomes especially noticeable. This fact evidences the necessity to jointly
assign the initial and final cycle parameters of a power unit considering their
relation. This interrelation becomes especially important for spacecraft
power units where the heat abstraction into the ambient space is provided by
radiation, while the Rc/Rh ratio increases by almost two orders. To confirm
the said, let us consider a particular example of the American orbital station
elaborated by General Electric (Favorsky and others, 1970). Its steam tur-
bine power unit comprises a nuclear reactor with a thermal power of NT = 8

221
MW, a steam turbine in complete with a generator with a net electric power
of Ne = 1 MW and a finned-tube radiator cooler. The power unit features
two main liquid-metal circuits. Liquid lithium is the working medium of the
first circuit, while potassium – of the second one. The temperature of the
potassium entering and leaving the turbine is 1,0650C and 704.40C, respec-
tively. The maximum allowable temperature of the fuel element is 1,2210C
and the temperature of the potassium entering and leaving the reactor is
1,037.8oC and 1,093.30C, respectively, which corresponds to the mean inte-
gral heat application temperature
°С opt opt
T1 , T2 thru cycle Т̄ 1 ≈ 1,324 К. The cy-
700
Th cle thermal efficiency ηt corre-
sponding to the heat abstraction
600 temperature Т̄ 2 = 977.5 К is
T1opt
equal to 0.26. The equilibrium
500 temperature of the radiator cooler
Tc (in the absence of Q2) found
400
T2 opt from the heat balance between
the solar radiation flow onto the
300 radiator cooler surface and the
Tc
Rc/Rh heat abstraction from the radiator
0.0 0.2 0.4 0.6 0.8 1.0
cooler into the ambient space by
Thermal resistance ratio radiation is tentatively 750 K. At
opt opt
Relation between T1 and T2 considered these conditions the Rc/Rh ratio
The above relation not considered found from expressions (12.5.6)
is equal to 2.1–2.2, which ex-
Fig.12.4. Interrelation between Heat Applica- ceeds by two orders its value
tion and Heat Abstraction Temperatures in
NPP Cycles with Water-Cooled Reactors usual for ground thermal power
plants. Optimal heat application
and abstraction temperatures
found for these parameter values from (12.5.6)–( 12.5.7) are Т̄ 1opt = 1350 К
and Т̄ 2 opt = 960 К, respectively – quite close to their true values. At the same
time it is noticeable that the plant cycle heat abstraction temperature appears
to be 2100C higher than the conditional temperature of the sink. This
stresses the necessity of interrelated search of optimal heat application and
abstraction temperatures in the cycles of power units.
The examples considered herein evidence that thermokinetics substan-
tially expands the applicability of finite-time thermodynamics allowing not
only to extend its methods to systems with various value of optimal load,
but also to clarify a number of its deductions.

222
 Chapter 13
CONVERTERS OF FIELD FORMS OF ENERGY

“…It is just a matter of time,

how soon mankind will succeed in
connecting its machines to the very
power source of the ambient space”
N. Tesla

Matter is known to exist in two forms, viz. substance and field. Up to

date mankind have used the energy of only the former of them. Such are, in
particular, the chemical energy of fuel and the nuclear energy of spontane-
ously fissionable elements. The final product of the substance energy con-
version is in most cases a substance in its modified state, which is accumu-
lated on the planet directly endangering its ecological stability. The problem
is aggravated by the concentration of population within huge megapolises
and by increasing consumption of energy resources. The energy generation
has become increasingly centralized, hydropower dams and thermal power
plants have become increasingly gigantic, oil and gas pipelines, as well as
power transmission and heat supply lines – increasingly extended. This just
facilitates an ecology crisis and aggravates the consequences of natural dis-
asters. Fossil fuel resources are being depleted, while the share of renewable
energy sources remains extremely low. In search of new energy sources the
scientific community takes the ever growing risk postponing for an uncer-
tain time the solution of the problem of nuclear waste disposal and conser-
vation of the exhausted nuclear plants. Vast funds are being spent to harness
the thermonuclear reactions. It seems as though there is no a way out of the
situation.
Meantime, in the present circumstances mankind ought to pay closer at-
tention to such alternative forms of renewable energy as the energy of the
field forces surrounding us.

13.1. Theoretical Possibilities of Creating Alternators

According to information received to date there are more than couple of

tens of devices in operation with an output exceeding the measurable input,
and more than a hundred of patents granted for such devices in different
countries. Such devices are often called the “free energy generators”, “over-
unity devices” (with efficiency in excess of 100%), “surplus power genera-

223
tors” and even “perpetual motion machines”. The authors of such publica-
tions are not willing for various reasons to recognize the energy sources not
estimable by present-day means. It comes down even to the fact that the
ambient energy consumption is neglected just due to the absence of…the
tangible costs associated. This can not be justified by whatever considera-
tions, the more so because all renewable energy sources feature low ex-
penses.
Since the physical content of these terms absolutely violates energy
conservation, it should be rather construed that such devices would use un-
accounted energy sources alternative to not only conventional organic and
nuclear fuels, but also to renewable energy sources. Therefore they will
hereafter be termed alternators 1) for short. However, the authors of publica-
tions for various reasons do not wish to recognize as energy sources those
inestimable by modern means. It comes down to even disregarding the en-
ergy from unknown sources because of unavailability… of the tangible
costs associated.
At these conditions it looks like reasonable that alternators be analyzed
via application of thermokinetics which body of mathematics, unlike the cy-
cle method of classic thermodynamics, operates exclusively parameters of
energy converting system itself rather than energy (heat) sources. It is just
apropos for alternators since the energy sources and their streams, as well as
their parameters, remain unknown in most cases.
According to thermokinetics, for thermodynamic analysis of an arbi-
trary machine converting some ith field form of energy into the jth form, the
law of energy conservation (2.3.9) should be used, which for the case of
vector forces acting gives:

dU = đQд – Хi·dZi – Хj·dZj = 0. (13.1.1 )

This equation differs from the Gibbs’ generalized relationship the clas-
sic thermodynamics operates since it does not contain the terms ΣiψidΘi de-
scribing transfer of the energy carrier Θi across the system borders and, on
the contrary, contains the term đWj = Хj·dZj describing the work done by the
force field and working medium itself. In the particular case of electrical
and magnetic fields the parameters Хi and Zi refer to the working medium,
while the parameters Хj and Zj – to the object the work applies to. The terms
Хi·dZi < 0 and Хj·dZj > 0 characterize here, respectively, the work done on
the working medium by the force field and the work done by the working
medium itself on the object the work applies to; the term đQd describes the

1)
The term “alternators” refer presently to the only class of the “over-unity devices”, viz.
magnetic motors and generators. We will refer to it in the wider sense.

224
elementary dissipation heat released for the time t within the system volume
V. As can be seen, to analyze the cycle involving a non-equilibrium working
medium, there is no need to know the parameters of the energy source. This
makes it possible to analyze the processes running in energy converter
without knowing the parameters of the energy source. Such an analysis, as
shown in Chapters 13 and 14, includes finding the motive forces Хi, Хj and
the energy carrier flows Ji, Jj for energy forms being and having been con-
verted, formulating the laws of energy conversion (11.4.1–11.4.2) in a par-
ticular machine, finding the similarity criteria (11.4.6–11.4.7), determining
the power efficiency of the machine and revealing the deviation of its opera-
tion from optimum for this particular type of energy converters. These prob-
lems outstep the classic theory of heat engines.
Unfortunately, such an analysis does not give a unique indication of en-
ergy source. However, as shown in Chapter 14, with regard to renewable en-
ergy sources for which investments do not noticeably contribute to the net
cost of the energy generated, their consumption is not of fundamental impor-
tance. Therefore the alternators are to be considered as ranked with other ma-
chines operating on renewable energy sources. However, the fact should not
be left out of consideration that a field form of energy, from the positions of
thermokinetics, is caused by heterogeneous distribution of “field-forming”
material objects in space and, therefore, belongs to them, but not to space it-
self (ether or physical vacuum), whatever properties we may try to attribute to
this space. It remains to show that whatever “free energy generation”, “sur-
plus power” and “over-unity efficiency” are out of the question. To do so, the
equation (15.1.1) is easier to be represented in the form of the energy balance
equation

dU/dt = Ni – Nj – Nd. (13.1.2)

Here Ni means the power input to the alternator; Nj – net power of the
device; Nd – dissipation power in the system. From here the power effi-
ciency expression ensues, which for cyclic machines (dU/dt = 0) according
to (11.2.9) takes the form:

ηN = Nj /Ni = 1 – Nd/Ni ≤ 1 . (13.1.3)

Thus the efficiency of any real energy converter (with Nd/Ni > 0) is
always below unity, i.e. the work of alternators is always done in
strict compliance with the laws of thermodynamics and without viola-
tion of its first law.
It should be also clearly understood that using a field form of energy in
order to convert it into other (more convenient for user) forms is possible
inasmuch as it is ordered. Let us now show that a field form of energy meets

225
this requirement, too, despite the known expression for the wave energy Uw
valid for waves of any character (acoustic, hydraulic, electromagnetic) and
not containing vector parameters:

Uw = ρAw2ω2/2 . (13.1.4)

Here ρ – oscillating medium density; Aw – oscillation amplitude; ω – oscilla-

tion angular frequency.
Let us show that any wave with an amplitude of A as a bilateral devia-
tion of the oscillating value Θw from its equilibrium value is characterized
by some distribution moment Zw. To do so, we have to consider a stretch
out of an arbitrary wave sequence with the oscillating value monotone in-
creasing or decreasing (in our case, the density ρw of the parameter Θw (Fig.
13.1)).
In this figure the crosshatched regions feature the part Θw* of the value
Θw transferred from some domain of space into another. As a result, the cen-
ters of the value Θw* for each quarter-wave part λ/4 shift from the positions
rо' and rо" they would occupy at uniform
ρw A distribution of ρw for quarter-period to the
positions r' and r". The distribution mo-
ment Zw is defined in this case by the
r same expression (1.5.7), but with the
ro elements dΘi replaced by Θw*:
ro r
r Zw = r' Θw*' + r"Θw*". (13.1.5)
λ/4
Since Θw*' = – Θw*", expression
-A (13.1.5) may be represented in the same
form as the dipole moment:
Fig.13.1. To distribution moment
generating in single wave
Zw = Θw*∆r , (13.1.6)

where ∆r = r" - r' means the dipole moment arm. It is significant that, like
in the case of dipole, the terms of (15.1.5) are not mutually neutralized, but
summarized since the signs of (r'- r'о), (r"- r"о) are opposite like for Θw*',
Θw*". Thus the oscillation of whatever medium may be considered as a kind
of its polarization. This allows finding the motive force Хw to the process of
radiant energy exchange by the same way, i.e. as the derivative of the wave
energy Uw with respect to the distribution moment Zw, i.e. Хw = -
(∂Uw/∂Zw). However, in this case it is even easier to use expression (2.5.2)
giving:

dUw = – Θwdψw = – ρAω dAω, (13.1.7)

226
 From this it directly follows that the “wave potential” is equal to the
product of amplitude and wave frequency: ψw = Aω, while the motive force
of oscillation energy exchange Хw = – ∇Aω, i.e. is expressed in terms of
negative gradient of this potential which we will hereinafter refer to as “am-
plitude-frequency”. As can be seen, this type of energy exchange is in the
general case directed toward equalizing amplitudes and wavelengths. It oc-
curs with a disturbance of matter–field equilibrium, which manifests itself
in any disturbance of amplitude balance or in frequency shift between ex-
ternal and natural oscillation processes. This refers, in particular, to radiant
energy exchange which, according to existing concepts, features the elec-
tromagnetic character. This energy exchange is by no means reducible to
heat exchange that occupies merely a minor part in the spectrum of electro-
magnetic oscillation with a wavelength λ of 0.4 to 4 μm 1)1) and has its own
motive force defined for absolute black bodies by the difference between
the forth powers of absolute temperatures.
The energy exchange arising between matter and field explains the
functioning of alternators without recourse to whatever hypotheses and pos-
tulates lying beyond the existing scientific paradigm. Such a position basi-
cally differs from the opinion common for both the traditional academic
ambience and alternative scientific forums which declare that, to substanti-
ate the principle of operation and to develop the “over-unity” devices (with
efficiency > 1), some novel “exotic” branch of physics is needed.
The further thermokinetic analysis of the alternators is advisably to be
conducted in application to a particular type of working medium that pro-
vides the energy conversion in some of the stationary force fields. Hereafter
we will consider examples of the energy converters for the force fields
known, viz. gravitational, electrostatic and electromagnetic. Such a consid-
eration is basically aimed at demonstrating the fact that the alternators fol-
low the unitary thermokinetic laws of energy conversion.

13.2. Gravitational energy Converters

The idea to create the “perpetuum mobile” (perpetual motion) technical

devices is rooted in philosophy of the Ancient Orient. This is the place
wherein the first documentary evidences about the creation of a gravitation-
based engine originate. Such is, in particular, the Bakhaskar “self-rotating
wheel” (India) with tubes tangentially arranged and half-filled with water,
which demonstrated in mid 12th century. In Europe the first projects of such

1)
Photosynthesis, photo-effect, photo-ionization, photoluminescence, photo-acoustic
phenomena, photonuclear reactions, etc, involving radiant energy exchange – all those
are evidence of the fact.

227
mechanical “perpetual motion machines” appeared in the 13th century: the
V. Honnecourt’s wheel with seven weights (France, 1245), the similar V.
Maricourt’s wheel (France, 1269). Leonardo da Vinci left a number of
drawings behind him with a device where vertically falling weights or water
rotated a wheel and thus worked efficiently. Mariano di Jacopo built in Italy
(1438) a system of eight rods radially arranged in the plane of rotation and
all bendable in one direction. Due to this fact the left half of the system dif-
fers from its right half in weight, thus, providing the rotation. Cornelius
Drebbel, alchemist and magician, created a perpetuum mobile in 1610 (pre-
sumably). Robert Fludd had developed a plenty of devices of such a type up
to 1630. Later on in 1870 many people tried to obtain patents for Fludd’s
device variations. Ulrich von Carnach developed a “ball” perpetuum mobile
in Germany in 1664. The famous scientist Jean Bernoully (1667-1748) put
forward a fluid-energy device project. Bockler designed a “self-rotating”
mill in 1686 based on the Archimedes’ screw in various versions.
In England the first perpetuum mobile patent was granted in 1635 for
the E. Somerset’s 4-meter wheel with 14 weights per 25 kg each. The ma-
chine was brilliantly tested in London in the presence of King Charles with
a record available in the archives. One of such wheels E. Somerset demon-
strated to the King of France in 1638. It is significant that already 600 pat-
ents for such devices had been granted in England up to 1903.
The wheel of J. Bessler (better known as Orffyreus) gained the most
distinction in the early 18th century. For several tens of years he publicly
demonstrated various models of his wheel set in motion with weights inside
it, which developed an asymmetrical moment at wheel rotation. The last of
his designs was 6 feet in diameter, 12 inches in thickness and rotated with a
speed of 42 rpm. That wheel not only kept continuous rotation, but also al-
lowed to do useful work, e.g. lifted a weight with a mass of 16 kg to a
height of 1.5 m. Its operation was more than once scrutinized by famous
scientists and officials, while German Prince Karl let into its design granted
Orffyreus a certificate for “Perpetuum Mobile”. Under control of a compe-
tent commission of 11 members (Professor Gravesaint, Newton’s intimate
friend, among them) the wheel kept working in a room locked and sealed up
for 40 days and at a sudden inspection showed as-before rotation with a
speed of 26 rpm. The tests repeated many times since the commission scru-
tinized the availability of secret drives (the reason wherefore the wheel
changed its location).
As one can see, those were not charlatans at all who were occupied with
perpetual motion machines. Therefore, it is hardly to credit of the Parisian
Academy of Sciences, one of the highest authority at that time, that it re-
solved in 1775 (i.e. before the energy conservation law discovered) to ex-
clude such projects from consideration. It is not out of place to note that still
earlier that Academy had refused to consider the existence of meteorites as
an absurdity of rockfall down from the heavens. It is well known what a

228
confusion it caused. It is quite possible that if the French academicians had
not disregarded a careful and unbiased consideration of such machines,
power engineering and science in whole could have followed quite other
path in their development. However, in reality an absolutely opposite situa-
tion developed: when an operating machine pretending to be the “perpetuum
mobile” was publicly demonstrated, no reasonable refutations followed, but
usually the case closed, as for the Orffyreus’ wheel, with a magazine pam-
phlet.
Assuming that the task of real science is not at all the devotion to a
paradigm having become antiquated long ago, let us consider, from the po-
sitions of thermokinetics, the operation of one of such devices, e.g. the
Fraga’s wheel (Cuba, USA Patent IL60915,1987).
This device constitutes a self-
rotating wheel with non-balanced
weights secured at the ends of levers
(Fig. 13.2). The opposite ends of the
levers accommodate movable joints
enabling the levers to “tilt” when tran-
siting the unstable equilibrium point.
The wheel has a number of cogs secur-
ing the weights with the levers till the
position of the maximum possible
torque (left part of the wheel). On the
Fig. 13.2. Fraga’s Wheel contrary, the weights in the right part of
the wheel are always in the position of
the minimum possible torque. The device is induced to stable rotation by a
minor push. For our case the basic equation of thermokinetics (2.2.5) has
the form:

dU = ТdS + Fg·dRg + M·dφ , (13.2.1)

where ТdS = đQ + đQд – total heat the system receives from outside (đQ)
and releases due to friction (đQд); Fg – gravity force; Rg– center-of-mass ra-
dius vector of the wheel; M=Fg×Rg – torque applied to the wheel from
gravitational field; φ – angle of wheel rotation.
Comparing this expression with the combined equation for the first and
second laws of classic thermodynamics (2.2.7) wherein the expansion work
pdV is superseded by other (the jth) kinds of work đWj

dU = ТdS – đWj , (13.2.2)

gives that, should a heterogeneous system have the internal energy of gravi-
tational interaction between system parts, the term đWj includes the weight
displacement work Fg·dRg and the wheel rotation work M·dφ.

229
 Let us apply this equation to the circular process the Fraga’s wheel runs:

∫ dU = ∫ ТdS + ∫ Fg·dRg + ∫ M·dφ . (13.2.3)

This expression can be essentially simplified given the circuital integral

of any state function, including system energy, equal to zero ( ∫ dU = 0) and
the friction and external heat exchange neglected ( ∫ ТdS = 0). Furthermore,
since the gravity force Fg remains essentially invariable in circular process,
then factoring Fg outside the integral sign and given ∫ dRg = 0 gives that the
second term on the right-hand side also becomes zero. Thus, from the posi-
tions of Newton’s mechanics not catering for rotational motion (in the ab-
sence of the term ∫ M·dφ), the Fraga’s wheel should be attributed to the
“perpetual motion of the first kind”. However, considering the third term on
the right-hand side of (13.1.3) gives immediately that the Fraga’s wheel cy-
cle work Wc = ∫ đWj is equal to

Wц = ∫ Mg·dφg . (13.2.4)

This expression shows that the cycle work of gravitational converter

may differ from zero if gravitational field does work for rotation of the
wheel. This work may differ from zero since the field of torques is not po-
tential. Let us now clarify the conditions making this work other than zero.
To do so, divide circuital integral (13.1.4) into parts 1–2 and 2–1 wherein
the angle ϕ varies from 0 up to 1800 and from 1800 down to 0. For the sake
of simplicity represent the integrand (13.1.4) in terms of scalar values of M
and φg and denote the torque and rotation angle for these parts of the cycle
with single or double stroke, respec- M 2
tively (M΄, φ΄ and M˝, φ˝). Then based
on (13.1.4) and given the opposite M˝
signs for the torques on these parts
(M˝= – M΄) one may write down in-
stead of (13.1.4): M΄

2
Wц = ∫ ( M ′ − M ′′ ) ⋅ d ϕg ' . (13.2.5) 1
1
0
π φ
The M–φg cycle of such a device Fig. 13. 3. Cycle of Unbalanced
Wheel in Gravity Field
for one of the weighted levers is plot-
ted in Fig. 13.3. Let us assume that at
initial time this weight takes the extremely lowest position (angle φ makes
230
00 to vertical). With an external force applied let us turn the wheel
counterclockwise at an angle of 1800
so that the weight arm Rg' remains
minimal (process 1–2). This action is
associated with an external work ex-
erted and expressed by the cross-
hatched area. With the apex position
reached a weight “throw-over” will
occur and the further rotation of the
wheel will proceed as a self-motion
(process 2–1). The self-motion work
expressed by the hatched area under
curve 2–1 will exceed the exerted
work due to increasing Rg” and the
corresponding average torque M”.
The weight lift along 1–2 will further
continue without an exerting interfer-
ence since the average counter-torque
M” will be lower. This is the princi-
ple of operation of the self-rotating
wheels which do not thus violate
whatever laws of physics and are not
the “perpetual motion” in the com-
mon sense of this term. It remains
only to show that the operation of
Fig. 13.4. A. Kosta’s Wheel such a device follows the same uni-
versal load curves as the other en-
gines. In our case the forces Хi and Хiо are construed as the moments M =
gRg, so the load criterion take the form В = 1 – Rg΄/ Rg˝ < 1. As for the Q-
factor Ф, it remains unknown. However, given the wheel friction, it is cer-
tainly below infinity, so from the curve in Fig. 14-2 it follows that the power
efficiency ηN and the engine power N become zero at the balance of arms
Rg΄= Rg˝ which is in strict compliance with experiment. From this it also fol-
lows that the device should be tuned to find the optimal relation Rg΄/ Rg˝–
the fact as well noted by the inventors of these devices. It remains only just
to instantiate the above with several present-day designs of such devices.
One of the most impressive machines of such a kind is the device (Fig. 13.4)
A Kosta built (Patent France No95/12421, 1995). His 18-meter unbalanced
wheel is the best wordless evidence of the inventor’s earnest approach and
confidence in integrity of the design he has created. This wheel operates to
the same principle and contains 236 movable elements providing its rota-
tion. It can rotate both clockwise and counterclockwise. As the inventor
notes, the main problem of its fabrication was to reach the variation of the
mass positions “in proper place and at proper time”.

231
 In 2006 the power company Environ
announced its readiness to bring to the
market a generator SPEGG producing
electric power without fuel consumption.
The generator is a self-rotating wheel
Warranlinc that is coupled to a current
generator and has 16 spokes. Eight of
them contain weights being in translation
movement and thereby compressing
springs. The wheel is designed so that it is
always heavier on one side. This disturbs
its equilibrium in gravity field forcing it to Fig. 13.5. Warannlinc Wheel
rotate. Thus the great idea to use inex-
haustible “living forces” of nature is starting to take real shape.

13.3. Generators Using Electrical Field energy

Let us now consider the generators using the energy of the natural elec-
trostatic field. The Earth and its ionosphere are known to constitute a
“spherical condenser” which energy content according to recent calculations
(E. Rauscher, 2002) is about 3·109 kWh
with a capacity of 15·103 μF. The intensity
of the electrostatic field it generates aver-
ages 100 V/m. This gives hope to mankind,
according to N. Tesla, “to connect its ma-
chines to the very power source of the am-
bient space”. One of the possible designs he
offered looks like an antenna in the form of
a metalized balloon lifted above the ground
and serving as an electric charge integrator.
Being connected to an energy converter
through a cable, this integrator is capable to
use the “gratis” energy of atmospheric elec-
tricity. J. Swenson has conducted a series of
simple experiments aimed at “energy recov-
Fig. 13.6. Swenson’s
Generator Schematic
ery from air”. Natural electrical oscillation
frequency of the planet, as known yet from
Tesla’s time, is 7.5Hz. Swenson dealt with a resonance frequency of 375
kHz and a 10-meter antenna (Fig. 13.6). This version of the energy recovery
from the ambient space has been implemented in the Efimenko’s electro-
static generator described in the book “Electrostatic Motor” (1973). A cy-
lindrical rotor rotates in the potential electrical field generating a power of
about 70W by means of a usual dynamo. The electrical field of the Earth

232
serves as a field source (6,000W), for which reason the machine has an an-
tenna and grounding. Such machines using the potential difference between
the planet surface and ionosphere have been known yet from 1800’s. In this
case not just posing a problem, but its solution is of interest.
In a number of cases the electrostatic fields can be generated artificially,
which has been proven by the energy generator Testatica invented by P.
Bauman and built in the Methernita Christian community. Several of such
fuel-free generators have produced for 30 years an electric power of 750kW
covering the entire community’s demands (including also the workshops lo-
cated in the same place). From the technical standpoint this device is the
Wimshurst modernized influence machine. The machine features two discs
with steel or aluminum segments fixed to the discs. The separation of the
charges on the disc segments occur due to the brush friction. The brushes
also pick up the charges on the disc segments. The self-rotation of the discs
is provided by, evidently, the mutual repulsion of the wheel segments rotat-
ing in the opposite directions due to delay of the electrostatic interaction
forces. The reaction force in this design can decrease by the relative dis-
placement of the brushes assuring decreasing the charge on the segments at
the moment the segments are drawing together. Such a machine with a disc
diameter of 20 cm generates a power of about 200W; the large-scale ma-
chine has discs with a diameter of 2 m and generates about 30kW of power.
A special diode module and Leyden jars provide frequency regulation. The
design also includes permanent horseshoe magnets with coils connected to
the Leyden jars and intended to maintain the frequency due to resonance, as
well as a device stepping the voltage from 100kV (and more) down to
200V. Using permanent horseshoe magnets in the modern converter version
is noted to considerably increase EMF.
In this aspect it is much more promising to use “electrodynamic” effects
associated with the pulsed operation of devices using natural electricity. Yet
N. Tesla, when conducted experiments with high-voltage DC generator,
spark discharger and high-voltage air-core transformer (Tesla transformer)
in the ninetieth of the 19th century, noticed a considerable increase in the
power the device released into the environment against the high-frequency
AC generator with a conventional step-up transformer. Tesla called that
power increment “radiant electricity” having attributed it to ether and op-
posed to the conventional electric current as an electron stream. That state-
ment was based on the effect Tesla discovered, which generated a voltage
spatially distributed. The latter grew along the secondary coil length of his
non-electromagnetic transformer and might thousands times exceed the
spark discharger initial voltage and not to be proportional to the number of
ampere-turns (since there was no current in the winding). That “electro-
radiant” effect appeared when the high-voltage direct current was discharg-
ing to spark gap and broke before the back (reverse) current appeared. The
effect intensified with a capacitor included in the high-voltage DC power

233
supply. The energy flux caused by that effect spread in the form of a longi-
tudinal electrostatic “light-like” beam that moved perpendicularly to the coil
turns like incompressible gas under pressure and did not penetrate into the
conductor. Fast interruption of direct current with magnetic interrupters
generated a shock wave in the laboratory wherein the Tesla transformer was
installed. This wave might be sensed as a sharp shock and an electric dis-
comfort (“tickle”) that, unlike the conventional electromagnetic radiation,
penetrated through metal shields and most of dielectrics. Nevertheless, the
pulses the Tesla transformer radiated, when shorter than 100 microseconds,
were absolutely harmless to people and did not cause heating. The magni-
tude of those pulses depended on their duration and on the spark discharger
voltage. They generated light effects in the vacuum tubes and caused a “re-
sponse” in metals in the form of electric charge accumulation.
Thomas Henry Moray (USA) was among the Tesla’s followers. In
1909–1910 he created a number of devices operating, as he believed, on
electrostatic charge of the Earth and ionosphere. One of such demonstra-
tional devices was represented to experts for review (with the exception of a
small box that the author preferred having kept in his breast pocket). The
device consisted of capacitors, step-up transformer, spark discharger and a
panel with two radio tubes, as well as of a permanent magnet, switchers and
lamps of 100W and 20W as a load. Whatever batteries were absent in the
device, however, the design included antennas with a length of up to 200
feet elevated about 80 feet and a grounding deepened about 7 feet. The de-
vice was tested for a time period lengthy enough at both home conditions
and in the mountains (far from power transmission lines). Some of the de-
vices developed a load power of up to 650W at a voltage of up to 250kV.
As can be seen, there are many elements in this device similar to the Tesla’s
technique.
Many of such elements can be revealed in the “over-unity” engine-
generator of Edwin Grey (USA). For a period of 1961–1986 he built and
patented several prototypes of self-sustained EMA devices (Electric Mag-
netic Association) capable to generate electric power without fuel used and
to supply the needs of an apartment house, automobile, train or plane. Re-
peated demos of his technique gained enthusiastic comments in the press,
while Grey himself was honored with a title of “Inventor of the Year”
(1979) and with the “Quality Certificate” from R. Reagan, who was the
Governor of California at that time. Like the Tesla’s device, the Grey’s gen-
erator was powered with high-voltage direct current. However, instead of
the high-voltage DC generator, a battery was used there with the output in-
terrupted through a multi-vibrator. The pulses from the multi-vibrator were
delivered to the primary coil of a conventional transformer with the high-
voltage winding connected to a bridge-type rectifier. Like in the Tesla’s de-
vice, the high-voltage DC power supply periodically charged a high-value
capacitor. The spark discharger was one more mutual component. However,

234
instead of the transformer, the Grey’s device used an “electrical conversion
switching element tube” which consisted of a resistor, spark discharger and
“charge receiving plates”, i.e. combined three components of the Tesla’s
device. It was the electrical conversion switching element tube that did work
when powered an engine, TV set, radio set, incandescent-filament lamps,
etc. The “cold electricity” circuit, as Grey called it, operated likely in the
following way: the storage battery voltage was stepped up to 3,000 V and
stored in a high-value capacitor. Then the pulses discharged through a spark
gap governed by electronic tubes so that the pulse duration was less than 50
microseconds. That intermittent pulse sequence passed through the conver-
sion switching element tube where was entrapped on the charge receiving
plates. The load was provided as a transformer that stepped the voltage
down to power the electronic tubes and other loads, as well as an additional
step-down transformer to charge a secondary battery. Having the batteries
periodically switched, Grey not only recovered the power supplied, but also
received an imposing “surplus” power. Nevertheless, all his attempts to ap-
ply his technique for commercial purposes failed. In the late 1970th the
company Zetec, Inc. bought up the Grey’s technology, while the inventor’s
attempts to intrigue the US Government and Senate in his work were not re-
sponded. Edwin Vincent Grey died in April 1989 in his workshop in Sparks
(Nevada) at the age of 64 under the mysterious circumstances.
However, ideas never die without leaving a trace, which is once again
confirmed by the Testatika energy generator invented by Paul Baumann and
built in the Methernita Christian community (Switzerland). Several of such
fuel-free generators have produced for 30 years an electric power of 750 kW
covering the entire community’s demands (including also the workshops lo-
cated in the same place). Like in the case of E. Grey, the idea was suggested
to P. Baumann through studying lightning. Unlike Tesla and Grey, to obtain
a high-voltage DC source, the Wimshurst (1832–1903) generator was used,
which application was practically ceased with electromagnetic generators
introduced. The Wimshurst generator consisted of two discs which rotated
in the opposite directions and had steel or aluminum segments fixed to
them. The separation of the charges on the disc segments occurred due to
the brush friction. The brushes
also picked up the charge on the
disc segments. The charge then
drained into the Leyden jars and
stored there. In the Testatika (see
Fig. 13.7) the high voltage from
these capacitors is supplied to the
top of large metallic jars (which
contents P. Baumann never
showed to anyone and which
Fig. 13.7. Testatica Generator caused a lot of guess-work – from

235
capacitors with uranium additives and to a fancy crystals–magnets combina-
tion) and then removed out of the bottom to spark dischargers. In the figure
two permanent horseshoe magnets can be seen along with the spark dis-
chargers. Thus the Baumann’s generator uses the same elements to enhance
spark formation and arc breaking as in the Tesla’s machine.
The self-rotation of the discs after an initial impact is provided by the
mutual repulsion of the segments in two wheels due to the forces of elec-
tromagnetic interaction since the rotating electrified disc of the generator
forms together with the external circuit a high current loop where the cur-
rent runs in the discs in the opposite directions. Electrostatic repulsive
forces may also take part in the process. The rotation is adjusted by a rela-
tive displacement of the brushes. A prototype of such a machine with a disc
diameter of 20 cm generated a power of about 200 W. The large-scale Tes-
tatika machines have discs with a diameter of 2 m and generate about 30
kW of power. A special diode module and the Leyden jars provide fre-
quency regulation. The design also includes a device stepping the voltage
from 100 kV (and more) down to 200 V.
The fact the machine exists and operates is confirmed by the reports
available from 12 recognized scientists who visited the community at differ-
ent times to investigate and check the Testatika integrity (including Prof S.
Marinov, who built working models of the generator and finally perished
under the mysterious circumstances). All of these reports come to the rec-
ognition of the fact that the principles of Testatika’s operation remain ob-
scure.
The William Hyde’s device (Patent USA No4897592, 1987) is one of
the operating up-to-date designs close to the Baumann’s machine in its
technical solution. The author called it a “system generating power from
electric field”. It includes conventional elements of an electrostatic genera-
tor using rotating discs with segments like in the Swiss M-L converter Tes-
tatika. In the prototype of 1987 Hyde used up to 240 rotor segments and 480
stator segments. His generator rotates with a speed by an order of magnitude
exceeding the Testatik speed. Furthermore, Hyde supplemented his design
with several new elements, viz. stator discs on electrode plates, external 3
kV-power supply charging these electrode plates, etc. With such a potential
the pulse voltage on the stator reaches 300 kV. The double rotor discs in his
machine rotate in the same direction. Due to this fact the paired stator seg-
ments appear to be periodically screened against the polarizing effect of the
exciter. Each stator segment in the machine is electrically coupled with its
pair through a circuit where the pulse voltage is reduced and rectifies for the
output circuit. To accelerate the machine rotor, Hyde uses the potential elec-
trostatic field on that path of the process where the field work is positive.
Where the field decelerates the rotor, Hyde partially shields it. So an unbal-
ance of the forces Fi' and Fi" is created demanded by thermokinetics and al-
lowing to obtain the energy from the stationary field source. The generator

236
output is 22.9 kW with an input of 2.4 kW and a total load power of 20.5
kW. Due to its features, as well as to its description available, the self-
rotating electrostatic Hyde’s generator is one of the most attractive alterna-
tor designs.
One of the recent domestic designs is worthy of notice – the hydromag-
netic dynamo by O. Gritskevich – patented yet in 1988 in the USSR as a
“method of generation and electrostatic plasma generator as its embodi-
ment”. Considering our planet as a huge electrostatic generator the inventor
offered a method to convert the force field of the planet into useful energy.
The method was based on known physical principles, but used unique de-
sign solutions. The “non-mechanical” dynamo by Gritskevich was config-
ured as doughnut-shaped bread roll with water as the rotor. Those genera-
tors agnominated as “water-warmers” had no pumps and none of the rub-
bing parts. They were operable for years practically without maintenance
and not consuming even a gram of fuel. At the same time they were com-
pact to be accommodated in any house and even a car. The cost of energy
they generated was fortyfold lower than by a nuclear power plant, twenty-
fold lower than by a thermal power plant and fourfold lower than by a wind
motor. The first prototype successfully operated in the mountains of Arme-
nia and supplied the on-site needs of a scientific research station. However,
despite the Rospatent certification and approval of scientific community, the
hydromagnetic dynamo by Gritskevich has not found the market demand
neither in Russia nor in the USA, where the inventor and seven of his mates
from the design bureau emigrated in the late 1999. This circumstance evi-
dences ones again that the main point here is not only the scientific consis-
tency of the alternators pertaining to this class and even not the search of
their most promising design solutions, but the geopolitics of the energy park
holders.

13.4. Current Generators Using Magnetic Field energy

The current generators based on permanent magnets constitute the most

numerous and diversified class among the devices attributed to the “perpet-
ual motion” category. The capability of the permanent magnets to do useful
work (e.g. to lift metallic objects) has been known since several centuries
ago. Since that time enthusiastic individualists have endeavored to use them.
Yet seven centuries ago P Piligrim offered the first magnetic engine. In the
16th century Jesuit priest J. Taisnerius was possessed with idea to create a
magnetic “perpetuum mobile”. Later on the number of such projects rose to
an avalanche and to date has reached such a scale that it becomes possible to
introduce a branched classification of such devices by various criteria. The
specific category of USA patents granted under the presentation of an acting

237
model of the device claimed is the official confirmation of the integrity of
some of the devices. Therefore, it is a matter of interest to demonstrate that
the alternators of this class do not either violate whatever laws of thermoki-
netics and are attributed to the “over-unity” devices just through misunder-
standing.
The magnetism of substances is known to be caused by basically the or-
bital motion of electrons, as well by their spin magnetic moments. Each of
the magnets has its own “magnetic ordered energy” content, i.e. a kind of
the “energy capacity” measured by the work which the magnet can do be-
fore its “depletion”. This capacity is highest for the relatively expensive
magnets made of rare-earth metals, whereas much lower for the magnets
made of the Alnico alloy.
The fact that magnets, in a great number of cases, do not lose their
properties when doing work may be explained by only their “charging”
from the ambient space. It is found that the permanent magnets, if shielded
from the ambient space with a magnetic screen, will “deplete” much sooner
in their operation under load. This will be confirmed hereafter based on the
testing the Perendev Co’s magnetic engines. In operation without load, as
long-duration tests show, permanent magnet coercive force changes negli-
gibly (from zero for the rare-earth magnets throughout 3% for the Alnico
magnets for 104 test hours). This supports the idea that if a permanent mag-
net is kept far from power supply lines, other magnets, high temperatures
and other factors affecting it adversely, this magnet will forever preserve its
magnetic properties. Shocks and vibrations do not affect the modern mag-
netic materials unless and until they physically damage them.
The temperature of magnets dropping is a sure evidence of disturb-
ing equilibrium between the electromagnetic field and the permanent mag-
nets operating under load. This effect has been discovered in all without ex-
ception operable machines on permanent magnets. In the Floyd generator
that temperature drop reached 200C. The energy exchange between perma-
nent magnets and the ambience is not evidently restricted to the thermal ra-
diation range. All bodies interpret the nonthermal radiation as work done on
them. Such phenomena as photoeffect, photosynthesis, etc. evidence this. It
is also known that the magnetic induction depends on the area electrons
“cover” in their orbital movement and may not only decrease in the process
of radiation or the work the permanent magnet does, but also recover in the
process of electromagnetic wave absorption increasing the energy of the or-
bital electrons. Thus the “mechanism” of such energy exchange is in general
known: this is the absorption of energy of various-frequency electromag-
netic oscillations or some “radiation” energy differing from the directional
flow of photons (radiant energy) by its isotropic character. This energy may
be attributed to oscillations of the ether as an all-penetrating medium which
stressed state causes the existence of the force fields. However, from the po-
sitions of thermokinetics as a non-hypothetic theory it does not matter. What

238
really matters is only that in any case the admissible continuous power of
magnetic engines is limited to the rate of their energy exchange with the
power source. This means that, from the positions of thermokinetics, the
permanent magnets should be treated in the same way as any other working
media running through a cyclic process, and whatever “perpetual motions”
are out of question. In this case the operation of the generators on permanent
magnets may be shown as based on the same mechanisms as for other heat
and non-heat engines. To this end let us use the basic equation of thermoki-
netics (2.4.5) which in our particular case looks like:

Wц = ∫ X м ⋅ dZ м , (13.4.1)

where Хм = В, Zм – magnetic induction and magnetization of the magnet in

whole, respectively.
Let us consider, as before, the scalars Хм and Zм and divide the circular
process into two parts, 1–2 and 2–1, in which limits the Zм variation follows
the same sign (dZм > 0 или dZм < 0). Then denoting Хм in “there” and
“back” directions with single and double prime, respectively, and consider-
ing dZм" = – dZм' gives instead of (13.4.1):
2
Wц = ∫ (Bм′′ − Bм′ )⋅ dZ м′ . (13.4.2)
1

From this it follows that if the mean magnetic induction of a material is

the same for both processes – its magnetization recovery (В'') and doing
work (В') – there will not be any work done for the cycle. The said is illus-
trated in Fig. 13.8, where an arbitrary cycle of a magnetic engine is de-
picted, which resembles the minor hysteresis loop.
The work for this cycle is defined by the cycle area. Hence, it is neces-
sary to somehow change the character of the “there” 1–2 and the “back” 2–1
B processes for the cycle area to become
2 other than zero. This may be provided
through e.g. Serl effect (magnet self-
B˝ rotation), or temporarily shielding the
magnetic field in-between the rotor
B΄ and stator at the stretch where the
magnets are closest to each other, or
varying the magnetic induction in the
1
rotor or stator by temporary variation
Zм of their temperature (nearby the point
Fig. 13.8. Magnetic Engine Cycle of phase transition), or varying the
path and field configuration when the
rotor and stator magnets are moving to and off each other, or temporarily

239
demagnetizing one of the magnets by electrical pulse at the moment of their
opposition, etc. We will see hereafter these are the means the inventers of
the permanent magnet-based devices of doubtless reality have recourse to.
The “Gramm’s generator” (Z.T. Gramm, 1869) is among the first of
such devices. It features a rotating annular rotor with the toroidal coil touch-
ing two conducting brushes diametrically opposite. The rotor rotates within
the poles of a stationary permanent magnet. The “asymmetry” of the mag-
netization and demagnetization processes for the annular rotor was provided
by shifting the moment of voltage supply to the toroidal coil.
Later in 1996 Russian engineer
A. Frolov upgraded the Gramm’s
generator. His design features a sta-
tionary ring with coils, while one
more coil in the center is used as the
alternative magnetic field source
(Fig. 13.9). Two magnetic fluxes
from the two load coils are mutually
offset and, thus, there is no reaction
in the primary circuit. This design
concept has become popular (O.
Fig. 13.9. Frolov’s Generator
Berens, Sweden; D. Hofmann,
USA; V. German, Germany; S.
Hartman, USA; etc.). One of the designs based on this concept and termed
by its authors as “Frolov’s generator” is illustrated in Fig. 13.10.
The central permanent magnet is here set into rotation by a small-scale
electric motor. To smooth the speed fluctuations, a massive flywheel applies
(Fig. 13-9). In August 1999 V. German with his team in Germany suc-
ceeded to obtain more that 1,200W on load. In 2003 S Hartman (USA) de-
signed a toroidal generator based on the standard 10kW-generator powered
from a car storage battery. The generator input current was 0.8A at a voltage
of 12.92V, the output current was
40A at a voltage of 6.5V. Thus the
power obtained was 25 times over.
John. Ecklin offered in 1975 an-
other way of the “asymmetry” crea-
tion (Patent USA No3879622 dated
22.04.75). His engine features mag-
netic shielding and opening in turns
and includes two permanent horse-
shoe magnets, a motor rotating “win-
Fig. 13.10. German’s Motor
dows” – magnetic shields, and a
magnet keeper of magnetic material,
which is in turns attracted to one of the magnets not shielded at the moment.
The keeper swaying transforms into the rotary motion through a crank gear.

240
The patent notes that with field strengths, magnet forms and materials, etc.
properly selected the energy obtained through the translation motion can ex-
ceed the one necessary to open and shut the “windows”. Ecklin did not
manage to create a “self-starting” machine, nevertheless, his concept served
as a basic for a number of USA patents: Jaffe (No3567979, 1976); Monroe
(No3670189, 1976); E. Gray (No3890548, 1976); V. Rivas (No4006401,
1977); H. Johnson (No4151431, 1979); F. Richardson (No4077001, 1987);
D. Regan (No4883977, 1989); V. Hyde (No4897592, 1990); H. Aspden
(No4975608, 1990), etc.
Some of the permanent magnets-based devices do not feature moving
parts at all. The Vacuum Triode Amplifier (VTA) of S. Floyd’s design
(USA) is one of them. Such a name is caused by the fact that the similar
principle of governing powerful flux through weak signal is used in triodes.
In the VTA design barium magnets were used prepared by repeated mag-
netic reversal at a frequency of 60Hz. That facilitated the transit of the field
from one direction to another when applying a weak signal to the control
winding from an external generator, thus, providing its “trigger” operation.
One of the VTA designs under demonstration included two sets of the
(4×6×1)-inch magnets positioned against two walls of the casing in such an
arrangement that an attraction appeared between them. The axes of the out-
put windings were parallel to the field lines, while the axes of the control
windings were perpendicular to them. The output power of the Floyd’s de-
vice was partly looped in feedback to excite the process resulted in a sig-
nificant power appeared in the output winding. Many investigators who
successfully repeated the Floyd’s experiments (e.g. J. Naudin in France,
whose design is shown in Fig. 13.11) noted that the best results of the mag-
netic substance “conditioning” were obtained when discharging alternate
current arc of the same frequency
as the control signal directly
through the permanent magnet
ceramics. Thus a bistable solid-
state substance was created
where acoustic resonance
showed at the weak control mag-
netic field oscillation frequency.
Special attention is attracted to
Fig. 13.11. Naudin’s Generator the current generators based on
using the “self-sustaining rota-
tion” effect of John Serl (Mortimer, Borkshire). In the fifties of the last cen-
tury he discovered that adding a small alternate current (~100 ma) radio-
frequency (~10 MHz) component in the manufacturing process of perma-
nent ceramic magnets endued them with new and surprising properties.
These consisted in the unusual interaction of the permanent magnet with the
magnetic rollers located on its surface. That interaction showed in a sponta-

241
neous rolling of the rollers after applying a minor momentum to one of
them.
In Russia the Serl effect was studied in the Institute of High Tem-
peratures of Russian Academy of Science. V. Roschin and S. Godin, re-
searches of this Institute, in 1992 built a generator similar to the Serl’s one,
which they termed the “magneto-dynamic converter”. It constituted a stator
with sector permanent magnets 1 and an annular rotor with rotating mag-
netic rollers 2 (Fig.13.12).
15 1 2 9 13 11
8 2 3 4 8
10
10
11
12
12

5 7
14
6

Fig. 13.12. Roschin–Godin’s Converter

The rotor had a diameter of 1 m and a weight of 500 kg. The rotor seg-
ments were made on the basis of rare-earth magnets with a residual induc-
tion of 0.85 T. They magnetized by the bank of capacitors discharging
through an inductor. Unlike the Serl’s disc, the ac biasing was not used in
the Roschin–Godin’s device. The “engagement” of the rollers with the rotor
annular magnet was provided to the gear principle through transversal mag-
netic inserts made of NdFeB with a residual induction of 1.2 T and arranged
in the stator and the rollers. There was an air gap of 1 mm provided for in-
between the stator surface and the rollers. The elements of the magnetic sys-
tem were assembled as a unit construction on a platform made of non-
magnetic alloys. The platform was equipped with springs, dampers and was
free to move vertically along three guides, which was measured by an in-
ductance transmitter 14. The stator 1 was fixed, while the rollers 2 secured
on a mutual movable separator 3 with the help of air dynamic bearings. The
separator was rigidly coupled to the shaft 4 and, through friction overrun-
ning clutches 5, to the starting motor 6 and electro-dynamic generator 7.
There were electromagnetic transducers 8 with open magnetic conductors 9
positioned along the rotor. The load 10 was arranged as electric incandes-
cent lamps. The machine was put into operation by speeding the rotor up
from an electric motor. At ~550 rpm the speed of rotor rotation became
spontaneously increasing despite the motor stopped with the electro-
dynamic generator coupled to the shaft. To keep the speed, a load in the

242
form of tubular electric heating elements was stepwise connected to the
generator. The power output of the machine was 7 kW.
There were a number of unusual effects observed in the machine be-
sides the “surplus power” generation, viz. platform weight reduction (to
35% of the initial weight); corona discharge in the form of a blue-pink
glowing; vertical concentric zones of increased magnetic intensity (to order
of 0.05 T) and abnormal temperature fall (by 60C–80C) in close proximity to
the converter. The impossibility of the existing theory to explain all these
effect in whole evidences its deep inferiority.
An interesting version of a switching-reluctance device generally
recognized as one of the most successful “free-energy” engines was offered
by Robert Adams (New Zealand) in 1977. In the Adams’ motor-generator
(Fig. 13.13) a rotor having per-
manent magnets with like poles
oriented radially outside rotates
generating inductive currents in
the stator coils positioned
around the rotor in the plane of
rotation. From the traditional
electrical engineering stand-
point a motor-generator without
the closed core (coil cores have
the bar configuration) is ineffi-
Fig. 13.13. Adams’ Motor
cient. However, this is the open
core that makes it possible to
generate power without rotor braking. There is no the electromagnetic in-
duction phenomenon here in the true sense of the word, but only the mag-
netic induction, i.e. the magnetization and demagnetization of the stator core
in the rotor permanent magnet field. It may be seen a full analogy with the
electric induction, i.e. the “electrization by induction” as they would say be-
fore. The “electrization by induction” differs from the electromagnetic in-
duction since the secondary magnetic field generated in the generator wind-
ing does not decelerate the rotor and does not interact with the primary mag-
netic field. Robert Adams is
working together with H Aspden
on acquisition of patent for his
system. It is significant that op-
erability of this machine can be
completely explained as based
on the Faraday’s law.
The K. Minato’s wheel (Pat-
ent USA No5594289, 1997) is
another rotating magnets-based
Fig. 13.14. Minato’s Wheel
engine even more attractive for

243
reproduction. It consists of a rotor (a bicycle wheel rotating on a horizontal
axis) having a multiplicity of permanent magnets attached and with their
like poles oriented in the rotor rotation direction, as well as stabilizers in-
tended to balance the rotor (Fig. 13.14). Each of the permanent magnets at-
tached to the rotor is positioned at an angle to the wheel radius. An electro-
magnet with electric current periodically induced depending on the rotor ro-
tation is arranged nearby the external periphery of the rotor closely to it.
Eric Vogels (Sweden, 1997) repeated and improved the Minato’s results
having split the track of the magnets into a multiplicity of minor tracks.
Many patents on magnetic engines have been granted in Russia (V.
Alekseenko, No5037775, 1996; V. Rykov, No2000101256, 2001; A.
Rjumin, No2001123502, 2003; V. Levkin, No5032711, 1995; M. Ostrikov
et al, No95103846, 1996; A. Starostin et al, No95112010, 1997; A. Kalinin,
No94019782, 1996; P. Imrish, No94026259, 1996; V. Dudyshev,
No2128872, 1998; Y. Pilipkov, No2000119415, 2002, etc).
American television network
recently released the information
about a revolutionary technique de-
veloped by J. Newman. His elec-
tromagnetic motor (Fig. 13.15) is
capable to power a production area,
residential building or farm. It is
based on the concept that the mo-
tion of electrons in coil wires re-
sembles the motion of micro-
gyroscopes as electrons rotate. For
high-inductance coils this motion
engenders special effects showing
Fig. 13.15. Newman’s Engine in the “surplus power” generation.
It is significant that more that 30
physicists, nuclear engineers, elec-
trical engineers and electricity ex-
perts signed the letter that con-
firmed the revolutionary character
of that invention.
The homopolar induction ef-
fect known yet from Faraday’s
time is also capable to create elec-
tromotive force at the rotation of a
Fig. 13.15. De Palma’s Generator
metallic rotor. De Palma’s ho-
mopolar generator (1991) is one of the practical designs the alternators of
this class feature. The testing results on this generator (Fig. 13.16) show that
its rotor deceleration due to the back EMF appears to a less degree than in

244
the traditional generators. Therefore, system power output surpasses that re-
quired for the rotor rotation.
It is worth noticing that the creation of the alternators has already gone
into a practical development stage. Swiss company SEG has quite recently
announced its intention to bring to the market a generator operating on the
Serl effect. It is planned to launch in the first place a compact 15-kW gen-
erator with approximate size parameters of 46×61×12 cm, which can be
tuned to generate direct or alternate current of various voltage within a
range of 12V to 240V. Each of such generators is capable to produce
60MW/h of energy before the reversal of the polarity would become neces-
sary. The generator model D15AP offered is illustrated in Fig. 13.16.
It consists of three four-layer concen-
tric rings, each of the rings being made
of composite. The rings are concentric
to each other and attached to a base.
Rollers freely rotate around each of the
rings: 10 pieces around the first ring,
25 – around the second ring and 35 –
around the third one. There are coils
arranged outside the rollers at the ex-
ternal ring periphery. The coils are
connected in different manner, which
enables generating either direct or al-
ternate current of various voltage. The
output coils have to be calculated to
Fig. 13.16. Serl’s Generator provide an output voltage of 240V at a
power output of 15kW. The generator
is a set of frictionless bearings in a way and, at the same time, a system of
three rotary transformers in one casing with a current of an extremely high
voltage at the output.
The LUTEC generator of Australians Brits and Christie (International
Patent No00/28656, 2000) has been one of the recent additions to devices of
this class. The simplicity of this engine is provided by arranging switchable
coils on the stator, while a permanent magnet – on the rotor. The direct cur-
rent input to the stator coils causes the magnetic repulsion force and is the
only current to create the “aggregate motion” (Fig. 13.17).

245
 A number of private enter-
prises are currently taking orders
for commercial motor-generators
on permanent magnets. In particu-
lar, the GMC Holding Corporation
(Orlando, Florida, USA) claims
that as a result of the 12-year study
it has created a device on perma-
nent magnets capable to solve the
world economic problems in
power engineering. Another com-
Fig. 13.17. LUTEC Generator
pany, Perendev (abbreviation of
“perpetuum energy device”)
claims that a 30kW-magnetic motor of its production is ready to find market
acceptance (Fig. 13-18).
Cost estimate of the first devices is about 8,500 €. Truth to tell, Keith
Anderson, whose company was invited to test the Perendev motor and built
two operating analogs, claims all of them depleted their load magnets.
Hence, further investigations are
necessary to clarify the balance
conditions for the internal energy
the magnets release and their re-
plenishment from the ambience.
The key problem of the designs on
permanent magnets is to calculate
the magnetic flux distribution in the
magnetic circuit that may include
permanent magnets, air gaps, high
magnetic permeability elements
and electrical currents. Accurate so-
lutions on magnetic fields demand Fig. 13.18. Perendev Magnetic Motor
a complicated analysis of many fac-
tors, though approximate solutions are also acceptable based on certain sim-
plifying assumptions. To obtain an optimal design with permanent magnets,
an experience and trade-offs are often needed. Meantime, the Perendev
company takes orders from those who have the adventurous spirit alive and
appreciate risks and restraints of this early stage.
One more magnetic engine modification called as Cycclone1) has been
recently created on funds of an American company in Australia (Fig. 22-
19). A demonstrational prototype of this engine intended for a car applica-
tion was shown on TV.

1)
The name reflects the engine combines cyclic motion with magnetic field “cloning”.

246
 It is quite evident that when estimating the availability of such devices it
is impermissible to consider the magnetic energy as “gratis” – its prime cost
needs the same expense accounting as for any other power units on renew-
able power sources. These expenses depend on the magnet classes. The sa-
marium-cobalt (SmCo) and neodymium (NdFeB) magnets sintered from
rare-earth elements are the newest addition to the earlier known ferrite (ce-
ramic) and aluminum-nickel-
cobalt (Alnico type) magnetic
materials. They feature the mag-
netic energy level achieved as
45–50 MGOe (mega gauss oer-
sted). The recent developments
in the magnet fabrication tech-
nique contributed to this
achievement. These develop-
ments have opened new thrilling
vistas toward technology upgrade
of the permanent magnet en-
Fig. 13.19. Cycclone Engine gines. Besides their production
costs it is reasonable to estimate
also the efficient design effort and theoretical in-depth studies since the im-
possibility to explain the multiple effects revealing themselves in applica-
tion to the alternators in operation evidences that theory is far behind ex-
periments.
Thus we can expect the market will see small-scale power units capable
to efficiently power offices, residential buildings and farms located far away
from power transmission lines.
In conclusion it is desirable to underline that despite the fact the opera-
tion of alternators is obscure in many respects there are enough grounds at
present to search for the most successful technical solutions on creation of
new-generation converters using the practically inexhaustible energy of the
ambient space.

13.5. Converters of Radiation Field energy

A dozen phenomena are known to date where after the “activation” of

some working medium the heat generation during its relaxation exceeds the
energy input. Such a “production” of heat energy may be observed in the
oxygen–hydrogen electrolytic tanks on usual and heavy water (V. Filimo-
nenko, 1957; S. Johns, 1989), in the electrical arresters (A. Chernetsky,
1971); vortex heat generators (Y. Potapov, 1992); at plasma and plasma-
chemical dialysis (A. Frolov, 1998; F. Kanarev, 2001); at “sonolumines-
cence” (R. Taleyar Khan, 2002), etc. Although such phenomena were dis-

247
covered a long time ago (F. Latchinov, 1888), these became known to most
people after Stanly Pons and Martin Fleishman, researchers of the Utah
University, had reported the results of their experiment they construed as the
“cold thermonuclear fusion”. Pons and Fleishman claimed they had ob-
served a heating of the electrochemical element that used heavy water as an
electrolyte. In that case the heat released much exceeded the electric power
consumed. Fleishman assumed that the energy was generated inside the pal-
ladium cathodes in the course of the nuclear reaction where two deuterons
combined into 4He in some manner yet unknown. However, no products of
that reaction were discovered which would correspond to the heat genera-
tion as claimed. The scientific community did not recognize that claim for a
scientifically substantiated explanation for several reasons. Firstly, the Cou-
lomb barrier impedes the combination of deuterons at room temperature.
The mechanism when deuterons could approach each other so that the syn-
thesis would occur was unknown. Secondly, if they could approach each
other for a reaction to occur, usual synthesis products should have been ob-
served as they appear very soon. Thirdly, a reaction when two deuterons
combine and form 4He usually runs with releasing gamma rays of order of
24 MeV. However, such a flow of gamma-ray emission was not observed
during the experiment. Finally, the reaction in whole ran a million times
faster than at usual conditions. Therefore, after many of the laboratories that
attempted to reproduce that experiment had failed the scientific community
came to a conclusion that the data of those experiments was incorrect.
Meantime, yet in 1989 S. Johns made another statement that during electro-
chemical reactions in heavy water he had observed the deuteron synthesis
reaction. The scientific community’s response to that statement was skepti-
cal as well since the signal/noise ratio was minor, while the theoretical con-
siderations did not allow treating as significant the effect Johns over-
stressed.
Therefore, it was an unexpected sensation when in 2002 the extremely
prestigious scientific magazine “Science” published an article about a desk-
top thermonuclear unit R. Taleyar-Khan’s team (USA) had created. There a
small cylinder containing acetone with the hydrogen nuclei replaced by the
deuterium ones was irradiated with a powerful flow of acoustic waves si-
multaneously with a neutron flux. The operation of the unit was claimed to
have been based on the sonocavitation effect when the acoustic waves
shook the water thoroughly having a multiplicity of bubbles produced there
with a diameter of to 1 mm (much greater than usually), which then “col-
lapsed”. Something like this has been known since yet the thirties when it
was discovered that some substances started to glow when ultrasound was
passed through them (a phenomenon known as the “sonoluminescence”).
Physicists affirmed that in that case acetone was heated to such tempera-
tures that the deuterium nuclei started to merge. As in the case with the
“cold thermonuclear fusion”, the investigators immediately encountered

248
troubles. The magazine “Science” that can not hazard its reputation when
publishing the like sensations without credibility check gave the floor to
also other investigators who attempted to reproduce the experiment. They
also discovered the neutrons. However, when they started to measure the
neutron flux with a detector more advanced than in the initial test, the parti-
cles immediately disappeared. Besides, nothing proved those neutrons had
something to do with the thermonuclear reaction. As a result, a good part of
scientists came to a conclusion the process of such a kind just did not exist
in nature.
Nevertheless, a number of investigators continue to insist on the fact
that the water itself is the source of energy in such phenomena. They often
refer herein to an installation known as the “Patterson cell” (USA). This in-
stallation is an electrolytic element filled with fine plastic beads covered
with superfine layers of nickel. The latter, like palladium, is capable to col-
lect and hold the heavy isotopes of hydrogen. When current is passing
through the layers, electric charges appear on them. Such a device, accord-
ing its producer – Patterson Power Cell (USA), steadily outputs 5W of heat
power per each 1.5W input. In Patterson’s opinion the “surplus” heat in his
device appears due to the cold nuclear fusion. The Patterson’s electrolytic
thermal cells are currently produced by the ENECO Corporation that has
collected above thirty patents with key technological solutions into a unitary
patent package. The Nova Resources Group Inc. (Canada) has also launched
the cells into production.
However, the physical proofs the advocates of the “cold nuclear synthe-
sis” theory and of water as the “energy of future” source adduce are so du-
bious that many investigators are inclined to call it after R. Park the “voo-
doo science”. To make sure that some effects of the “cold nuclear synthesis”
are not the reason, but rather consequence of an external impact on a sys-
tem, it is enough to consider the operation of the machines wherein the
“surplus heat generation” is observed at the conditions when the nuclear
synthesis is excluded.
The quite sensational results Stanly Myer obtained confirming the reso-
nance effect of electrostatic field on water molecules are among most
known. In the late eighties he developed and fabricated a “low-current” wa-
ter fuel electric cell (Fig. 13.20) that makes it possible to separate the usual
tap water into hydrogen and oxygen with much less energy consumption
than at the usual electrolysis. As S. Myer’s investigations showed, when wa-
ter is exposed to a frequency coinciding with its natural molecular fre-
quency, the chemical affinity of the hydrogen and oxygen obtained (gener-
ated as heat in the process of their subsequent synthesis) appears to exceed
the energy consumed from the current source.
The design of the Myer’s cell is simple. Its electrodes are made of
stainless steel plates either parallel or concentric. The gas output is usually
proportional to the distance between them – a good result is assured with a

249
patented distance of 1.5 mm. Considerable distinctions from the usual elec-
trolytic cells lie in the power supply to the
cell. Myer uses an external inductance that
forms an oscillation circuit with a cell capaci-
tor to create a parallel single-tuned network.
It is actuated by a powerful pulse oscillator
that along with the cell capacitor and a recti-
fier diode constitutes a pumping circuit. The
high frequency of the pulses generates an in-
step rising potential until the point reached
where the water molecule disintegrates and a
Fig. 13.20. Myer’s Cell
short current pulse appears. The meter-
current supply circuit reveals this step and
locks the impulse source for several cycles allowing the water recovery.
Whereas the usual water electrolysis requires a current measured in am-
peres, the Myer’s cell provides the same effect at millamperes. Furthermore,
usual tap water requires an electrolyte (e.g. sulfuric acid) to be added to in-
crease conductivity, whereas the Myer’s cell operates with pure water pro-
viding a tremendous throughput. According to eye-witnesses the most strik-
ing aspect of the Myer’s cell was that it remained cold even after several
hours of the gas production. The recorded gas output was enough to show
the oxyhydrogen flame that instantly melted the steel. To say nothing of the
abundant release of oxygen and hydrogen and minimal heating of the cell, it
was surprising that the water even inside the cell disappeared quickly hav-
ing transformed into its components in the form of a mist spray as a great
number of tiny bubbles that covered the cell surface. Myer claimed his oxy-
hydrogen mix converter was working for last four years based on a chain of
six cylindrical cells. The Myer’s invention is covered by a series of USA
patents (Patents 4.936.961, 4.826.581, 4.798.661) filed before 1991 under
heading “101” (where granting a patent for invention is subject to its suc-
cessful demonstration). The inventor explains the cell operation by the po-
larization of water molecules under the electrical field gradient and the
resonance within the molecule, which enhances the effect. This standpoint is
close to thermokinetics which considers the low-current electrolysis as a
remarkable example of the work done by high-frequency electromagnetic
oscillations applying the resonance effect to the electrons relatively weak-
retained.
Some investigators explain the surplus heat appearing by the water mo-
lecular structure reconfiguration. Actually, tens of the water molecular
structure modifications are known which evidently differ in also their inter-
nal energy. In such a case the water transit from one configuration into an-
other with a lower energy level must involve a heat release. In particular, as
Dr. R. Mills (USA) sees it, the hydrogen atoms in the water molecule may
be at different energy sublevels including also those corresponding to frac-

250
tional quantum numbers. He called the water molecules with hydrogen at-
oms being at a lower (than usual) energy level as “hydrino” (see also S.
Nesterov, 1995). However, this standpoint contradicts quantum physics and
meets, therefore, many opponents.
Meantime, there may exist another and quite adequate to the existing
paradigm explanation of the surplus heat release phenomenon in the above
processes. For this it is necessary to recognize the fact that the “heat genera-
tors” above described operate as heat pumps. They feature the reverse cycle
“water electrolysis – hydrogen and oxygen synthesis” where the dissipated
radiant environmental energy is a heat well analog, while the work spent for
electrolysis – an analog of compression work in heat engines.
Released heat Q to power supply energy W ratio called as heat trans-
formation coefficient ηт in thermodynamics is known to be the efficiency
criterion for inverse cycles:

ηт = Q/W > 1, (13.5.3)

This criterion always exceeds unity if only W is construed as the work a

source of current supply spends. However, this not at all means the “over-
unity” efficiency ηt = W/Q obtained since the efficiency is construed as a re-
ciprocal, viz. work W obtained in engine related to heat Q supplied from
”hot’ well.
The experiments conducted by Wm. Lyne (1996) are further evidence
that the field (distinguished from substance) forms of energy of the ambient
space are a source of surplus heat energy at the spontaneous synthesis of
hydrogen and oxygen molecules in the Myer’s installation. In 1981 Lyne
built and tested a heating system on atomic hydrogen. The hydrogen in his
installation, like at the usual hydrogen welding, passed through electric arc
which decomposed it into “atomic” hydrogen. Then the atomic hydrogen
recombined under heat release. Thus the hydrogen played a part of a “me-
diator” that ran through a cyclic process, while the electric power consump-
tion to maintain the arc was considered by Lyne as an “activation” energy.
The energy “lacking” for the hydrogen dissociation was taken, in Lyne’s
opinion, from the “ether”.
Similar results are obtained in tests with heat generators on atomic hy-
drogen, which have been conducted in Russia since 2003 by A. Frolov,
founder of the Faraday Lab. Ltd (JSC LNTF). His design is based on a pow-
erful electron-vacuum diode with the tungsten directly heated cathode (Fig.
13.21).
The generator includes a cylinder with inlet and outlet channels where
the water stream flows around a closed internal chamber filled with hydro-
gen at a pressure of 0.2 at. The diode tungsten filament with a diameter of
0.25 mm in the center of the device serves as a cathode whereon the hydro-
gen changes from molecular state into atomic one. Then during the H–to–H2

251
transit a “surplus” heat is released, which is removed with the cooling water.
Hydrogen is not consumed here. To create the “low-current” splitting of H2
into H, a potential difference with pulses within 200 V to 300 V is applied
between positive anode and negative cathode. The pulses are generated by
direct current with a frequency of to 10 MHz. When using the cathode pulse
heating of 12 V with a frequency of 51 Hz and a
pulse/break ratio of 5%, a multiple excess of the
heat produced over the electrical power consumed
is obtained. It is significant that the anode voltage Cooling System
pulses generated by alternate current do not cause Water Inlet Outlet
this effect. These experiments obviously enough
confirm the above “thermokinetic” interpretation H2 at a p = 0.1 at
0.3 liter
of the “surplus” heat production origin. The ab-
sence of the hydrogen consumption evidences that
hydrogen is here just a “mediator” (working me-
dium), but not the energy source, while the ab- Tungsten
sence of the “surplus heat generation” effect with Cathode
Filament
ac supply to anode confirms the fact that this ef-
fect is obtained not by the “excitation” of orbital
electrons at their mean energy level invariable, but
rather by the unilateral action of the voltage pulses
in the direction of removing the orbital electrons
from the hydrogen nucleus (in resonance with
pulses of the electromagnetic field surrounding the
device). The selection of the anode voltage pulse Fig. Generator
13.21. Frolov’s

frequency is necessary in this case in order to ob-

tain resonance between the external field action and the “activation” pulses.
French investigator J. L. Naudin has recently improved the operation of
this generator replacing the power unit by a large-capacity battery with a 10
MHz-pulse generator. On this basis Naudin managed to better form the an-
ode voltage pulse, which configuration appeared have been very important,
and thus to build up the “heat output”. His “atomic hydrogen generator” was
in operation for an hour with an efficiency exceeding 2,000%, i.e. with 20-
fold exceed of the heat generation over the electrical power consumed. He
used two types of absolutely different measuring instrumentation which
confirmed the reliability of input power measurement for the installation.
Y. Potapov’s “heat generator” (Patent of RF No2045715, 1993) is one
of the widely known devices intended for practical using “surplus” heat
generation. It constitutes a vertical cylindrical tube with a water flow tan-
gential input in its upper part (Fig. 13.23). The velocity of the water is such
that there are cavitation phenomena causing quick water heating and ob-
served in the tube along with the intensive flow turbulence. If this water is
removed from the maximal temperature zone and directed to usual heating
devices with return to the lower part of the vortex tube (wherefrom it is

252
taken by a usual centrifugal pump and pumped again to the upper part of the
tube), the amount of the heat removed for heating appears to exceed the
amount as to the balance of the dissi-
pative (hydrodynamic) losses for
friction and cavitation. From the au-
thor’s initial statements his heat gen-
erator produced up to 3 to 4 kW of
heat per 1 kW of the electrical power
consumed. However, at comparative
testing in NPO «Energia» in 1996 the
heat generation in the Potapov’s de-
vice appeared to have been only 23%
higher that in the 3-phase ac elec-
trode boiler and 42% higher than in
the electrical boilers with standard
heating elements (tubular electric
heating elements). In Potapov’s opin-
ion the surplus heat generation in his
installation is explained basically by
the integration of water molecules in
“associates” under the cavitation ef-
fect and, to a lesser degree, by the
cold nuclear fusion reactions (Y. Po-
Fig. 13.23. Y. Potapov’s Heat Generator tapov et al, 2002). This standpoint is
supposedly supported by the discov-
ered in the HPO «Energia» 1.5-increase in the hard gamma radiation (not
mitigated with steel shield) and the generation of carbon radioactive iso-
topes if the installation filled in with an anti-freezing agent.
At the same time the author has repeatedly stressed that the water in the
installation may not be interchanged for several years. This means that the
water and the anti-freezing agent were actually just a working medium par-
ticipating in the closed process, whereas a primary energy source should be
searched for beyond it. This is the opinion F. Kanarev (2004). He thinks
physical vacuum to be the “surplus” heat generation source.
The surplus heat generation phenomenon theoretically allows creating
autonomous (self-sustained) power supply sources. To create them, it is
necessary to combine electrolysis cell with fuel cell or heat engine convert-
ing heat Q obtained at burning of oxyhydrogen mixture into electric power
W. According to thermodynamics the efficiency of such a machine will be
dependent on the temperature T of the heat Q yielding from synthesis or re-
combination. If the temperature of the heat obtained as a result of synthesis
or recombination exceeds that of the ambient space To≈ 300 К by just 30K,
ideally the Carnot engine built on this heat source will have the thermal ef-

253
ficiency only 9%. This means that, to cover the energy expended on elec-
trolysis, a more that 11-fold “surplus heat” output is needed.
Nevertheless, the thermokinetic analysis undertaken herein is meant to
mitigate the suspicion of formal science toward the phenomena above de-
scribed and to switch the problem of creating electrical energy alternative
sources to practical course. Several firms are already selling cavitation heat-
ing equipment. Formal science is looking askance at this trade activity, the
more so because the surplus heat generation in this equipment is not sup-
ported with independent expert appraisal data. However, profit of trade ap-
pears to be dominating. As reported at the Genesis World energy consor-
tium’s site, they have already developed a self-contained and self-sustained
energy generator “Edison Device”. In its size it is approximates to the ex-
ternal air conditioning system, which makes it possible to quickly and easily
install it at home or in office to obtain a practically unlimited energy from
any available water source. The commercial model of the device is capable
to generate 100 kW of energy per diem. The “mechanical part” comprises
minor pumps and micro-valves providing a circulation, which makes the
device noiseless and not demanding a special maintenance for the designed
20-year service life. The feasibility of such environmental energy converters
are also confirmed by the valid tests on the Mayer’s converters in motor car
application. Water consumption per a haul of 100 km was there about 3 li-
ters (Pat. USA No 149.407). Thus the point is just to find the most promis-
ing engineering solutions in this field.

Conclusions

Applying thermokinetic to systems doing useful work has made it pos-

sible to set and solve a number of new interesting problems. Laying founda-
tions of similarity theory for linear power converting systems should first of
all be noted among them. This theory has first confirmed in most general the
unity of general regularities of energy conversion in heat and non-heat, cy-
clic and non-cyclic, direct and inverse machines. It has thereby uprooted the
inconsistency of attempts to restrict the application of classic thermody-
namic laws to only heat engines. Practical importance of this theory is that it
allows transferring the investigation results from one (well studied) power
and process installations to another (little studied). The examples given in
this chapter confirm this.
Laying foundations of productivity theory for technical systems has be-
come not less important achievement of thermokinetic. Enabling the synthe-
sis of thermo-kinetics and thermo-economics this theory supplements the
classic analysis of thermodynamic efficiency for power units by correlating
their power and economic efficiency. Thereby it approximates the thermo-

254
dynamic estimates of limiting potentials for irreversible processes to reality.
At the same time this theory allows finding the economically most advanta-
geous operational modes for various power, process and transport units,
which is of great practical importance. The theory application examples
given in this chapter demonstrate that the conclusions from thermokinetic
are of the same strictness and generality degree as those from thermody-
namics with regard to reversible processes.
However, most “exotic” are perhaps the deductions of thermokinetic
about the possibility of using field forms of energy. These results allow ex-
plaining the functionality of a number of devices presently referred to “per-
petuum mobile”. The hermokinetic analysis given in this chapter and per-
taining to devices consuming the energy of gravitational, electrostatic and
electromagnetic fields demonstrates that the operation of these devices does
not exceed the bounds of thermokinetic and does not conflict with its laws.
Lifting theoretical bans on creation of such devices opens new vistas to us-
ing renewable energy sources alternative to the known ones and actually in-
exhaustible.

255
 Part 5

EXTENSION OF ENERGY TRANSFER THEORY APPLICABILITY

Various thermo-mechanical, thermo-diffusive, thermo-electrical,

thermo-magnetic, etc, effects widely used in modern heat engineering
devices and arising with dissimilar non-static processes simultaneously
running in the same domains of space have given rise to a keen interest of
science and engineering in boundary fields of knowledge. This causes the
necessity to synthesize the heat engineering fundamentals with other
scientific disciplines. One of such attempts was made in thermodynamics of
irreversible processes. However, the theory of irreversible processes (TIP)
was constructed by extrapolating classic thermodynamics beyond the strict
framework of applicability of its system equilibrium and process
reversibility concepts. This has led the TIP to its losing the rigor and
completeness intrinsic for the classic thermodynamic method. In this
context it becomes a question of significant importance to provide a
consistent thermodynamic substantiation of the TIP fundamentals from the
more general positions of thermokinetics.

Chapter 14

THEORY OF NONLINEAR TRANSFER PROCESSES

This chapter is dedicated to a new method proposed to find and describe

these effects. The method consists in finding for each of the independent
transfer processes a resultant force which disappearance ceases the process.
The method allows to dramatically facilitate the thermodynamic description
of irreversible processes and to cut the number of the kinetic coefficients
256
required for such a description. It casts light upon the origin of the so-called
stationary superposition effects for different-kind processes, which will be
obtained here from the conditions of partial (incomplete) equilibrium as a
result of mutual compensation of the components the resultant force consists
of.
The effectiveness of the method proposed and the validity of the
conclusions received within the frames of thermodynamics will be
confirmed here by ample experimental data.

14.1. Proof of Existing Motive Forces’ Resultant in Independent

Transfer Processes

When having extended the L. Onsager’s formal theory of velocity to the

vector transfer processes, H. Cazimir (1945) and I. Prigogine (1947) kept to
the same concept according to which any of the independent flows Ji was
linear with all the thermodynamic forces Xj acting in the system, i.e. they
used the same Onsager’s law (4.1.6) but with the scalar flows Ji and forces
Xj substituted there for the vector values Ji and Xj. Thus obtained kinetic
equations (4.4.1) added to known Fourier’s, Ohm’s, Fick’s, Darcy’s,
Newton’s, etc. laws the “non-diagonal” terms LijXj which characterized the
contribution of the “alien” force Xj to the ith flow Ji (j ≠ i). E.g., to describe
the thermo-diffusion phenomenon (the kth substance flow arising under a
temperature gradient) the right side of the Fick’s law of diffusion is added
with a linear term proportional to the temperature gradient. This is
equivalent to representation of the flow Ji as a sum of several summands Jij
= Lij Xj :

Ji = Σj Jij = Σj Lij Xj . (14.1.1)

Such a description corresponds to a notion that the non-matching flows

running simultaneously in the same space areas are as if “superimposed”
each one onto other (S.R. De Groot, P. Mazur, 1964) mutually “entraining”
and intensifying each other (A. Veinik, 1973; N. Bulatov, A. Lundin, 1984).
More than half a century this notion has been indisputable and reproducible
in all study guides on thermodynamics of irreversible processes but with a
proviso that, according to Curie symmetry law, only processes of the same
(or even) tensor order and kind can interact (be superimposed) (S.R. De
Groot, P. Mazur, 1964). This law first substantiated by P. Curier (1947)
257
stated that due to the possible spatial symmetry in anisotropic systems the
number of coefficients in linear equations would decrease in such a way that
not all Cartesian flow components would depend on the force components.
However, it appeared important for isotropic systems, too. As De Groot and
Mazur showed (1964), in anisotropic systems the Cartesian components of
thermodynamic forces of different tensor order and kind would be
transformed at rotation and inversion in such a way that only the links
between the same tensor-order flows and forces would remain invariable.
Thus for isotropic systems the Curie law may read as follows, “In isotropic
systems the phenomena described by thermodynamic forces and flows of
different tensor order and kind do not influence each other (I. Gyarmati,
1974). In other words, any vector flow Ji may depend on only forces of the
vector character.
The investigators were not at all confused with the fact that the
Onsager’s postulate disagreed with the centuries-old foundations of
mechanics according to which each independent process (movement,
acceleration, setting of mechanical equilibrium, etc.) could be associated
with the only (resultant) force which disappearance would cease the
process. The anisotropic heat conduction and electric conduction equations,
which, as Onsager admitted, had prefigured his “phenomenological” laws,
also evidenced the motive force uniqueness. The components of the only
motive force – temperature gradient and electric intensity, respectively –
were in those equations under the sum sign, too. Ultimately, it was not at all
a secret that the said “superposition effects” reached their extremum at the
so-called stationary conditions when some (the so-called “non-fixed”) flows
simply disappeared and therefore could not in principle be superimposed
onto the remainder flows. E.g., in electrolytic solutions, where the
phenomena of electric conduction and diffusion take place, the voltage
(Quincke effect) is maximal with the current disappeared. (Haase, 1967).
The same situation is about Soret effect – generation of the kth substance
concentration gradient within the initially homogeneous system with a
temperature gradient generated, the said concentration gradient reaching
peak value with the diffusive flows disappeared. Hence the reason of such
superposition effects arisen should have been searched for anywhere but not
in the interaction (superposition) of the irreversible phenomena themselves.
In fact, the generation of gradients or differentials of whatever potentials
meant the system withdrew from that kind of equilibrium, i.e. anti-
dissipative processes appeared in the system. Meantime any relaxation

258
phenomena are always directed toward setting equilibrium. In other words,
the components Jij of the generalized rate Ji of any relaxation process in
(5.1.3) have always the same sign. Therefore some of these flows Ji
becoming zero in stationary state could be caused by only mutual
compensation of the components Jij of this flow. However, all terms in
equations (14.1.1) have intrinsically the same sign as they describe the
relaxation processes. Hence the appearance of the “superposition effects”
could have been explained by only the generation of anti-dissipative
processes in the system. However, the investigation of such processes was
evidently beyond the Onsager’s theory. Therefore such an assumption could
not arise in its depths.
To reveal the fallacy of the Onsager’s postulate, let us note that the
coefficients Lii and Lji in the components Jii = Lii Xi and Jji = Lji Xi of flows
(14.1.1) are expressed in terms of the matching components Jii and Jji of the
flow Ji :

Lii = dJii/dXi ; Lji = dJji /dXi . (14.1.2)

According to (14.1.2) there is a simple relation between the diagonal Lii

and non-diagonal (“cross”) coefficients Lij of the transfer equations in
matrix form:

Lji = Lii (dJji /dJii) . (14.1.3)

However, at Xj = const dJji = dJj and dJii = dJi , therefore (14.1.3) may
be changed to:

Lji = Lii (∂Jj /∂Ji)Xj . (14.1.4)

From this it follows that providing the flows Ji and Jj are not
interconnected, the coefficients Lji, at Xj = const, become zero along with
the derivative (∂Jj/∂Ji)Xj. This means that for the independent flows Ji and Jj
the reciprocity relationships are satisfied trivially (Lij = Lji= 0) just as
expected. In other words, in the absence of the motive force Xi as matching
the flow Ji no other force whatever can induce this flow providing it is truly
independent. Since the independence of the flows Ji and Jj is laid into the
foundation of the Onsager’s theory by the flow Ji definition as itself, kinetic
equations (4.1.6) are really valid for interconnected flows.

259
 So in the absence of additional constraints imposed the Onsager’s
postulates should be superseded by a statement reading that the generalized
rate of whatever irreversible process (flow Ji) depends on only the
components of the sole(resultant) motive force Fi. This statement
corresponds to the kinetic transfer equations of the (2.6.11) kind:

Ji = Ki(Fi – Fiо) = Li(Xi – Xiо), (i = 1, 2, ..., ni) (14.1.5)

where the coefficients K ij = Кi(Ψj, Fi ), unlike (4.4.1), are arbitrary functions

of thermostatic parameters (temperature, pressure, concentrations of the kth
substances, etc.), as well as of any forces Fi. The principle difference of
equations (14.1.5), besides their non-linearity and the “threshold” values of
the forces Fiо which can be taken into consideration, is that they contain the
only kinetic coefficient K ij – the fact that considerably facilitates the
investigation of transfer process kinetics.
It is worth noticing for the sake of justice that the possibility to
transform kinetic equations (4.4.1) into diagonal form (14.1.5) containing
the only motive force does not contradict TIP according to which the only
thing necessary and sufficient for that is the linearity of laws (4.4.1) and the
symmetry of phenomenological coefficient matrix therein (S. De Groot, P.
Mazur, 1964). However, this procedure provided in TIP by the linear
transformation of flows and forces does not advantage at all since it does
not reduce the number of the independent phenomenological coefficients Lij
in the initial equations and does not remove the constraints intrinsic for
linear TIP. Unlike TIP, thermodynamics allows to directly find the
generalized form of the Fourier’s, Ohm’s, Darcy’s, Fick’s and Newton’s
equations by substituting the forces therein for the more general (resultant)
motive force Fi or Xi.
The existing of such forces assumes they have the components Fij =
ΘjXij as it takes place in the anisotropic heat conduction and electric
conduction equations. Therefore laws (14.1.5) may be written in the form:

Ji = Кi (Fj) Σi Fij , (i, j = 1, 2, ..., n). (14.1.6)

It is a matter of no small importance that the components Fij of the

resultant force Fj in equation (14.1.6) have the same dimensionality and
260
unitary physical meaning of force in its traditional (Newtonian) conception.
This allows their summing in any assortment as applied to phenomena of
various nature and, thus, finding a resultant force for various poly-variant
systems. Naturally, such “diagonal” laws may be transformed to their initial
(matrix) form (4.4.1) by placing the coefficients Кi(Fj) inside the sum sign
and substituting the product Θi Кi(Fj) for the coefficient Lij. The
“phenomenological” coefficients Lij thus obtained will naturally be any
more neither pure kinetic nor pure thermodynamic values, which is
particularly emphasized in TIP (S. De Groot, P. Mazur, 1964). However, at
such an “inverse” transformation the possibility will be lost to investigate
separately the impact of thermodynamic and kinetic factors upon the
transfer process, which itself is no small importance. We will see hereafter
that the “diagonal” form of transfer laws (14.1.6) allows cutting the number
of the pure kinetic coefficients in the transfer equations and makes
superfluous the assumption of their constancy. However, the main
advantage of such an approach lies in the possibility to express the
superposition effects exclusively in terms of thermodynamic variables and
to give them a simpler interpretation as a result of the cross compensation
between the non-matching components of the resultant force. Furthermore,
with the resultant motive force of a transfer process found, the impact of
kinetic and thermodynamic factors upon the process may be separately
investigated in a number of cases.

14.2. Superposition Effects as Found from Partial Equilibrium

Conditions

Macro-processes in complex (poly-variant) physicochemical, biological,

ecological, etc. systems having many degrees of freedom can practically
never cease simultaneously for all degrees of freedom intrinsic to them.
Usually first comes equilibrium of a one (ith) kind characterized by ceasing
the ith process followed by equilibrium of the jth kind and so on. The state
characterized by ceasing a part of processes is often termed “incomplete”,
“current” or “intermediate” equilibrium. We will term it “partial equilibrium
of the ith kind”, which will allow specializing its type (thermal, baric,
diffusive, electrical, etc.). Such is the state of mechanical (more definitely –
baric) equilibrium occurring well earlier than other thermodynamic
processes and characterized by no cubic strain of subsystems or the system

261
as a whole1), as well as thermal equilibrium in systems with a slower
diffusive process taking place. Unlike the stationary state of the kth order,
where external forces maintain the invariability of parameters in time,
partial equilibrium does not demand external forces to be applied and the
parameters in this state are variable.
The notion of stationary state order is known to have been introduced
into practice after I. Prigogine proved (1960) a theorem reading, “minimal
generation of entropy corresponds to a state where the flows Jj associated
with the non-fixed forces Xj disappear”. If in a system described by n
independent forces Xi (i =1,2,…, n) k of them are maintained constant (with
the help of whatever external effects), such a state is termed the stationary
state of the kth order. According to this definition, when the forces Xj
numbered k+1, k+2, etc. are not fixed, the flows Jj matching them disappear,
and the system passes spontaneously to a stationary state of less order (with
less entropy generation) subsequently until it reaches the zeroth-order
stationary state, viz. equilibrium (with the zeroth entropy generation).
Thereupon all superposition effects arising in the system with disappearance
of whatever ith flow Ji started to be termed stationary effects. Their formal
difference from equilibrium conditions is that there are the
phenomenological coefficients Lij presenting in their analytical expressions,
whereas the classic equilibrium conditions are expressed exclusively in
terms of thermodynamic variables.
It is easy to understand why the coefficients Lij appear at the stationary
state conditions if to find them from Onsager’s phenomenological laws
(4.1.6). For the simplest case with two flows Ji and Ji these equations are:

Ji = Lij Xi + Ljj Xj, (14.2.1)

Ji = Lji Xi + Ljj Xj . (14.2.2)

From (14.2.2.) it follows, in particular, that for the state with Ji = 0 the
stationary effect is expressed by the relationship:

(Xj / Xi)ст = – Lij / Ljj . (14.2.3)

1)
Such specialization is needed to distinguish it from the equilibrium inside a liquid flow
with layers moving with the same velocity (without momentum interchange), which refers to
the mechanical equilibrium category, too.

262
 As follows from (14.2.3), the multiple superposition effects for
different-kind processes are expressed in TIP via empirical coefficients
depending on kinetic factors which value, unlike the thermodynamic
parameters, are not known beforehand. Therefore TIP is unable to predict
value of these effects. However, even more important is the fact that for
isolated systems, bio-systems, oscillatory processes, media chemically
reacting with reactions simultaneously running, etc., non-stationary states
are inherent. The existing stationary TIP is inapplicable to such systems.
This forces to search for other methods of investigating the superposition
effects therein arising. Such are the methods based on the partial
equilibrium conditions, which may be as well found for systems non-
stationary in whole. These methods are based on the phenomenological laws
represented in “diagonal” form (15.1.5). It follows from these equations that
the ith kind process ceases (Ji = 0) when the components Fij of the resultant
force Fi mutually compensate each other:

Fi = ΣjFij = 0, (14.2.4)

In the particular case of j = 1,2 Fi = ΘjXj + ΘjXj = 0, then we obtain:

Fi = ΘjXj + ΘjXj = 0, (14.2.5)

which gives

(Xj / Xi)ст = – Θj /Θj . (14.2.6)

As will be shown hereafter, such a structure is attributed to

unexceptionally all superposition effects described within thermodynamics.
These relationships do not contain anymore the kinetic coefficients and,
therefore, refer to the partial equilibrium conditions as if equilibrium in the
field of centrifugal forces or the gravity field. The possibility of
investigating non-stationary processes substantially extends the TIP
applicability.

11.2. Value of Superposition Effects as Predicted from Known

Thermodynamic Parameters

Characteristic features of the method proposed may be conveniently

considered for a rather general class of phenomena involving diffusive
263
mixing of the kth substance in a closed heterogeneous system with invariable
volume. In this continuous system under consideration, along with the
external heat exchange process, the internal processes of thermal
conduction, electric conduction and diffusion may also run. Let us
specifically consider the diffusion of the kth independent component within
such a system. The diagonal form of kinetic equation for such a process
hereafter termed for short as thermodynamic form looks like:

Jk = Кk Σi Fki = LkXk. (14.3.1)

In accordance with (4.5.6) the integral motive force of this process is

understood as a difference of diffusive potential of the kth component Xk = –
Δζk. To expand the expression for Xk, we will use the relation between the
diffusive and chemical potentials ζk = μk + Tsko (6.3.6). From here:

dζ k = dμk + Tdsko . (14.3.2)

Considering (14.3.2) with the expression of exact differential of the

chemical potential p,T,сℓ :

dμk = Σk(∂μ k /∂cℓ )dcℓ – sk dT + υkdp, (14.3.3)

gives that at p,T = const the concentration relationship of the diffusive

potential coincides with the similar relationship of the chemical potential,
dζ k = dμk, since skо = skо(p,T):

[dζk]p,T = [dμk ]p,T = Σk μkℓ dcℓ , (14.3.4)

where μkℓ – abridged symbol for the derivative ∂μk/∂cℓ . Due to this in the
general case:

dζk= dζk= Σk μkℓ dcℓ – (sk – sko)dT + υk dp . (14.3.5)

From here the extended expression of resultant diffusion motive force

ensues:

264
 Fk = – [Σk μkℓ∇cℓ + (sk – sko)∇T – υk∇p] . (14.3.6)

Here Fkℓ = Σkμkℓ∇cℓ – component of the resultant force Fk responsible for

usual (concentration) diffusion; FkТ = (sko – sk)∇T – component responsible
for thermal diffusion (substance transfer due to temperature gradient); Fkр =
υk∇p – component responsible for pressure diffusion (substance transfer due
to pressure diffusion).
Thus the kinetic equation of diffusion in extended form is:
Jk = Кk [Σk μkℓ∇cℓ + (sko – sk)∇T + υk∇p] . (14.3.7)

Let us consider for example the Soret effect – a concentration gradient

∇ck arising in a binary mixture being in mechanical equilibrium (∇p = 0),
where a temperature gradient ∇T has been created. This phenomenon means
the absence of balance between the components FkТ and Fkр of the resultant
force Fk, which causes a flow of the independent kth component. When
partial equilibrium occurs (Jk = 0), the expression for stationary Soret effect
directly follows from (14.3.7):

(∇ck /∇T)ст = – qk∗/Tμkk , (14.3.8)

where qk∗ = T(sk – skо) – the so-called heat of kth component transfer; μkk =
(∂μk/∂ck). The expression reciprocal to the above one:

(∇T/∇ck)st = – Tμkk/qk∗ (14.3.9)

characterizes the stationary Dufour effect (temperature difference arising as

a result of component diffusion).
Similarly from (14.3.6) the so-called thermo-mechanical effects in
continuums may be predicted as showing in a gas or liquid flow arising
under the temperature difference. One of them is Knudsen effect arising in
vessels with initially equal ideal gas pressure after communicating them
with a capillary tube. The gas overflow ceasing, stationary temperature and
pressure differences set in in the vessels (in the capillary – temperature and
pressure gradients). Relation between them (stationary Khudsen effect) may
be determined directly from (14.3.6). In this case ∇ck = 0, and due to mutual
compensation of the last two terms it follows that

265
 (∇p/∇T)st = – qk∗/Tυ. (14.3.10)

This effect is explained from thermodynamics by a distinction between

the entropy sk in the pores or capillaries of the membrane and its value
beyond them (in the vessels) sko. This is most evident from Knudsen effect
(1910) when the pressure difference Δp disappears with increasing the
diameter of the capillary or the width of the gap connecting the vessels with
gas of different temperature (Haase, 1967). In this case the invariably
negative value of the transfer heat qk∗ in Knudsen effect evidences that the
entropy sk of the gas transferred thru the capillary system is less than the
entropy sko of the same gas in its “free” state. This is natural since the gas
transferred thru the capillaries lacks a part of the degrees of freedom in
mechanical motion of molecules (i.e. it is more ordered).
In the same way the sedimentation effect may be predicted widely used
at the uranium enrichment in centrifuges. The component separation is
caused therein by a pressure gradient ∇p arising in binary solutions as
generated by centrifugal forces. In the state of sedimentation equilibrium in
the centrifugal field the first and the third terms of expression (14.3.6) are
mutually compensated at ∇T = 0, which directly gives the stationary effect
expressed as:

(∇ck/∇p)st = – υk /μkk . (14.3.11)

It is significant that this result has been obtained within the linear theory
of irreversible processes also from the equilibrium conditions, whereas
Soret, Dufour, Knudssen, etc. effects – from the stationary state conditions.
This tells to the TIP inconsistency that evidences also from the fact that the
said effects maintain their character also after the independence of both
flows (heat and substance) has been provided.
Thus the method proposed allows predicting the value of the
superposition effects in the states of partial equilibrium from the
thermodynamic parameters Θj and Θj known (V. Etkin, 1999) or finding the
hard-to-measure thermodynamic parameters of the sk, μkℓ, etc. type as based
on measuring the stationary superposition effects of the (14.2.6) type.

266
 14.4. Simplification of Phenomenological Laws without Onsager’s
Reciprocal Relations Applied

As shown in Chapter 4, the application of TIP to investigate the effects

of superposition (interaction) of irreversible processes having various nature
starts off with setting up the balance equation for mass, momentum, charge,
angular momentum and energy in order to subsequently identify therein the
terms defining the entropy rise rate in a system under investigation and to
find from them the cofactors that may be interpreted as the generalized rates
of irreversible processes (the flows Jj) and their motive forces Xi. To set up
such equations, thorough knowledge of the entire complex of scientific
disciplines is required, which equations contain (unlike thermodynamics)
time as a physical parameter and which processes effect the entropy
generation. Then based on these balance equations a similar entropy
equation of the (4.3.1) and (4.3.2) type is formulated followed by
identifying therein terms associated with the entropy “generation” and
dividing them into flows and forces. Such a division into cofactors may be
implemented by different ways, which supposes certain arbitrariness in
choosing flows and thermodynamic forces, their dimensionality and the
meaning associated. Next step is to set up Onsager’s kinetic equations
(phenomenological laws) (4.1.6) for particular processes under
investigation. These equations are then considered jointly with Onsager’s
symmetry conditions (4.1.7) or (4.1.8) which interrelate the flows Ji and Jj.
And only after that expressions of the so-called “stationary superposition
effects” are determined as corresponding to disappearance of one of the
flows Ji or Ji with the non-equilibrium state of the system maintained.
Thus the determination of the TIP-based superposition effects is a quite
complicated multi-stage problem involving profound knowledge of many
special disciplines. However, this but adds to the proposed method that
allows finding the shortest way to solution to the problem. This becomes
possible due to the fact thermodynamic equations of the (2.5.9) type already
contain time, flows and thermodynamic forces and, therefore, do not need a
clumsy form of balance equations for mass, charge, momentum, energy and
entropy. Besides, thermodynamics eliminates whatever arbitrariness in
choosing motive forces and generalized rates of any (both scalar and vector)
processes – their meaning and dimensionality are unambiguously defined by
the character of the coordinate Θi as a quantitative measure of the particular
energy form carrier. In this case, according to (2.5.8), the thermodynamic

267
forces are expressed exclusively in terms of the negative gradients of the
generalized potential associated with the Θi, while the flows Ji or ji are the
product of these values and their transfer velocity under the action of the
forces Xi.
Furthermore, thermodynamics considerably simplifies the
phenomenological laws reducing them to the so-called “diagonal form”
(14.1.5) which contains minimal number of the kinetic coefficients K ij .
These laws are based on the statement proved in the previous paragraph and
reading that for each independent process the only (resultant) force exists
which disappearance ceases the process of this kind.
Let us consider for example the equations of multi-component isobaric-
isothermal diffusion, for which L.Onsager proposed (1945) the
phenomenological laws of the kind:

ji = –Σj Lij ∇μj , (i, j = 1, 2, …, K–1). (14.4.1)

This equation assumes that the flow jk of any of the K–1 independent
(emphasis added) components of a system, nevertheless, depends on all
thermodynamic forces acting in the system, which are, in the case of
isobaric-isothermal diffusion, identified with the negative gradients of
chemical potential of each of such jth components μj. The additional (non-
diagonal) terms i ≠ j of the first sum of this expression were introduced by
L. Onsager to allow for the interrelation between flows, which he used to
explain the “ascending diffusion”, viz. transfer of a substance in the
direction of its concentration increasing.
Since the chemical potential of any of the jth substances is a function of
temperature T, pressure p and concentration ck of the independent kth
components, its differential dμj at р,Т = const features exclusively its
dependence on concentration, which may be expressed as:

dμj = Σk (∂μj/∂ck)dсk = Σk μjk dсk , (14.4.2)

where μjk – abridged symbol for the derivative ∂μj/∂ck. Therefore equation
(14.4.1) has actually the form:

ji = –Σj Lij Σk μjk ∇сk , (i, j, k = 1, 2, …, K–1) (14.4.3)

268
i.e. contains a double sum of terms, which sharply complicates determining
the already hard-to-determine cross phenomenological coefficients. The
formal simplification of these equations as

ji = – Σj D΄ik ∇сk , (i, j, k = 1, 2, …, K–1) (11.4.4)

by introducing the diffusion coefficients

D΄ik = Σk Lijμjk (14.4.5)

does not matter since keeps invariable the former number of the kinetic Lij
and thermodynamic μjk factors influencing the diffusion of the ith
component. The double sum in (14.4.3) with the number of summands K(K–
1) results in an extremely complex relationship between the said values and
makes the statement of the problem of finding the non-diagonal diffusion
coefficients D'iℓ in metals and alloys with the existing methods of
experimental determining the fields of impurity concentration (including the
X-ray diffraction analysis) mathematically incorrect (Krishtal, Volkov,
1985). This forces investigators to apply a number of assumptions. When
investigating the diffusion in metals, both sums in (14.4.3) are most often
neglected and the so-called Birchenall-Mehl approximation is used:

ji = – (Lij RμT/ai ) ∇ai = – Di∗∇ai , (14.4.6)

or Darken approximation

ji = – Di∗(∂ai /∂ci )∇ci = – Di∇ci , (14.4.7)

wherein the chemical potential μi = μio + RμT lnai is represented by a known

way in terms of its standard value μko and the activity ai of this component
by Lewis, while the dependence of the ith diffusive flow on the
concentration gradients of other components is allowed for indirectly in
terms of the “true” Di = LiiRμT/ai or “effective” Di = Di*(∂ai/∂ci) diffusion
coefficients with the help of a number of empirical or semi-empirical
relationships (M. Krishtal, 1972). Without these additional relationships
equations (14.4.6) and (14.4.7) can not describe the ascendant diffusion
phenomenon, concentration discontinuities on the welding border, bulk
effects like Kirkendal effect (replacement of diffusion pair border), etc.
269
Therefore it is a matter of interest to compare the simplifying assumptions
proposed with those ensuing from thermodynamics.
Let us consider for example an arbitrary discontinuous system like the
diffusive welding pair on Fig.14.1). Such a pair is formed as a result of
vacuum diffusion welding of two plates having different compositions. One
of the subsystems (left-hand) is a three-component system containing, along
with iron (Fe), also chromium (Cr) and carbon (C), while the other (right-
hand) – iron (Fe) and carbon (C). The dots on the figure denote the
experimentally found atomic concentrations of the components.
As follows from Fig.14.1, in the course of high-temperature annealing
%
of the system under
Fe + C + Cr Fe + C consideration (Fe + Cr + C) –
0.6 (Fe + C) carbon transfer is
Carbon content

observed in the direction of

Welding area
0.4 carbon concentration increase
(carbon ascending diffusion)
0.2 leading to a carbon
X concentration discontinuity in
–0.4 –0.2 0.2 0.4 the diffusion we lding area with
δ a thickness of δ. This
Distance from welding area
Calculation Experiment
discontinuity can not be
anymore explained by the
Fig.14-1. Carbon Ascending Diffusion in Alloys generation of the so-called
“dissipative structure”
(according to the Prigogine’s terminology) supported by the irreversible
processes running in the system since in this case the system in whole is
closed and none of the thermodynamic forces –∇ck is artificially maintained
therein. This means that the steady-state concentration distribution among
the substances in the system refers indeed to partial equilibrium.
Let us apply the thermodynamic approach to describe the phenomenon
of ascending diffusion. From the positions of thermodynamics there is the
only motive force existing for a flow of any independent component ji,
which engenders this flow with all the associated superposition effects. For
the isobaric-isothermal diffusion in continuums according to (14.3.5) the
negative gradient of the diffusion potential ζk is the resultant force of the
process. However, since ζk = μk + Tsko – рυko, then at р,Т = const the
diffusion and the chemical potentials have the same dependence on
concentration because [dζk]p,T = [dμk ]p,T. Therefore the negative gradient of

270
the chemical potential of the ith component may still remain the motive force
of the concentration diffusion, i.e. XD = – ∇μi. Since this potential is the
function of concentrations of all independent system components, the Fick’s
generalized diffusion law takes the form:

ji = – LiiΣk μik∇сk = – Σk Dik∇сk, (14.4.8)

where Dik= Liiμik – generalized diffusion coefficients. This expression

features much simpler form of diffusion coefficients, which consequences
allow a direct experimental check with the up-to-date means of investigating
diffusion in metals. One of such consequences is the simple relationship
between the thermodynamic μik and kinetic Lij factors of diffusion:

Dii/μii = Dik /μik . (14.4.9)

The relationship of such a kind was set up earlier in Darken

approximation (Brown, Kirkaldy, 1969). Let us use now this condition of
the simultaneously disappeared flow ji and its thermodynamic force XD.
This condition should be referred to the partial (incomplete) equilibrium
state since it occurs for components relatively fast diffusing with the
redistribution of other components (in this case – chromium) going on. This
state features, as experiment shows (Fig.14.1), a carbon concentration
discontinuity on the welding border. It may be more conveniently described
if changing to the integral form of the diffusion motive force XDс = –Δμk =
Σk μik Δсk, where μik – mean value of the coefficient μik. Then from
expression (14.4.9) it may be found:

Σk μik Δсk = 0 . (14.4.10)

In the particular case of three-component systems with the third

component slowly moving:

D23 = D22 (c2΄ – c2˝)/(c3˝– c3΄) , (14.4.11)

where c2΄ , c2˝ , c3΄ , c3˝ – impurity concentrations on both sides of the
welding border.

271
 This relationship evidences the generation of the concentration
discontinuities (c2΄ – c2˝) in the fast diffusing substance when its equilibrium
distribution sets in, which is experimentally confirmed. One more
relationship for the diffusion pair under consideration may be obtained
based on the expression:

Δμi = – Σk μik Δсk . (14.4.12)

When comparing (11.4.12) with the similar expression Δμi = – μik Δсi
for a binary system and using relationship (14.4.9) one can find an
expression for the so-called “distribution coefficient” of the ith component
ci΄/ci˝, which is the relationship between its concentrations in the binary
system and that under investigation:
æi = 1 + Σk χik ck , (11.4.13)

where χik = Dik/Diick – the so-called parameter of cross-impact between the

ith and the kth components.
This expression differs from a number of empirical relationships such as
Wagner formula (Krishtal, 1972) and is a matter of principal interest since
directly relates equilibrium concentrations and the hard-to-measure
diffusion coefficients Dik (V. Etkin, 1994). To check it, let us apply to the
amply available data on the carbon distribution in the system (Fe + 6.5%Cr
+ 0.4%C) – (Fe + 0.49%C) after the 5-hour annealing at a temperature of
1,0000C (see Table (14.1):

Table 11.1
Carbon Distribution Coefficient in Austenite

Content of components, Experiment, Calcula

atomic percentage atomic percentage tion
C, Cr,
[1] [2] [3] Mean
Iron Alloy Alloy
2.28 7.81 6.38 0.35 0.57 0.38 0.44 0.47
4.76 6.58 0.52 0.46 0.41 0.46 0.44
1.15 2.85 6.70 0.43 0.41 0.40 0.41 0.42
0.74 1.64 6.78 0.47 0.38 0.41 0.42 0.41
272
 The C (second component) – Cr (third component) cross-impact
coefficient remains practically constant within the given concentration range
and equal to χ23 = –10,9 (Chipman, Bruch, 1968). The calculated f values
have been found from expression (11.4.11) and compared in this table with
the data from three independent experiments (M. Krishtal, 1972; Chipman,
Burch, 1968; Shenk, Kaiser, 1960). As follows from the table, the calculated
and mean measured data complies quite satisfactorily. It is worth noting that
this result was earlier obtained by balancing the activities (supposing the
activities ak΄ and ai" being equal in both parts of the pair at partial
equilibrium) and confirmed experimentally on a number of metallic alloys
Fe-based 1), which is illustrated by Fig.11.1 (Krishtal, 1972).
As another example, let us consider a three-component system Fe + Si +
Cr that is a matter of not only theoretical, but also practical interest in the
context of siliconizing of chromium steels. Diffusion pairs (Fe + Si) – (Fe +
Cr) with different content of silicon and chromium were investigated.
Experimental value æ was determined as the ratio of the Si concentrations in
iron and alloy; the ratio D23 /D22 was found from expression (14.4.10). The
calculation and experimental data is shown in Table 14.2.

Table 14.2
Calculated and Experimental Values of Silicon Distribution Coefficient
in Welded Diffusion Pairs

Parameter Distribution
Content Ratio
coefficient
Si, % Cr, % D23 /D22 χ23 Calculation Experiment

1.0 6.7 0.03 3.00 1.20 1.30

2.0 6.7 0.07 3.50 1.20 1.2
3.0 6.7 0.11 3.66 1.26 1.33
4.0 6.7 0.15 3.75 1.26 1.33
3.0 1.0 0.10 3.33 1.04 1.15

1)
It is easy to note that the activities balancing method proceeds from the same assumptions
as thermodynamics
273
 It should be emphasized that theoretical calculations of such systems
forming substitution solutions are especially complicated; therefore the
possibility to find “cross” diffusion coefficients featuring their interference
is a matter of no small consequence.
The data of the above table evidences quite exact compliance between
the experimental and calculated æ values. The minor systematic deviation of
the calculated æ values from the experimental ones may be explained by
neglecting the c3' value calculated from (14.4.11). Thus the data of the
considered experiments confirms that the diagonal form of the multi-
component diffusion laws includes the whole information of superposition
effects despite it is much simpler than that Onsager postulated. Thereby the
conclusion of thermodynamics is once again confirmed that the motive
forces in non-related irreversible processes are unique (Etkin, 1989) and the
Onsage’s postulate should be superseded for such processes by a statement
each flow depends on only all components of the resultant thermodynamic
force. Since all summands of this force Xi have the same tensor order, such
a revision of the physical concept of Onsager’s postulate intrinsically
corresponds to Curie principle that excludes the interrelation between
phenomena of different (odd) tensor order. This is such (resultant) force that
should have been named thermodynamic since the fact itself of its existence,
as well as its particular form for each of the independent processes is
defined by the basic equation of thermodynamics. Such a force found allows
avoiding the “over-determination” of the flow given in expression (14.4.2)
as a function of thermodynamic forces and retaining the simple form for
known laws of thermal conduction, electric conduction, diffusion, filtration,
momentum transfer at their generalization to the case of superposition of
different-kind phenomena.

14.5. Investigation of Membrane Processes with the Reciprocal

Relations Violated

The Onsager’s symmetry conditions Lij = Lji or Onsager-Cazimir’s anti-

symmetry conditions Lij = – Lji are known to be violated (S. De Groot, P.
Mazur, 1974) providing the coefficients Lij and Lji are inconstant, i.e.
depend on the parameters of the system, in particular, the thermodynamic
forces Xi and Xj. This fact significantly restricts the existing TIP

274
 applicability since the existence of such dependence is doubtless. The
situation changes when the transfer equations are represented in diagonal
form (11.1.6), while the superposition effects are determined from the
partial equilibrium conditions. In this case, as shown hereinbefore, the
kinetic coefficients are cancelled out and the superposition effects are found
from the condition of mutual compensation of the resultant force
components without appeal for the reciprocity relationships to be used.
Let us show this by a quite general example of multi-component
systems divided into two subsystems with a porous partition, capillary,
valve, finite-thickness membrane, etc (Fig.14.2).
The partition has generally finite dimensions and is, therefore,
considered as one of the subsystems within such a “combined”
(discontinuous-continuous) system. The system in whole is a spatially
heterogeneous medium featuring temperatures T, pressures p,
concentrations ck, etc. continuously varying across the thickness of the
membrane, whereas distribution of temperatures, pressures and
concentrations outside the membrane being practically homogeneous
(intrinsically equilibrium) (respectively, T', p', ck' and T", p", ck", etc.).
2 3
1 The effects arising in such a system
T΄ P˝
is more conveniently to be
T(r,t) Subsystem 2
j΄ considered starting off with the
Subsystem 1

Subsystem 3
jv
jq Xq Xu P(r,t) simplest case of a single-component
j(r,t) X Xk C˝
jk
e j e medium (ck′ =ck″ = , υk = υ) having
p΄ T˝ intrinsically the same pressure on
c΄ ck(r,t) j˝ both sides of the porous partition (p′
=p″). With the temperature
difference (ΔT ≠ 0) generated in
Fig.14.2. Combined (Discontinuous- such a system a gas or liquid flow
Continuous) System
1 – membrane (continuum)
across the partition arises resulting
2, 3 – reservoirs of heat and substance in pressure difference. This
phenomenon was first described by
Feddersen (1873), who observed a flow of air across a plaster wall in the
direction toward the higher temperature and termed it thermo-diffusion.
This phenomenon is presently known as thermo-osmosis. The thermo-
osmosis ceases with both components of the osmotic force, FkТ = (sko – sk)ΔT
and Fkр = – υkΔp, mutually balanced. From here, given the relationship (sko –
sk) = qk∗/T (where qk∗ – the so-called transfer heat of the kth component), a
275
known expression of Feddersen stationary effect directly follows (Groot,
1956):
(Δp/ΔT)st = – qk∗/Tυk . (14.5.1)

There is the inverse phenomenon also known, viz. a temperature

difference arising across the two sides of a partition with air or other gas
being forced thru it. Both these effects have the same nature with Knudsen
effect (1910), as well as with Allen - Jones effect (1938) in liquid helium II
(fountain effect) consisting in overflow of helium out from a bulb closed up
with a porous cork under a minor heating (Haase, 1967). Assuming for this
case (London, 1938) that the superfluid component of helium II passing thru
the porous cork or capillary has zero entropy (sko = 0) expression (14.5.1)
becomes:

(Δp/ΔT)st = sk/υk , (14.5.2)

which corresponds to the result London earlier obtained by the “pseudo-

thermostatic” way. There is the inverse phenomenon also known, viz. a
temperature difference arising with a pressure difference generated across
the two sides of a partition, which was termed as mechanocaloric effect
(Daunt-Mendelson).
In isothermal systems (ΔT = 0) under a pressure difference Δp generated
across a membrane the phenomenon of reverse osmosis arises, viz. binary
solution separation with extracting the kth component (usually solvent)
therefrom. This phenomenon is being applied in ever growing scale in water
purification units. The stationary concentration difference of the kth
component arising therein is described as:

(Δck /Δp)st = – υk /μkk . (14.5.3)

There is the inverse phenomenon also known, viz. an osmotic pressure

difference (osmotic pressure) Δp arising at redistribution of the kth
component, which plays an important role in biological systems. Both these
effects do not include the transfer heat and are reversible.
Now let us consider even more complicated case when the membrane is
permeable (though in different degree) for both of the components. There
are the flows of the 1st and 2nd components J1 and J2 arising in such a

276
system, for which the thermodynamic transfer laws, due to Δc2 = – Δc1, take
the form:

J1 = – L1 [Σkμ11Δc1 + (s1– s10) ΔT + υ1 Δp] ; (14.5.4)

J2 = – L2 [Σkμ21Δc1 + (s2 – s20) ΔT + υ2 Δp] . (14.5.5)

The separation of a mixture in such systems into components due to a

temperature difference maintained is sometimes termed as thermo-effusion
to distinguish this phenomenon from the thermo-diffusion (redistribution of
components in the absence of membranes and convective gas or liquid flow
(S. De Groot, 1956). Given the interrelation of the chemical potential
derivatives μ21 = – (c1/c2)μ11 ensuing from Gibbs-Duhem relationship at p,T
= const and solving the set of equations (14.5.4)…(14.5.5) for the Δc1 and
ΔT the expression of stationary thermo-effusive effect may be found:

(Δc1/ΔT)ст = (υ1q2∗ – υ2q1∗)/Tυμ11 . (14.5.6)

Similarly solving the same set of equations for Δp and ΔT and given ( s10
– s1) = q1∗; ( s20 – s2) = q2∗ the expression for thermo-mechanical effect may
be found (R. Haase, 1967):

(Δp /ΔT)st = – (c1 q1∗ +c2 q2∗)/ υ , (14.5.7)

where υ = c1υ1 + c2υ2 ..
All these results comply with those obtained within the frames of TIP.
However, there was no need to apply now to the Onsager-Cazimir’s
reciprocity relationships and assume the linearity of phenomenological
laws, as well as the constancy of phenomenological coefficients (their
independence on the thermostatic parameters). On the contrary, it is quite
evident that the processes in the system considered are non-linear since the
coefficients of thermal conduction, electric conduction, diffusion, filtration
and viscous friction in Fourier’s, Ohm’s, Fick’s, Darcy’s and Newtons’s
empirical laws depend on temperature, pressure, composition and number of
other parameters varying across the thickness of the membrane. This
violates their constancy requirement which in the Onsager’s theory bears the
principal character and constitutes a substantial part of his law (S. De Groot,
P. Mazur, 1964). In fact the forces Xi in his theory were determined by a
277
deviation of the system parameters ψi (temperature, pressure, component
concentrations, etc.) from their equilibrium values and were, therefore,
functions of these parameters. In such a case the dependence of the
coefficients Lij on these parameters implies their dependence on also the
forces Xj, i.e. the non-linearity of phenomenological laws (5.1.6). This
violates the conditions under which the statistic-mechanical substantiation
may be applied to symmetry of phenomenological coefficients matrix as
used by Onsager and, thus, excludes the possibility of using these
coefficients in the membrane processes. It is even more important that the
method proposed does not demand the application of the Onsager’s
relationships violated in such systems. The analytic expressions for the
superposition effects herein obtained do not depend on the fact whether the
phenomenological laws (see 5.1.7) are linear or not and whether the
Onsager-Cazimir’s reciprocity relationships are observed or not therein.
This means that the said effects bear a more fundamental character.
The fact attracts attention as well that all of the said effects reach their
peak values at stationary states when one of the flows (mostly the flow of
the kth component) disappears and therefore just can not superimpose on the
remainder flows. This once again evidences the inconsistency of the TIP-
associated interpretation of these effects as a result of superposition
(interaction) of the flows. As appears on close inspection, these effects arise
due to superposition of motive forces, i.e. in exactly the same way as this
occurs in mechanics and other disciplines. The principal difference between
these two interpretations is that the superposition effects in thermodynamics
are functions of state, but not of process, and arise from the partial
equilibrium conditions, but not from whatever order stationary state
conditions.

278
 Chapter 15
GENERALIZATION OF THE THEORY TO SYSTEMS STANDING
FAR AWAY FROM EQUILIBRIUM

This chapter is dedicated to thermokinetic description and investigation

of a number of interrelated non-linear transport processes based on the
method offered in the previous chapter. The method basically consists in
defining the superposition effects of irreversible processes from the partial
equilibrium conditions. In the case of interrelated processes it is realized by
using the differential reciprocity relationships allowing for the contribution
of an alien force to a phenomenon under investigation and thus making such
processes independent. This allows finding for each of them a resultant
motive force which disappearance ceases the process. Such an approach
enables the further (relative to TIP) reduction of the number of kinetic
factors in these laws and the expression of the superposition effects in the
same way as in Chapter 11 for independent processes.
The efficiency of the method offered and the validity of the conclusions
obtained from thermodynamics will be here confirmed on a numerous
experimental data.

15.1. Reciprocal Relations for Chemical Reactions

Classic thermodynamics with its laws considers only initial and final
states of a chemically reacting system not touching upon kinetics of
chemical reactions, i.e. the rate of the initial-to-final state transit for the
system. The factors affecting this rate, (such as temperature, concentration
of reagents, catalysts available, etc.) have been subject of chemical kinetics.
This postulates, according to the Guldberg & Waage mass action law, that
the rate of a chemical reaction is proportional to the product of reagents’
concentration. Since any chemical reaction runs in both forward and
backward directions, its resultant rate is defined by the difference between
the rates of direct ωr' and reverse ωr" reactions:

wr = ωr΄ – ωr˝ = ωr΄ (1– ωr˝/ωr΄ ) . (15.1.1)

279
 According to the detailed balance principle a reaction stops when the
rate of direct reaction becomes equal to that of reverse reaction. In this case
the affinity of the reaction Ar becomes equal to zero. From this it follows
that

Ar = RμT ln (ωr˝/ωr΄) , (15.1.2)

where Rμ – universal gas constant.

Considering (15.1.1) and (15.1.2) jointly the kinetic laws of chemical
reactions may become exponential dependences of the type:

wr = ωr΄ [1– exp (–Ar /RμT)] . (15.1.3)

These equations term the Guldberg & Waage laws. The dependence of
ωr' on temperature within a quite wide range is here described by the
Arrenius empirical law:

ωr = ωrо exp (–Ea/RT) , (15.1.4)

where ωrо – Arrenius factor; Ea – activation energy required to initiate

reaction
This non-linearity of chemical reactions leads to a violation of the
Omsager-Cazimir’s reciprocity relationships therein with the result that the
applicability of TIP to chemically reacting systems is restricted to next-to-
equilibrium states. The thermodynamic substantiation of more general
differential reciprocity relationships (4.5.3) offered in Chapter 4 allows
overcoming this restriction. Let us consider for example the case of
interrelated chemical reactions obeying the Guldberg & Waage chemical
kinetics laws (15.1.3). For this let us apply to the class of ternary
unimolecular reactions L. Onsager earlier considered (1931). These
reactions with substances L, M, N may be described by two different ways:
either as a two-linear-independent reaction model (R. Haase, 1967):

L ↔ M (reaction 1);

M ↔ N (reaction 2), (15.1.5)

or as a linear-dependent elementary reaction model (see Fig.15.

280
 M
L↔M (reaction 1),
M↔N (reaction 2), (15.1.6)
w1 w2 N↔L (reaction 3).
A1 A2 The first model features a purely macro-
approach, where a merely general process is
described with the help of mathematically
A3
L N independent equations. The second model
w3 describes a reaction mechanism, i.e. a real
Fig. 15.1. Ternary
chemical conversion process (R. Haase,
unimolecular reaction 1967). The above elementary reactions run
with rates w1, w2 and w3, respectively, being
expressed, according to the Guldberg & Waage laws, by the following
exponential kinetic equations:

w1 = ω1΄ [1 – exp(–A1/RμT)] ;
w2 = ω2΄ [1 – exp(–A2/RμT)]; (15.1.7)
w3 = ω3΄ [1 – exp(–A3/RμT)] ,
where ω1΄ , ω2΄ , ω3΄ – rates of the corresponding direct reactions; A1, A2, A3
– their current affinities interrelated to the Gess law as A1 + A2 = A3.
According to these equations the dissipation function Tσs for the system
under consideration is:
Tσs = w1 A1+ w2 A2 + w3 A3 . ( 15.1.8)

For the considered case the rates of the reactions and their affinities are
linear-dependent (i.e. represent a linear combination of each others). At
such conditions the symmetry conditions are known not to be guaranteed
(De Groot, 1956). Therefore let us introduce two new linear-independent
rates:

wα = w1 + w2 ; wβ = w2 + w3 . ( 15.1.9)

Two independent forces Aα and Aβcorrespond to these rates, which

allows equations (15.1.7) to be transformed, subject to the invariance of the
dissipation function Tσs, to the form:

281
 wα = L11 [1 – exp(–Aα /RμT)] + L12 [1 – exp(–Aβ /RμT)], (15.1.10)
wβ = L21 [1 – exp(–Aα /RμT)] + L22 [1 – exp(–Aβ /RμT)]. (15.1.11)
Here
L11 = ω1΄ + ω2΄; L12 = ω3΄ exp(–Aα /RμT )] ; (15.1.12)
L21 = ω3΄ exp(–Aβ /RμT )]; L22 = ω1΄ +ω3΄ . (15.1.13)

Thus a ternary reaction far from equilibrium may be described by two

non-linear kinetic equations with linear-independent rates and forces. Here
the coefficients ω1΄, ω2΄ and ω3΄ being functions of temperature, pressure
and concentrations of parent substances for the corresponding reaction do
not depend on its affinity. In fact, according to (15.1.2) the affinity Ar of any
of the rth reactions is defined by the relation between the rates of direct and
reverse reactions and does not depend on each of them separately. Hence
chemical reactions are described by exponential kinetic equations with
phenomenological coefficients independent on forces. In this case
generalized reciprocity relationships (5.5.3) should be true having in our
case the form:

∂wα /∂Aβ = ∂wβ /∂Aα . (15.1.14)

To make sure they are valid, it is enough to differentiate expression

(15.1.10) with respect to Aβ, given L11 and L12 independent on Aβ, and to
repeat the similar operation on (15.1.11) with respect to wβ. The result will
be as follows (V. Etkin, 1982):

∂wα/∂Aβ = ∂wβ/∂Aα = (ω3΄/RμT) exp(–A3 /RμT ). (15.1.15)

It can be easily seen that with approaching equilibrium, when A1, A2 and
A3 are simultaneously tending to zero, relationship (15.1.15) goes over into
known Onsager’s symmetry conditions for chemical reactions in the form:

L12 = L21 = ω3΄/RμT . (15.1.16)

This example shows that the generalized reciprocity relationships do not

follow from the Onsager’s symmetry conditions as usually considered, but,
on the contrary, these conditions themselves ensue from the differential
282
reciprocity relationships near equilibrium when the linear approximation
appears to be valid.
Thus thermokinetics appears to be quite compatible with the
exponential laws of Guldberg & Waage chemical kinetics, which opens up
additional vistas in the thermodynamic analysis of chemical technology
processes and the evolutional problems of biological and ecological
systems.

15.2. Investigation of Irreversible Processes in Systems Standing

far from Equilibrium

Measuring gradients of temperature, pressure, chemical, electrical, etc.

potential in non-equilibrium systems divided into parts (subsystems) with
membranes, capillaries, gates, etc., is practically excluded. In this case the
thermodynamic transport laws have to be written down in the integral form
where the thermodynamic forces are represented in terms of differentials of
the potentials mentioned. Then the phenomenological coefficients Кk which
are functions of thermostatic variables become implicit functions of the
thermodynamic forces Fk since these depend now on the fields of
temperatures, pressures, concentrations, etc. Such dependence makes the
transport laws linear since these fields vary with variation of the forces Fk
themselves. As shown on example of chemical reactions, the Onsager’s
reciprocity relationships in such systems are violated all the more the further
the system stands from equilibrium. That was brought out clearly in the very
thorough experiments the group of American investigators from the Oak-
Ridge National Laboratory (USA) carried out to solve the problem of
creating nuclear reactors with gas heat carrier for spacecrafts (Ewans,
Watson, Truitt, 1963; Mason, Wendt, Bresler, 1972). A system was
considered comprised of two hollows separated with a graphite membrane
of medium permeability. There was substantially pure argon (х1 = 0.9711)
on a one side of the membrane and substantially pure helium (х2 = 0.9917)
on the other side. Both gases were maintained in a state with the same
temperature. As a result, the counter flows J1 of argon and J2 of helium
(mole/s) occurred across the membrane. The pressure differential Δр having
been generated across the membrane, the filtration process superimposed on
the isothermal diffusion with the result of the volumetric flux JV (cm3/s)

283
arisen. By changing the pressure differential in value and sign a stationary
state could be obtained with the volumetric flux ceased 1).
As independent flows, the investigators considered the diffusion flux JD
(cm3/s) = J1/n1 – J2/n2 (where n1, n2 – molar concentrations of argon and
helium, respectively, mole/cm3), which does not vary with filtration due to
the equality of the volumes J1/n1 and J2/n2 therein, and the volumetric
(filtration) flux of the gas mix JV (cm3/s) =Vμ (J1+J2), which does not vary
with diffusion due to the equality of the flows J1= – J2 therein (i.e. the
constancy of the mole numbers in the subsystems). Diffusion motive force
was adopted as the mole fraction differential for one of the system
independent components (argon) XD = – Δx1, which remained practically
constant in the experiment, while filtration motive force was adopted as the
total pressure differential XV = – ∇р. Then the Onsager’s phenomenological
laws become:

jD = – LDD ∇x1 – LDP ∇р , (15.2.1)

jV = – LPD∇x1 – LPP ∇р , (15.2.2)

where jD, jV –diffusion and volumetric flux densities, respectively,

cm3/cm2⋅s; LDD, LDP – phenomenological diffusivity (the first index “D” in
both subscripts) resulting from diffusion and filtration (the second index
“P”) motive forces, respectively; similarly LPD and LPP – filterability of the
volumetric flux resulting from the same forces.
However, to apply these laws to a finite-thickness membrane, it was
necessary to change to integral forces expressed in terms of differentials of
the same values Δx1 and Δр. To do that, it was necessary to know how the
local factors LDD, LDP, LPD and LPP depended on the pressure and molar
concentration of one of the gases n1. For this purpose the investigators
applied to the Chapmen-Enskog’s kinetic theory of ideal gas transport.
According to this theory a little bit modified to the intermediate flow under
consideration in-between the normal diffusion and Knudsen fluxes the
binary gas in the porous membrane is approximated as a three-component
system where one of the components is the membrane material itself as if
distributed in the gas flow in the form of a more coarse suspended dust. The

1)
To ascertain those conditions was the main task of the investigation mentioned.

284
said “dusted gas” model certainly comprised the Onsager’s symmetry
conditions and provided explicit equations for diffusivity and filterability:
LDD = (D1/n1γ1 +D2/n2γ2)/р(n1γ1 + n2γ2) , (15.2.3)
LPD = LDP = (D1– D2)/р(n1γ1 + n2γ2) , (15.2.4)
LPP = (n1D1+n2D2)/р(n1γ1 + n2γ2) + Bo/ν . (15.2.5)

Here D1 = (1/D1k+1/D12); D2 = (1/D2k + 1/D12) – diffusion factors for argon

and helium, respectively; γ1 = D1/D1k; γ2 = D2/D2k; D1k, D2k – Knudsen
diffusion factor and inter-diffusion factors for argon and helium,
respectively; Bo – membrane permeability; ν – dynamic viscosity of the
system.
The said dependence of the factors Lij in relationships (15.2.3)-(15.2.5)
on the local pressure p and the molar concentrations n1 and n2 of the
components on the membrane made the diffusion and filtration integral laws
jD = – LDDср Δx1 – LDPср Δр , (15.2.6)
jV = – LPDср Δx1 – LPPср Δр (15.2.7)

non-linear since the fields of pressures and concentrations in the membrane

vary with the forces Δx1 and Δр; the averaged factors LDDav, LDPav, LPDav and
LPPav also appear to be an implicit function of these forces. Therefore the
associated targets of this investigation were to check the applicability of the
linear TIP and the ideal gas transport kinetic theory to this system. The
calculated results of one series of such experiments at a constant mean gas
pressure of p = 1.96 kg/cm2 and a constant gas composition are plotted on
Fig.15.2 as solid lines against the experimental values of the same
parameters dotted (Mason, Wendt, Bresler, 1972).
As follows from the figure, the calculated from kinetic theory
diffusion and volumetric flux densities jD and jV vs. pressure differential
well agree with the experimental data and have a complex non-linear
character complying with that experimentally found. At the same time both
the experiment and the calculation reveal a violation of the Onsager’s
symmetry conditions in the system under consideration, which aggravates
with the system drifting from equilibrium. In fact, the tangent of jD curve
inclination defining the diffusivity LDP changes not only its value, but also
the sign, whereas the vertical segments in-between the curves Δх1 = 0,963
and Δх1 = 0 characterizing the symmetrical filterability LPD change much
285
less in value and does not change the sign at all. Only close to the point Δp
= 0.2 atm, where volumetric flux becomes zero, while the jD and jV curves
may be quite accurately approximated by the linear phenomenological laws,
the Onsager’s reciprocity relationships apply with relative accuracy.
However, as experimenters stress, the space where the linear laws and the
Onsager’s reciprocity relationships are valid, becomes evanescent if
equations (15.2.6)-(15.2.7) are integrated with the fluxes jD and jV related to
the mean integral value of argon concentration (which ensued from the TIP
jD, jv, requirement for invariance of
sm3 dissipation function with change to
sm2s
j new forces and flows). It is also
6 D

5 significant that the tangents of jD

4
and jV curves inclination are
3
opposite on the major part of their
2
1 2 length (at Δp > 0.2 atm), which
1 j v
evidences the anti-symmetrical
j character of the reciprocity
0 v
relationships). This also
–1
–2
contradicts the TIP stating the anti-
–0.4 –0.2 0.0 0.2 0.4 ΔP, atm symmetry conditions show in only
– calculation; – experiment the case when the forces XD and
_______calculation;○ experiment XV have different parity relative to
1 – ΔxAr = 0.963; 2 – ΔxAr = 0. time reversal (H. Cazimir, 1945).
Meantime, in this case these forces
Fig.15.2. Superposition of Flows do not change the sign when time
across Graphite Membrane “reverses”, i.e. both of them refer
to even time function. It is worth
noticing that close to the stationary state (jV = 0) the Onsager’s reciprocity
relationships apply despite the extremely possible drift of the system from
material equilibrium (Δх1 ≈1), which evidences that the TIP requirement for
a system to be immediately close to equilibrium is superfluous. Thus we
encounter here the situation when practically none of the linear TIP known
statements applies.
It becomes even more attractive to check the applicability of
thermokinetics at these conditions. Note immediately that the independent
diffusion and volumetric fluxes jD and jV, as well as the volumetric flux
motive force XV = – ∇р, comply with the thermokinetic requirements.
However, the diffusion flux motive force at S,V = const is the negative
gradient of diffusion potential of any (e.g. the first) independent component

286
XD = –∇ζ1. Since the molar entropies sko and volumes υko remain constant in
this case, then according to (14.3.2)

XD = –∇ζk1 = – ∇μ1, (15.2.8)

so that it may be used a known representation of the ideal gas chemical

potential μi in terms of its standard value μiо and the partial pressure рi

μi = μiо + Rμ T ln рi . (15.2.9)

Applying the ideal gas equation рi = рxi = niRμT (where xi , ni , Rμ – mole

fraction of the ith gas, its molar concentration and universal gas constant,
respectively) the diffusion motive force may be expressed as XD = – (Rμ
T/рi)∇рi. Thus the forces XD and XV appear in this case to be interrelated so
that the local diffusion and filtration laws should be written down in the
matrix (Onsager’s) form:

XD = RDD jD + RDР jV (15.2.10)

XV = RРD jD + RРР jV (15.2.11)

где RDD , RDР , RРD , RPP – phenomenological resistibility as inverse value to

the phenomenological diffusivity and filterability LDD, LDP and LPD, LPP,
respectively (S De Groot, P. Mazur, 1964).
To change to the integral form of these equations where the diffusion
and filtration motive forces are expressed in terms of pressure differentials
X̄D = – Δр1 ; X̄V = – Δр), let us apply to the same phenomenological factors
vs. pressure and concentration as in 15.2.3-15.2.5. Taking into consideration
that X̄D = ∫XDdℓ; X̄V = ∫XVdℓ, where ℓ – membrane thickness, and given
constancy of the jD and jV fluxes, gives that change to the integral form of
laws (15.2.3) and (15.2.4) comes to calculation of the total resistibility to the
said flows Řij = ∫Rijdℓ = ∫Lij1dℓ. This leads to pseudo-linear integral laws of
the following kind:

X̄D = ŘDD jD + ŘDP jV (15.2.12)

X̄V = ŘPD jD + ŘPP jV (15.2.13)

287
 These laws differ from (15.2.6), (15.2.7) basically in the behavior of the
diffusion motive force vs. the pressure differential across the membrane
with the result that the non-linearity of the jD curves becomes less distinct. It
only remains to confirm the validity of generalized reciprocity relationships
(4.5.3) for such processes. The computation results based on (15.2.3)-
(15.2.5) for the same experimental conditions are described in Table 12-1
(V. Etkin, 1983).
As follows from the table, the generalized reciprocity relationships
(columns 3 and 4) apply well enough for even the close-to-limit drift of the
system from diffusion equilibrium. At the same time the computation results
show the Onsager’s reciprocity relationships go over into the Cazimir’s anti-
symmetry conditions thru the entire range of thermodynamic forces. This is
caused by the fact that the concentration gradients across the membrane for
argon and helium are counter-directed (i.e. are opposite in sign).

Table 12.1
Reciprocal Relations for Non-Linear Cross Diffusion
of Helium and Argon
Motive forces, atm Diff. reciprocal reltions, Onsager’s symmetry
cm3/cm2 ⋅s⋅ atm conditions,
cm3/cm2 ⋅s⋅ atm
X̄D X̄V ∂jD /∂X̄V ∂jV/∂X̄D LDP LPD
0 0 – 1.75 1.75 – 1.749 1.749
0 0.5 – 1.77 1.77 – 1. 765 1.915
0 0.9 – 1.83 1.82 – 1.837 2.12
0.5 0 – 1.91 1.905 – 1.875 1.85
0.5 0.5 – 1.70 1.76 – 1.786 1.97
0.5 0.9 – 1.21 1.25 – 1.81 2.15
– 0.5 0 – 1.75 1.77 – 1.835 1.745
– 0.5 0.5 – 2.87 2.96 – 1.85 2.08
– 0.5 0.9 – 4.51 4.55 – 1.89 2.21

In this case the equalization of concentration of one of the system

components (argon) causes the “ascending diffusion” of the other
component (helium), i.e. its motion toward its increased concentration. The
two flows (argon and helium) become now interconnected (“related”) so
288
that an enhancement of one of them suppresses the other. This means that in
the course of relaxation of one of the components useful work on the other
one is being done. Since TIP does not consider the useful work processes, it
could not predict this result. In this respect thermokinetics supplements the
existing concepts regarding the origin of anti-symmetrical reciprocity
relationships including in their number the cases of counteracting forces of
the same parity. It follows also from Table 15.1 that the anti-symmetry
conditions apply quite satisfactorily close to the stationary state associated
with disappearance of the volumetric flux jV, though the system in whole
stands in this case extremely far from material equilibrium (concentration
differential close to unity). This confirms the deduction of thermokinetics
that, to agree with the Onsager-Cazimir’s symmetry conditions, it is enough
that the cross terms in the Onsager’s phenomenological laws be linear,
which complies with partial (in this case, filtration) equilibrium. Thus we
encounter in this example a situation when practically all statements of
linear TIP do not come true.

15.3. Simplification of Transport Laws Based on Differential

Reciprocal Relations

A significant advantage of the method for investigation of irreversible

processes consists in the possibility to reduce the transport kinetic equations
to the “diagonal” form containing a single (resultant) motive force. This
allows dramatic reduction of the kinetic (empirical) factors these equations
contain. This may be instantiated most clearly by the interrelation between
anisotropic thermal conduction and electrical conduction of bodies in
magnetic field.
Artificial anisotropy, viz. a distortion of current lines and heat stream in
conductors placed in magnetic field, engenders a number of effects usually
referred to the “thermo-galvano-magnetic” group and subdivided into
“thermoelectric” (due to the interrelation between thermal and electric
phenomena), “galvanomagnetic” (due to the interrelation between electric
current and magnetic field) and “thermomagnetic” (due to the impact of
magnetic field on heat flow), while depending on the magnetic field
direction (lengthwise or across the conductor) – also into longitudinal and
transversal. An empirical approach to the description of such phenomena is
usually restricted to the 2D case (heat and charge streams lie in the x–y

289
plane) and introduces the tensors of thermal conduction and electrical
conduction thus providing a set of 6 interrelated equations (equal to the total
number of the electric current and heat stream vectors’ components) with 36
empirical factors (S. De Groot, 1956). Due to the isotropy of the system in
the x–y plane only 12 factors of 36 remain independent.
These are the laws that served for L. Onsager as a prototype to his
“matrix” phenomenological laws (4.4.1). According to these laws the
electric charge and heat stream vectors’ components are considered as
independent flows, while the temperature gradient and electric potential
vectors’ components – as independent thermodynamic forces. The linear
theory of irreversible processes (TIP) based on this assumption sets an
additional interrelation between some of the abovementioned phenomena
and reduces the number of the independent empirical factors to 9 due to the
application of the Onsager-Cazimir’s symmetry conditions (R. Haase,
1967).
Thermokinetics allows to go even further and to reduce the number of
the said factors more than twice by imparting the thermodynamic form to
the equations of anisotropic thermal conduction and electrical conduction.
This becomes possible by applying the reciprocity relationships not before,
but after the final (thermokinetic) form of anisotropic thermal and electrical
conductivities has been determined.
According to thermokinetics the artificial anisotropy of electrical
conduction may be allowed for by proceeding from the traditional definition
of work as the product of a force and the distance the object covers under
this force applied. According to (2.4.7), providing the force direction does
not coincide with the travel direction as it takes place in anisotropic bodies,
the dissipation function (dissipation process power) should be represented
as:

Nd = Σi ⏐Xi ⏐⋅⏐ji ⏐cos γi , (15.3.1)

where γ – angle between the vectors Xi and ji.

From this it follows that the thermodynamic form of Ohm’s and Fourier
laws for bodies with artificial anisotropy requires the angle γe between the
anisotropic electrical conduction motive force Xea and the current vector je
to be allowed for, as well as the angle γq between the anisotropic thermal
conduction motive force Xqa and the entropy flow js:
290
 je = σe Xea cos γe = Le (Xea) Xea , (15.3.2)

js = (λ/T)Xqa cos γq = Lq (Xqa)Xqa , (15.3.3)

where λ, σe – thermal conductivity and electric conductivity, respectively,
in the absence of magnetic field; Lq(Xqa) = λcosγq /T; Le(Xea ) = σecosγe –
phenomenological coefficients as functions of the thermodynamic forces Xsa
and Xea thru the angles γq и γe.
Before obtaining an extended expression for these laws, let us find their
thermodynamic form in the absence of anisotropy. We will consider
electrical conduction in metals as a particular case of the diffusion in a
multi-component system where free electrons are the only movable
component (Groot, Mazur, 1964). The chemical potential μk for a multi-
component system with components carrying the electrical charge per mass
unit ϑеk is known to give place to the electrochemical potential μk∗ = μk +
ϑеkϕ, where ϕ – electrical potential of a particular zone in the system.
Therefore assuming the “electron gas” to be, as usual, one of the
components of the system “free electrons + ions” and the uniqueness
conditions for the diffusion of free electrons to be the same as in the
diffusion of the kth component, gives the additional term ϑеk∇ϕ appearing in
the equation for the thermodynamic force Xk = –∇ζk of this process:

Xk = – [Σkμkℓ∇cℓ + (sk – sko)∇T – (υk – υkо)∇p +ϑеk∇ϕ]. (15.3.4)

Since mechanical equilibrium (∇p = 0) takes place in the system under

consideration and the gradient of ion concentration (second component) is
absent, i.e. ∇cℓ = 0, the expression for the diffusion flux of the kth
component in electrolytes becomes:

jk = – Lkk [ϑеk∇ϕ + (sk – sko)∇T]. (15.3.5)

To endue this equation with the form of Ohm’s generalized law, it is

necessary to change to the electrical charge flux with the kth component jе =
ϑеkjk and to simultaneously introduce the related force Xe proceeding from
the product jk⋅Xk to be maintained (dissipation function invariance). This
condition is associated with the force Xe = Xk/ϑеk so that Ohm’s generalized
law becomes:

291
 jе = σеXe = σе [E + Xeq] , (15.3.6)

where E = –∇ϕ – electric field; Xeq = – sе∗∇T – thermodynamic

representation of the so-called “thermomotive force”; sе∗ = (sе – sеo) –
specific (per charge unit) electron transport entropy (Haase, 1967); σе =
Lkkϑеk 2.
Thus there is an additional force of Xeq appearing along with EMF in
conductors with inhomogeneous temperature field and responsible for
homogeneous thermoelectric effects (Etkin, 1991). In magnetic field it is
supplemented with the Lorenz force FL = ve × B = υe je × B (where ve , B –
electron transport velocity and magnetic flux density, respectively; υe, je –
specific volume of “electron gas” and current density, respectively). This
force is a normal to the current direction and, therefore, does not do any
(including dissipation) work. However, this is the force responsible for the
artificial anisotropy of the initially isotropic heat and charge conductors
since it changes the direction of current and heat stream and thus – the
projection of the anisotropic electrical conduction resultant force Xea = E +
Xeq + FL on the electron transport direction. Given this force, Ohm’s law
acquires the following extended thermodynamic form:

je = – σecosγe [∇ϕ + sе∗∇T + υeje × B] , (15.3.7)

where

cos γe =⏐Xe⏐/⏐Xea⏐= 1 –⏐FL⏐/⏐Xea⏐. (15.3.8)

The dependence of cos γe on the current density je makes the anisotropic

electrical conduction law (see 15.3.7), strictly speaking, non-linear.
To determine the thermodynamic form of Fourier law for the case of
anisotropic thermal conduction, let us use the generalized reciprocity
relationships (see 4.5.3) which have the following form for the case under
consideration:

(∂je /∂Xqa) = (∂js /∂Xea) . (15.3.9)

Given (15.3.8) and the independence of ∇ϕ on Xqa , one can find at Xea
being constant:

(∂je /∂Xqa) = Xea(∂Le /∂Xqa) = σe Xea( ∂ cos γe /∂Xqa) = – σe sе∗. (15.3.10)

292
 Then according to (15.3.9) with the components FL and Xeq of the force
Xea being constant:
(∂js /∂Xea) = (∂js /∂E) = – σe sе∗ . (15.3.11)

This means that an additional term, viz. – σesе∗E, appears in the

thermodynamic form of Fourier law generalized for the anisotropic thermal
conduction case:

js = – (λ/T)∇T – σe sе∗E . (15.3.12)

The found diagonal form for the equations of anisotropic thermal and
electrical conductions allows obtaining all superposition effects inherent for
such systems from the equilibrium conditions, i.e. by the same method as
described in the previous chapter.

15.4. Further Reduction of the Number of Kinetic Factors

in Transport Equations

Let us apply the found equations of anisotropic thermal and electrical

conductions to determine the analytical expressions for the abovementioned
superposition effects. Let us start off with the magneto-resistive effects
consisting in resistance variation of the conductors in magnetic field and
used for the magnetic field measurement (in magnetometers).
From the positions of thermokinetics these effects are explained by the
deviation of the current direction from the electric field direction E. If, for
the sake of demonstrativeness (see Fig.15.3), the field direction E coincides
with the axis x, then at ∇T = 0 the current direction will make the angle γe
with this axis, while the current component along this axis, jeх(B), in the
transverse magnetic field will vary by a factor of cosγe against its value
without the field, which, according to the definition method of the electric
conductivity σe = je/E, will lead to reducing this conductivity in the
transverse magnetic field down to σe":

σe̋ = jeх /∇х ϕ = σe cos γe , (15.4.1)

293
where ∇х ϕ – projection of the vector ∇ϕ on the axis x.
As next step, let us consider thermoelectric effects in the presence of
magnetic field. Let the conductor plane and the current therein (see
Fig.15.3) are oriented in the magnetic field direction, i.e. along the axis z (je
= jez). Then the angle between the vectors je and B is equal to zero, so
Ohm’s law becomes:

jez = – σe cos γe [∇z ϕ + se ∗∇zT], (15.4.2)

where ∇z ϕ , ∇z T – projections of the vectors ∇ϕ and ∇T on the current

direction.
y а y б
a a
Xe Xq
Xey Xqy

jey je E Fe j qy jq Xqe
γe γe ∇T
x x
B j ex Xex B j qx Xqx

z z

Fig.15.3. Thermal Conductance & Electrical Conductance Anisotropy

As follows from this expression, the electric current is a result of the

joint action of the thermoelectric force se∗∇zT and the electric field ∇zϕ.
And otherwise, the current flow in the conductor generates a heterogeneous
temperature field with the conductor surface releasing (or absorbing) the
Thomson heat qт* = Tse* which is similar to the transport heat qk* of the kth
component. That phenomenon was first predicted by Thomson in his theory
of thermoelectricity (1864) and has been called Thomson effect since that
time. From the expression for the Thomson specific heat qт* it follows that
the Thomson effect value is proportional to the current density je, while for
the conductor as a whole – to the displacement current therein, which has
been experimentally confirmed (R. Haase, 1967). It is worth mentioning the
stationary state supported by the thermal conductance process and featuring
the current disappearing (jez = 0) due to the thermodynamic force Xe
components mutually balanced out in (15.3.7). This phenomenon is called
the Thomson stationary effect and estimated as the electric potential
gradient related to the temperature gradient in a conductor de-energized (jez
= 0):
294
 αТ΄ = (∇z ϕ/∇zT)st = se*, (15.4.3)

where αТ' – thermoelectric motive force coefficient 1).

The similar effect appears in the transverse magnetic field as well when
the current is directed along the axis x. In this case in Ohm’s law (15.3.7) je
= jex;∇ϕ =∇xϕ;∇T = ∇xT. For stationary conditions (je x= 0) from (15.3.7) the
expression for the Thomson cross effect follows as:

αТ˝ = – (∇x ϕ/∇xT)st = se*. (15.4.4)

Now let us consider the thermomagnetic effects. One of them – Righi-

Leduc effect – means a temperature gradient arising in the direction normal
to the heat flow jq and the field B in the absence of current (je = 0). This
effect is as well caused by the abovementioned thermal conductance
anisotropy in magnetic field. If, for the sake of convenience, the heat stream
direction coincides with the axis x and given the vector ∇T makes the angle
γq with this axis (see Fig. 15.4), the components ∇xT and ∇yT will be
evidently interrelated as:

αRL B = (∇yT/∇xT) = tg γq (je = 0), (15.4.5)

where αRL – Righi-Leduc coefficient.

a a
а 0 je jеx Xe б jq –∇x T Xqe
x 0
x
B γe B γq
jey E Fe Xqe
–∇yT E
z
Xe z
∇T
–y
–y

Fig.15.4. Definition of Thermo-Galvano-Magnetic Effects

1)
The left-hand side of the expression for an adequate effect will hereafter everywhere
contain the coefficient corresponding to the empirical description of this particular
phenomenon in the monograph by R. Haase, 1967.

295
 That effect was discovered almost simultaneously by A. Righi and S.
Leduc (1887). It is usually represented as the product of some constant αRL
and the magnetic intensity B thus stressing the proportionality of this effect
to the field B. Note – this fact directly ensues from (12.4.4) thereby
confirming the deductions of thermokinetics.
Another thermomagnetic effect consists in the emergence of a weak
electrical field of ∇yϕ in a de-energized conductor containing a temperature
gradient of ∇T caused by the heat flow jq in the direction normal to the field
intensity B. That effect was discovered by W. Nernst and A. Ettinshausen
(1886) and has borne their names since that time. It can be more
conveniently described with the vector ∇T oriented along the axis x (see
Fig.15.4). Then ∇T = ∇xT and, given the evident relation ∇yϕ/∇xϕ = tgγe
(Fig.15.4), the desired relationship can be represented as:

ΑENB = – (∇yϕ/∇xT) = – (∇xϕ/∇xT)(∇yϕ/∇xϕ) = se*tg γe , (15.4.6)

where αEN – Ettinshausen–Nernst constant. The proportionality of this effect

to the field B is also experimentally confirmed.
Let us now consider the group of galvanomagnetic effects in more
detail. Hall effect discovered in 1879 is the best known of them. It consists
in the emergence of the electrical field E in the direction normal to the
electrical current je and the magnetic field B. To derive the Hall effect
analytically, let us align the axis x with the current density je vector (Fig.
15.4). Then jey = 0; je = jex , while the force Xe will make an angle of γe with
the axis x so that its component along the axis y will become equal to ∇y ϕ =
∇х ϕ tgγe. Taking into account that in a conductor thermally homogeneous je
= – σe"∇хϕ gives:

αHB = – (∇y ϕ /jex )ст = tg γe /σe˝ ( jey = 0 ), (15.4.7)

where αH – Hall constant.

From (15.4.7) it directly ensues that Hall constant is nothing else but the
specific volume per electrical charge unit (positive or negative). Therefore
the Hall effect sign depends on the charge carrier sign, while the Hall effect
value – on the charge carrier concentration. These facts make Hall effect
one of the most efficient methods for investigating the charge energy
spectrum in metals and semiconductors. It is also used for measurement of

296
magnetic intensity, for current multiplication in analog computers and in
some kinds of MHD-current generators.
Another effect also discovered by Nernst in 1887 and bearing his name
consists in the emergence of a temperature gradient of ∇xT in the direction
of the current je = jex with the transverse magnetic field available when the
heat flow jq is absent. This effect differs from the Thomson thermoelectric
effect because arises in a conductor that initially is thermally homogeneous.
This effect results from setting the stationary state when the heat stream
ceases. Then substituting the product σeЕ for jex in Fourier’s generalized law
(15.4.5) gives that at this state jsx = – (λ˝/T)∇xT + se*jex = 0. From this it
directly follows that

αN = (∇xT/jex)cт = Tse∗/λ˝ ( jey = 0 ), (15.4.8)

где αN – Nernst constant; λ" – thermal conductivity in the transverse

magnetic field.
The effect of a temperature gradient of ∇zT emerging in the magnetic
field direction is similar to the Nernst effect providing the current flows in
the same direction je = jez, while the system in this condition is thermally
homogeneous. This phenomenon does not have a special name – R. Haase
proposed to call it Kelvin effect. Expressing, similar to the above, Fourier’s
law for the heat flow along the axis z gives jsz = – ( λ/T )∇zT + se*jez, so that
with the stationary state setting in:

αK = (∇zT/jez )cт = Tse*/λ ( jex = 0) , (15.4.9)

where αK is a value R. Haase proposed to call Kelvin constant.

Ultimately, there is one more effect among the galvanomagnetic effects,
which was also discovered by Ettinshausen in the mentioned 1887 and
consists in the emergence of a temperature gradient of ∇yT in the direction
normal to the current je = jex and the field B. It is also caused by the
different directions of the vectors Xe and ∇T. If, by way of explanation, the
axis x coincides with the current je direction, then the components ∇y T and
∇xT will be interrelated thru tgγq. Then substituting the expression ∇yT tgγq
into (15.4.9) gives instead of ∇xT:

αE B = (∇yT /jex)st = Tse*tg γq/λ˝ ( jey = 0), (15.4.10)

297
где αE – Ettinshausen constant.
Thus the entire spectrum of the thermoresistive, thermoelectric,
thermomagnetic and galvanomagnetic effects in the longitudinal and
transverse magnetic fields may be expressed using only 4 phenomenological
coefficients (λ, λ", σe,σe″). This substantially facilitates their theoretical
analysis and experimental investigation.

15.5. Setting Additional Constraints between Superposition Effects

The above-found thermo-galvano-magnetic effects allow setting a

number of relationships between them earlier obtained in TIP based on the
Onsager-Cazimir’s reciprocity relationships. This is first of all the
Bridgeman’s relationship (1929) connecting the Ettinshausen-Nernst effect
with the Ettinshausen effect:

TαEN = λ"αE . (15.5.1)

This relationship can be also obtained from thermokinetics by

comparing expressions (15.4.6) and (15.4.10). Another relationship
connects the Nernst effect with the Thomson effect in the transverse
magnetic field:

αN = T αT"/λ". (15.5.2)

This relationship generalizes the thermoelectric effects to the systems in

the transverse magnetic fields. It can be also obtained by directly comparing
expressions (15.4.9) and (15.4.8). A similar relationship takes place in the
longitudinal magnetic field as well:

αК = TαT'/λ. (15.5.3)

This expression extends the Thomson relationships obtained for

thermocouples to particular current-carrying conductors. It can be also
obtained by comparing expressions (15.4.8) and (15.4.14).
At the same time thermokinetics allows obtaining 5 additional
relationships between the thermo-galvano-magnetic effects. One of such
relationships directly ensues from comparing the thermo-EMF coefficients
298
in the longitudinal and transverse fields (see 15.4.8 and 15.4.9), which
evidences the identity of the Thomson longitudinal and transverse effects:

αТ' = αТ" = se* (15.5.4)

This provision, as well as the representation of the Thomson effect in

terms of the transport entropy, did not ensue from TIP. It means that the
thermo-EMF value does not depend on magnetic field – only the thermo-
EMF direction varies and to the same extent as the direction of the heat
stream creating this EMF. From this a physically significant conclusion
follows about the coincidence of the angles γq and γe, i.e. about a unified
mechanism of the magnetic field impact on both the heat and electrical
charge flows, which allows definition of the angle γ = γe = γq by measuring
the magnetoresistive effect (see 15.4.6):
Another relationship, due to γ = γe = γq, connects the Ettinshausen effect
(15.4.10) with the Nernst effect (15.4.8):

αN tgγ = αEB . (15.5.5)

One more relationship connecting the Righi-Leduc and Hall effects can
be obtained by comparing expressions (15.4.5) and (15.4.7) given the
equality γe = γq :
αRL = σe"αH . (15.5.6)

That relationship was earlier set within the electron theory of metals and
many times experimentally confirmed.
The relationship between the Righi-Leduc (15.4.5) Ettinshausen-Nernst
(15.4.6) effects can be found in the same way:

αEN = se*αRL . (15.5.7)

Considering relationships (15.5.6) (15.5.7) jointly gives Wiedemann-

Franz law stating constancy of the metal thermal conductance/electric
conductance ratio:

Λ = λ"/Tσe ″ = αH αEN /αRL αE , (15.5.8)

299
where Λ is the so-called Lorenz constant. Thus Wiedemann-Franz law, as
well as relationship (15.5.6), take a character of the deductions of
thermokinetics. This shows, in particular, in the fact that all the above
thermomechanical, thermochemical, thermoelectric and thermo-galvano-
magnetic effects appear to be expressed in terms of the exclusively
thermodynamic variables. This fact, as well as the possibility to set
additional relationships between the superposition effects, results from the
interrelation between the diagonal and non-diagonal phenomenological
coefficients in the Onsager’s phenomenological laws having been taken into
account. To reveal the nature of such a relation, let us apply again to the
thermokinetic form of transport equations (2.6.11) and (14.3.1) deriving
them for a particular case of transporting any kth substance in the form:

Jk = Кk Σi Fki = LkXk , (15.5.9)

where Fkj = αkj∇ψj – components of the resultant force Xk = – ∇ψk ;

αkj ≡ μkℓ, sk∗, θek, υk, etc. – thermodynamic values associated with the
generalized potential gradients ∇ψj ≡ ∇сℓ ,∇T, ∇ϕ, ∇p, etc.
To return to the matrix form of the phenomenological laws adopted in
TIP, let us introduce new phenomenological coefficients Lkj = Lkαkj. Тогда

jk = Σ j L kj Xj . (15.5.10)

It can be easily noticed that in such (onsager’s) form of

phenomenological laws the non-diagonal coefficients Lkj (k ≠j) are
connected with the diagonal coefficients Lkk = Lkαkk thru a simple
relationship:

Lkk /αkk = Lkj /αkj = Lk . (15.5.11)

The relationship of such a kind may be instantiated by the above-found

relation between the diagonal and non-diagonal diffusivities. Thus the
definition of the thermodynamic form of phenomenological laws (14.3.1) is
equivalent to imposing (n – 1) additional constraints between the diagonal
Lkk and the non-diagonal Lkj phenomenological coefficients by representing
their ratio in terms of relations of the corresponding thermodynamic
parameters:

300
 Lkj /Lkk = Xkj /Xj = αkk /αkj . (15.5.12)

This becomes possible because the kinetic factors conditioning these

coefficients (permeability of membranes, mobility of components, etc.)
equally enter in both Lkk and Lkj with no impact on their ratio. The number
of such constraints in each of the equations of the (12.5.11) type is evidently
equal to (n–1), which gives, with n of such equations, n(n–1) constraints of
the (12.5.12) type and allows reducing the number of the phenomenological
coefficients from n2 in empirical description down to n, whereas TIP
reduces the number of such coefficients to only n(n+1)/2. In particular, to
describe the diffusion, thermal diffusion and pressure diffusion of the kth
substance, one kinetic factor (along with the thermodynamic factors)
appears to be sufficient, while, to describe anisotropic thermal conductance
and electric conductance in the longitudinal and transverse magnetic field,
per one thermal conductivity and electric conductivity (λ , λ", σe, σe") are
sufficient. In this respect the thermokinetic approach may be considered as a
final stage of the transition from a purely phenomenological (pre-
thermodynamic) investigation of a phenomenon to a thermodynamic-
phenomenological investigation and further – to a purely thermodynamic
one.

Chapter 16

SPECIFICITY OF TRANSFER PROCESSES IN BIO-SYSTEMS

One of the most characteristic features of bio-systems consists in the

structuring processes running therein with a higher level of their
heterogeneity (non-uniformity) and enhanced interrelations between them.
In the late 19th century H. Driesch, a prominent German embryologist, based
on a comprehensive experimental data proved that not only and not so much
embryo cells as themselves would define bio-structure formation processes,
but rather their position in the germinating embryo considered as a single
whole, whence it followed that the properties of a whole bio-system would
not mean the sum of its constituents’ properties, but should be considered as
a single spatially heterogeneous system. This concept underlies the
principles of theoretical biology E. Bauer has formulated. His first principle

301
reads as follows, “All living systems are never being in equilibrium but
permanently doing work at the expense of their free energy against
equilibrium required by the laws of physics and chemistry under existing
environmental conditions”.

16.1. Vector Character of Biochemical Reactions

The useful (ordered) work can be provided in according to expression

(2.3.7) by solely the forces of vector character Fi or Хi enabling the
transport of the object they are applied to at the distance dRi. The validity of
this statement is confirmed by all available operation experience of various
power units. The chemical sources of electric energy, where for this purpose
the spatial separation of reagents and, respectively, electrodes is provided,
are not either exclusion. In the absence of that stated chemical reactions
inevitably acquire the thermodynamically irreversible character, while their
chemical affinity is realized in the form of the released dissipation heat
equivalent to work. The affinity of homogeneous chemical reactions relates
therefore in TIP entirely to the dissipation heat wells (see 4.1.3).
Nevertheless, the EMF of galvanic or fuel cells found by the chemical
affinity value corresponded to its experimental values. That had not resulted
in any conflicts until a contradiction with Curie law was revealed. This law,
as it has already been noted in Chapter 4, ensues from the condition of
conservation of constraints at the transformation of coordinates and
excludes the possibility of association (interrelation) of the phenomena
having different tensor order and kind. In the context of transfer phenomena
this means that the generalized rate of whatever process (flow Ji) in the
equations of (4.4.1) type may depend on only the thermodynamic forces Xj
of the same (or even) tensor order (S. De Groot, P. Mazur, 1967). In the
case under consideration this means that the homogeneous chemical
reactions can not interact with the transfer processes bearing vector
character. Due to the absence in classic thermodynamics the notion of
internal work this fact has for a long time been passed unheeded and has up
to date impeded the comprehension of impossibility to measure the useful
work We in terms of the Gibbs’ energy decrease.
This conflict may be settled with the help of thermokinetics stating the
difference in tensor order between the motive forces and the coordinates of
ordered and disordered works, as well as the vector character of transfer

302
processes and the chemical energy conversion for substances in biological
membranes. To show this, let us consider an arbitrary rht chemical reaction
with the kth substances running in an open system of a type of flow reactor,
fuel cell, plasma membrane, etc. In this case the flow Jk of the kth substance
is related with the total flow of reacting mixture Jс thru the evident
relationship Jk = Jссk, where сk = Nk/N – current (local) value of molar
concentration of the kth substance in a particular cross section of the flow
reactor. Expressing the сk value by a known way in terms of its value сko at
the flow reactor inlet, the degree of completeness ξr of the rth reaction and
the stoichiometric coefficients νkr of the kth substances in this reaction (R.
Haase, 1967; I. Gyarmati, 1974) gives:

сk = сkо + Σr νkr ξr . (16.1.1)

According to this expression the variation in the concentration сk of the

kth substance in an open system occurs both due to the diffusion of the
substances not participating in the chemical reactions (i.e. by equilibrium
mass transfer) across the system borders and due to the internal chemical
conversions. Now let the motive force Хc enabling the transfer of reacting
mixture composed of the kth substances be expressed in terms of the
negative gradient of mixture chemical potential Хc = – ∇μc = –Σk∇(μkсk).
Substituting this force into expression (16.1.1) reveals it as the algebraic
sum of motive forces enabling the processes of two kinds:

Хс = –Σk∇(μkсkо) – Σr ∇ψr = ΣkХk + ΣrХr . (16.1.2)

Here ψr = Ar(1–ξr) – local potential of the rth flow reaction; Хk = – ∇(μkсkо) –

motive force enabling the transfer of the kth substances not participating in
the chemical reaction (diffusion); Хr = –∇ψr – motive force of the rth flow
reaction. Both forces have certainly the same tensor order, which evidences
the vector character of such reactions.
From expression (16.1.2) it also follows that each summand of its first
sum, including the term ΣrArdξr describing the rth scalar chemical reactions
in homogeneous media, receives the matching additional summand of the
second sum Σr Хr·Jr having vector character and describing in this case the
chemical reactions running in the flow (in galvanic and fuel cells, plasma
membranes, Van’t Hoff’s box, etc.). The meaning of the motive forces Хr of
303
such reactions and their generalized rates (flows of reagents Jr) is
unambiguously defined by the general expression Xi ≡ – ∇ψi и Ji = Θivi. In
this case the motive force of the rth chemical reaction means, as also in other
cases, the negative gradient of the local potential ψr = Ar(1 – ξr) pertaining to
this reaction in a particular section of the flow reactor or plasma membrane,
while the generalized rate of the process means the flow Jr of the reagents
participating in this reaction:

Jr = – Lr Σr∇ψr, (16.1.3)

Thus under the conditions of spatial separation of reagents chemical

reactions take the directional (vector) character, which causes their
interaction with transfer processes in the full compliance with Curie law.
The validity of this law for chemical reactions can be easily confirmed with
the reagents preliminarily mixed before fed to the electrodes of a fuel or
galvanic cell: their EMF will then drop to zero, while the reaction becomes
thermodynamically irreversible.

16.2. Interrelation between Active and Passive Transport Processes

in Biosystems

One of the major targets of modern biophysics consists in stating laws

describing principles and processes of functioning living organisms.
Thermokinetics is most applicable to these targets. Firstly, it considers each
cell of a bio-system as a spatially heterogeneous system being the same full-
value micro-organism as the whole organs and organisms these cells form.
Secondly, unlike non-equilibrium thermodynamics, thermokinetics is not
limited to studying dissipation processes, but considers, along with them,
also the processes of useful energy conversion inseparably associated with
the vital activity of bio-organisms (E. Bauer, 1930). Thirdly, it is not limited
to the stationary non-equilibrium states, as TIP is, and allows investigating
the cyclic state variations characteristic for bio-organisms at all stages of
their evolution. Fourthly, it is applicable to also the states which are, like
bio-systems, rather far from equilibrium. Therefore the methods of
thermokinetics are most adequate to the circle of the problems biophysics
and bioenergetics solve.

304
 One of such problem consists in determining the motive forces and
generalized rates of biological processes. Specific role among them play the
reactions of adenosine triphosphate (ATP) synthesis from the molecules of
carbohydrates and fats nutrients contain and the reactions of the ATP
subsequent decomposition. The biochemical reactions of ATP synthesis
proceed with useful work done and both the cell and the system in whole
withdrawing from the state of equilibrium. This is achieved due to the active
transport of substances in biological membranes (i.e. due to their transfer in
the direction of increasing concentration). The differentials of chemical
potentials thus generated enable then the passive transport of the reagents to
the reaction zone. The combination of these processes provides the cyclic
process of ATP synthesis and subsequent decomposition with the associated
energy consumption and release.
The determination of such forces and rates is currently provided based
on expression (4.1.4) for the entropy generation rate. Meantime the
application of TIP to analysis of reversible processes is a method involving
certainly unfit means. The useful work bio-systems do in the course of their
vital activity relates, as per classic thermodynamics, to adiabatic processes
which in principle can not cause entropy variations. Therefore the entropy
generation can not assure correct value of the motive forces for biological
processes since their reversible component does not impact the entropy. E.g.
the EMF of a reversible galvanic element derived from (4.1.4) will appear to
be close to zero, the same for the entropy generation itself in this particular
case.
Furthermore, expression (4.1.4) can not be basic for even the definition
of the sign pertaining to these forces. In fact, in the active transport
processes the kth substances are transferred toward their increased
concentration, i.e. the Jk flows are directed against the motive forces Xk. In
this case the product Xk⋅Jk is always negative. Meantime, all the summands
of (4.1.4) are always non-negative as characterizing the share of any of the
independent irreversible processes in the entropy production. It is easy to
ascertain this if noticing that the sign of the derivative Xi =(∂S/∂αi) always
coincides with the sign of the force since the entropy production is always
non-negative. In these cases the authors of many study guides interpret the
situation Xk⋅Jk < 0 as the “nega-entropy production”, “negentropy
consumption”, etc., which is an apparent absurdity from the positions of
classic thermodynamics.

305
 Thermokinetics figures out an escape from the situation offering a more
general method to find motive forces as derivatives of the system energy. In
this case the terms of the second sum in (2.5.9) may have different sign
depending on whether the system itself does work (Хi·dZi > 0) or work is
done on the system (Хi·dZi < 0). This liberates from whatever violence
against thermodynamics when applying equation (4.1.4) to bio-systems
consuming the free energy (inergy) from the environment.
A significant advantage thermokinetics offers is the possibility to find
from its equations the resultant F = ΣiFi of forces of various nature, which
allows to directly obtain the simplest diagonal form of kinetic laws (2.6.11)
with a minimal number of the transfer coefficients K̄ ij therein.
In particular, with temperature and pressure being constant the laws of
passive transport of the kth substances in the finite-thickness biological
membranes become:
Jк = Lk(Хк)Хк = – Lk Δζk. (16.2.1)

Here Хк = –Δζk – diffusive potential differential across the membrane; Lk –

osmotic diffusion coefficients for the kth substances depending on the fields
of temperature, pressure and concentration of all independent components
of the system. This relationship means that transfer equations (16.2.1) are
non-linear since these fields themselves and the field-conditioned
coefficients Lk vary with the Хк variation. However, with the diagonal form
of laws (16.2.1) this non-linearity does not anymore impede the
determination of superposition effects. Let us show this by an example of
the filtration and osmotic diffusion of the kth substances running
simultaneously in a biological membrane. In this case the bulk (filtration)
flow Jv and the flows Jк of the kth substances are interrelated (Jv = ΣkυkJк).
This is especially evident for a binary system where only one of the flows is
independent (say, the solvent flow Jv=υ1J1). This flow is caused by the
solely motive force Хк = –Δζk. Using the expression for the exact
differential of this potential (see 14.3.5) law (16.2.1) may be written as:

Jк = – Lk[Σkμkℓ Δcℓ + (sko – sk) ΔT + υk Δр] , (16.2.2)

where μkℓ – abridged symbol for the derivative (∂ζk/∂cℓ) =(∂μk/∂cℓ). From
here at T = const, Jк = 0 the stationary effect of osmotic pressure generation
Δπs follows as:

306
 (Δπs/Δc2)ст = – (∂μℓ/∂c2)/υ1, (16.2.3)

where Δc2st – stationary differential of impurity concentrations on both sides

of the membrane. This expression allows using in (16.3.3) the
experimentally found value – Δπs (υ1Δc2/Δc2ст) instead of the value
(∂μℓ/∂c2)Δc2 which is hard to measure. In this case the laws of filtration and
osmotic diffusion in biological membranes will become as proposed earlier
(A. Katchalsky, P. Curran, 1967):

Jν = Lv (Δр – σсΔπs); (16.2.4)

Jк = Lk(Δр – σсΔπs), (16.2.5)

where σс = Δc2/Δc2ст – the so-called selectivity factor.

As can be seen here, due to the diagonal form (16.2.1) of the active
transport laws representation, the flow integration effects in such processes
can be found without Onsager’s symmetry conditions (4.1.7) or without
more general differential reciprocity relationships (4.5.3). However, it is
even more important that these effects can be expressed exclusively in terms
of the thermodynamic variables. This allows proposing a new method to
determine the hard-to-measure thermodynamic parameters based on
measuring the stationary superposition effects. In particular, by
experimentally measuring the value υ1(Δπs/Δc2)st the non-calculable
concentration behavior of chemical potential (∂μ1/∂c2) can be found from
the relationship (16.2.3).

16.3. The Role of External energy in Vital Activity of Biosystems

Despite the noticeable successes in interpreting the laws of bio-system

function from the positions of TIP and synergetics (A. Rubin, 1964; S.R.
Keplen, E. Essig, 1968; L. Blumenfeld, 1977; I. Prigogine, 1977; A.K. Pritz,
1980; D. Nickols, 1985; G. Edsol, H. Gotfrend, 1986; G. Gladyshev, 1988,
and others), finding a solution to the evolution problems on the rigorous
physics-mathematical basis encounters considerable difficulties. The major
part of them is attributed to the fact there are no potential state functions in
open systems, which variation could serve as quite general and rigorous
criteria of the development (ontogenesis) and evolution (phylogenesis)
307
pertaining to bio-systems. The entropy S is inapplicable here since in the
bio-systems exchanging substance with the environment it can vary due to
heat transfer or mass transfer and, on the contrary, remain invariable
providing the withdrawal of the system from equilibrium or its approaching
this is caused by useful work done. As known from classic thermodynamics,
a system can be drawn off equilibrium by only doing useful work on the
system, which relates to adiabatic effects and, in principle, can not change
the entropy of the system. The so-called entropy “production” Р = diS/dt
does not either meet these requirements since this criterion features the
minimum in only a quite particular case of stationary processes in linear
systems and states in the vicinity of equilibrium (I. Prigogine, 1947),
whereas the evolution processes are in principle non-stationary and mostly
obey exponential (non-linear) laws. Ultimately, in all cases of useful work
done on a bio-system the flows Ji appear to be directed against the
thermodynamic forces Xi, while the production Xi·Ji becomes negative.
When these forces and flows are derived from TIP based on the expression
of entropy generation rate (see 4.1.4), the share of the terms Xi·Ji in the
entropy generation takes negative sign too. It is usually interpreted as a
consequence of the “negentropy” inflow to the bio-system, which results in
perverting the physical meaning of bio-system ordering processes and in
construing it as a consequence of “order arising from chaos” rather than
acknowledging the evident fact of ordered energy inflow to bio-systems
from outside. The fact there are no potential state functions in
thermodynamic of open systems restricts the application of the
thermodynamic potentials method to biological, ecological, cosmological,
etc. systems. Thermokinetics enables a solution to this problem by
introducing the concept of the external (ordered) energy E as a function of
the external coordinates Ri for a spatially heterogeneous system. This
characteristic state function generalizes the notion of exergy as the maximal
work a system can do before thermodynamic equilibrium with the
environment has been set (J. Szargut, R. Petela, 1968; A. Andrushchenko,
1975, and others). From unifying the object of investigation with its
environment in a single non-equilibrium whole and representing the ordered
energy as a function of state variables for such an “extended” system the
external energy becomes the most general measure of useful external
working capacity of open and closed, simple and complex, homogeneous
and heterogeneous systems. Let us consider the additional possibilities
ensuing from the notion of the external energy E introduced as the

308
convertible part of energy for a spatially heterogeneous system and the most
general measure of how far it stands from equilibrium. First of all, let us
show that the working capacity function in terms of energy carrying flows
in the form of (11.1.2) enables a shorter way to obtain the same results as
exergy in its traditional interpretation.
Let us assume an available heat Q well with a constant temperature of
T1, and the environment with a temperature of T0 as a heat sink of so large
capacity that the entropy flow does not change its intensive parameters. The
power of the heat engine using such energy well can be found from equation
(11.1.2). The entropy flow Isе = dS/dt thru the heat engine is related with the
heat flow at its inlet Iq = đQ/dt simply as Isе = Iq/Т1. In this case, integrating
(11.1.2) over a temperature interval of Т1 – То and assuming Isе = const
(which is equivalent to assuming reversibility of Carnot heat engine based
on these heat wells) gives:

Nq = – ∫ (Jq/Т1)dT = (1 – То/Т1) đQ/dt, (16.3.1)

which corresponds to the expression for heat well exergy (A.

Andrushchenko, 1975) Wеmax = Q(1 – То/Т1).
Similarly the exergy of working medium can be expressed taking into
account that in this case a substance well with a chemical potential of μ1
exchanges not only the mass M with the environment (the mass flow Jmт =
dM/dt), but also the entropy S so that the energy converter power is derived
as the sum:

NТ = – ∫ Jmеdμ – ∫ JsеdT . (16.3.2)

Integrating this expression over temperature and chemical potential

intervals of Т1 – То and μ1 – μо, respectively (где μ1 = h1 – T1s1; μо = hо –
Tоsо) gives:

NТ = [(h1 – T1s1) – (hо – Tоsо)] dM/dt, (16.3.3)

which corresponds to the expression for working medium exergy

(A.Andrushchenko, 1975) Wеmax = H1 – Hо – Tо (S1 – Sо). Similarly the
chemical exergy of reacting systems can be found.

309
 The identity of the results obtained based on both interpretations of the
working capacity function is caused by the fact that these interpretations
refer to the same value. However, unlike exergy, expressions (16.3.1) and
(16.3.3) derive the value of work maximum possible and with system
parameters inconstant. Thus the concept of external energy is a further
generalization of the exergy concept. This imparts all advantages of the
characteristic state function to external energy. First of all, its value as a
state function does not depend on the path the system comes to equilibrium.
This excuses from the necessity to invent the ways of system reversible
state-to-state transfer, which was characteristic for any working capacity
functions in classic thermodynamics.
Furthermore, the ordered energy (inergy) E of an extended system is
expressed directly in terms of the energy carrying flows and the
thermodynamic forces Хi. Due to this, it can be found under the boundary
conditions set by the flows Ji or forces Xi, i.e. similar to those in the heat-
mass transfer theory. One more advantage of the external energy is that it
may be differentiated by the degrees of freedom of the system. It is
important also that, unlike Gibbs or Helmholtz energy, the decrease of the
ordered energy E defines the useful work value at any process conditions
(not only at T,р = const or T,V = const). This further extends the sphere of
application of the thermodynamic potentials method. Due to all these merits
the ordered energy is one of the most universal thermodynamic potentials.
Ultimately, exergy, being a state function of the system under
investigation itself (regardless of the environment), is irreplaceable under
the necessity to define the “reserve” of the system reversible energy in the
absence of a heat/substance “inflow from outside”. This is quite important
when defining the external energy of an “independent” system not needing
the environment to be used. The external energy in this case allows
estimating the reserve of the ordered energy per each of its components
Еi = Xi·Zi. This facilitates the analysis of system energy form-to form
conversion.
To illustrate the universality of the ordered energy as an equilibrium and
stability criterion, let us use the expression for its exact differential (see
2.3.7). To enable a more convenient comparison of the results with those
already known, it makes sense to use the specific values εi of ordered energy
and to express the specific work wi in terms of thermostatic variables. Since
according to (2.5.8) the forces xi = –∇ψi, while the variation of the specific

310
distribution moments zi in distribution processes dzi = θidri, then for the
general case of non-stationary processes the elementary work becomes:

– dε i = đwi = xi·dzi = – θidrψi , (16.3.4)

where drψi = (dri·∇)ψi – variation of the potential ψi in the direction of the

energy carrier Θi1) transfer.
In this case the equilibrium and stability criteria for a system with any
degree of freedom become, respectively:
δεi = – θidrψi = 0 , δ2εi = – (∂θi/∂rψi) > 0 . (16.3.5)

Let us use this expression to set particular equilibrium and stability

conditions for various systems. For a particular case of mechanical degree
of freedom Θi = – υ, ψi = р. From this it follows that at equilibrium the
pressure in all points of the system is the same (∇р = 0), while (∂υ/∂rр) < 0,
i.e. the gas specific volume decreases with the pressure rise. It is easy to see
that these conditions are equivalent to the mechanical equilibrium and
stability conditions well-known in classic thermodynamics (I.P. Bazarov,
1991). Similarly for the thermal degree of freedom (θi = s, ψi = T) we have
∇T = 0, (∂s/∂rT) > 0, which means the same temperature all over the system
and entropy rise with temperature. The electrostatic equilibrium conditions
are the like ∇φ = 0, (∂θе/∂r φ) > 0 meaning the same electrostatic potential φ
all over the conductor and charge increase with electric potential.
Let us now consider the conditions of material equilibrium, i.e. the
equilibrium distribution of the kth substance among system parts.
Proceeding, as usually, from Gibbs relationship (4.5.1) gives the potential of
the kth component as its chemical potential μk. In this case the material
equilibrium conditions become ∇μk = 0, which corresponds to Gibbs
conditions μk' = μk''. However, given heat and bulk mixing effects, the
component potential at р,Т = const is the diffusion potential ζk (Chapter 9).
In this case the Gibbs material equilibrium conditions give place to the
equality of diffusion potentials ζk' = ζk''.
The phase equilibrium conditions are substantially the same, but
demanding diffusion potential differential comes to zero rather than gradient

1)
Such is, in particular, the expression for the useful work of gas in stream đwре = – υdrp,
where θi = υ (gas volume transferred), ψi = p (gas pressure).
311
(due to the system discontinuity). Thus the material and phase equilibrium is
understood as the equality of diffusion potentials on the interface of
subsystems. It is significant that the heat and bulk mixing effects at phase
transitions are in this case construed as phase transition heat and work (like
the evaporation heat and the expansion work at vaporization).
Thus all equilibrium and stability conditions known in thermodynamics
may be found from the extreme of only one state function – the system
external energy. Another advantage of this criterion consists in its
applicability to open and closed, isolated and non-isolated, homogeneous
and heterogeneous, simple and complex systems regardless of the set of
thermostatic variables accepted as independent.
The method of equilibrium conditions definition based on external
energy is not only straighter and shorter than that based on the max entropy
principle. As follows from equilibrium conditions (4.6.6), the equality of
pressures, as well as chemical, electrical and other potentials at equilibrium
takes place only subject to condition (4.6.7) T' = T". The conclusion usually
drawn from this is the thermal equilibrium conditions are of a more
fundamental importance as compared with others 1). It is commonly stated,
in particular, that material equilibrium can not set in, as a matter of fact,
unless the thermal equilibrium conditions are met since a “heat-
impermeable membrane is simultaneously also a mass-impermeable one”
(R. Haase, 1964). This conclusion is rebutted by the experiments on
annealing of welded diffusion pair (see Chapter 12), which prove that the
flow of one of the movable components can cease even in the absence of
thermal equilibrium. Besides, with internal processes (e.g., diffusion mixing
processes) under consideration, whatever membranes are absolutely
unnecessary. The apparent independence of the Zi and Zj parameters in
equilibrium criterion (2.5.12) unambiguously evidences that the conditions
of partial equilibrium of one (the ith) kind are independent on equilibrium of
other (the jth) kind and thus rebuts the sentential opinion of a specific role of
thermal equilibrium.
One more advantage of external energy consists in the fact that the
stable equilibrium can be distinguished from the metastable (relatively

1)
Note this conclusion ensues from construing entropy not as an independent variable (as by
Clausius), but rather as the state function S = S(U,V,Θi) varying with the system energy in
adiabatic processes (V,Θi = var). Such an extrapolation of the entropy concept, if unheeded,
causes a number of confusions.

312
stable) equilibrium and the labile (absolutely unstable) one. Equilibrium for
the metastable state can be disturbed only at finite variations of the system
parameters ∆θi и ∆zj

δ2εi = – (∂xi/∂zi) > 0 при ∆θi , ∆zj ≠ 0, εi > 0, (16.3.6)

while for labile states – at infinitesimal variations of the system parameters:

δ2εi = – (∂xi/∂zi) > 0 при dθi , dzj ≠ 0, εi > 0. (16.3.7)

For whatever process to arise in a system being in a retarded or

metastable equilibrium, some “energy barrier” has to be overcome. Such a
“barrier”, as a matter of fact, exists for most of the forms of energy. For the
thermal form of energy it shows in a certain temperature difference to be
generated (e.g., a certain overheating or overcooling of one of the phases to
initiate phase transition); for mechanical motion of bodies – in overcoming
the static (rest) friction; for chemical and bio-chemical processes – in
supplying “activation energy”; for spontaneous fission of uranium heavy
nuclei – in introducing fast neutrons to enough amount; for thermonuclear
reactions – in raising heat up to tens million degrees, etc. The value of such
a barrier (its exergy) is allowed for in kinetic equations by introducing a
“threshold” force value of Fiо, below which the process fails to arise.
With introducing the external energy concept it becomes also absolutely
clear that the forms of the energy supplied to a system and accepted by the
system are different things. It becomes clear that not only the form (ordered
or disordered) of energy supply to a system is a matter of importance, but
also how the energy has been accepted by the system. In particular, supply
and abstraction of dissipation heat (disordered energy) can change the
external energy of a system in the same way as the work done on the
system. A body may be, e.g., non-uniformly cooled. Then parts
(subsystems) will appear therein having different temperatures and capable
to serve as heat wells and heat sinks. Such a body will acquire an ability to
do useful external work before thermal equilibrium has set in there. The
form itself of accepting by a system the energy supplied from outside is, to a
great extent, defined by the system properties, i.e. by the forces the system
will set against the external action. If these are forces with a resultant of Fi,
the supplied energy will, at least partially, be accepted by the system as
ordered work done on it. Such is, e.g., the radiant energy partially consumed
313
for photosynthesis of plants, while partially (within the thermal radiance
range) accepted as heat. With the resultant force absent, the system will
accept an external action as disordered work. This work can replenish only
the equilibrium part of the system energy, viz. the anergy. All the said refer
to not only technical, but also biological systems since their vital activity is
based on the external work they do.

16.4. Maximum of inergy as Maturity Criterion for Bio-Organisms

Biological objects relate to the category of open systems developing due

to interchange with the environment of rather substance than energy. These
are the metabolism processes (substance exchange) associated with
chemical reactions running in bio-systems that are a well of the ordered
energy they dispose of.
At the same time the processes of energy conversion in non-equilibrium
bio-systems bear their own specific character. Though in each of the
elements on the surface of the cell membrane these processes, as it has
already been mentioned, bear the vector (directional) character, it is
impossible to define a direction predominant for the whole set of the cells
comprising the organism. This means that chemical reactions for a bio-
system as a whole bear the scalar character. In this case the general measure
of how far this bio-system stands from equilibrium and the main source of
its vital activity is its internal ordered energy (inergy) E represented in
(2.3.6) in terms of the deviation of the system general potentials Ψi from
their equilibrium values Ψi . The value of inergy may be found directly from
known fields of temperatures, pressures, concentrations, etc. both for the
system as a whole and for each of ith degrees of its freedom. This makes the
inergy one more feasible and extremely “physical” criterion of how far any
system (including a bio-system) stands from equilibrium.
To define within the energy of some system E the non-equilibrium
(convertible) part associated with heterogeneous reactions, let us apply to
relationship (16.1.1). According to it the equilibrium concentration of the kth
substance = c̄ k (at ξr = 1) is equal to = c̄ k = сkо + Σrνkr. Substituting this
expression into relationship (16.1.1) gives:

сk = c̄ k – Σr νkr (1– ξr ). (16.4.1)

314
 Multiplying (16.4.1) by μk and summing over all kth components easily
gives an expression for the chemical potential μm of mixture (Gibbs specific
energy):

μс ≡ Σk μkсk = Σk μk c̄ k + Σr Ar(1– ξr). (16.4.2)

This expression represents the Gibbs specific energy as a sum of the

equilibrium Σkμkc̄ k and non-equilibrium Σr Ar(1– ξr) components. These
components describe independent processes since mass transfer
simultaneously and equally changes both сk and ck leaving the ξr invariable,
whereas chemical reactions, on the contrary, change the coordinates ξr,
while leaving the = c̄ k invariable. Note by the way that the last term in
(9.1.4) is different from zero insofar as the chemical potentials μk of the
reagents being introduced into the system are unequal to their equilibrium
values μ̄ik (at ξr = 1). Thus in locally non-equilibrium systems both
equilibrium and non-equilibrium parts of the Gibbs energy G = ΣkμkМk .
(sometimes termed as chemical energy of the system under consideration)
may be distinguished. Multiplying the Gibbs chemical energy by the mass
of the kth substance Мk = ΣrМkr and introducing (to make the description
general) Zr = Мνkr (1– ξr) as the coordinate of the rth reaction gives:

Uх = Σk μk M k + ΣrArZr . (16.4.3)

Thus the energy of media chemically reacting may also be resolved into
the non-equilibrium (convertible) part which, according to the classification
herein adopted, should be termed as the chemical inergy

Еch = Σr Еr = ΣrArZr , (16.4.4)

and the equilibrium (inconvertible) part of this energy, i.e. the chemical
anergy:

Ūch = Σk μk M k . (16.4.5)

The chemical inergy of a system features its convertible part regardless

of what this energy conversion will be – dissipative or useful (convertible).

315
The latter means doing useful internal work, i.e. conversion of some forms
of chemical inergy into other forms necessary to provide the vital activity.
This is the specific character of bioenergetics. Strictly speaking, chemical
reactions are also caused by spatial heterogeneity in distribution of atoms in
chemical elements within chemical compounds. Chemistry in this sense has
always been a science of substance restructuring. This process has the
vector character when referred to molecular level. However, at macro-level
a chemical reaction acquires the scalar character due to the chaotic
arrangement and motion of all set of atoms and molecules.
Substance restructuring processes are a peculiar analog of chemical
reactions. A convertible (useful) component available in these processes,
differs them from the non-equilibrium phase transitions of the 1st kind, i.e.
the substance restructuring processes imply a transition of some kind of
energy to not only heat (anergy), but also to inergy of other kind. Such are,
in particular, the processes of biological structures formation sometimes
termed as “self-assembly” (G. Gladyshev, 1976).
It can be easily seen that the inergy Еch possesses all properties of the
characteristic state function. Its decrease defines the work being done in the
course of the chemical reaction:

– dЕch = Σr Ardξr . (16.4.6)

Thus the inergy possesses potential properties. Unlike Gibbs energy,

using the chemical inergy does not require the process character
identification (p,T = const). Furthermore, the derivative of Yх with respect
to the reaction coordinate Zr, according to (16.4.4), defines the motive force
of the reaction – its affinity:

Ar = (∂Еch /∂Zr). (16.4.7)

The inergy-based method of chemical equilibrium conditions definition

(∂Е ch /∂λr) = 0, Ar = 0, (16.4.8)

is straighter and shorter than that based on max entropy since does not
require constancy of the equilibrium parameters Θi. This also refers to using
inergy as the system stability criterion

316
 δ2Е ch = – (∂Ar/∂ξr) > 0 . (16.4.9)

One more advantage of inergy consists in the fact that the true
equilibrium Еch = 0 can be distinguished from the so-called “retarded”
equilibrium (when chemical conversions cease because of the absence of
enzymes or catalysts), as well as from the “metastable” (relatively stable)
equilibrium and the “labile” (absolutely unstable) one. These states differ
from the true (full) equilibrium in their inergy that exceeds zero (Еch > 0),
but may decrease if some “activation energy” is supplied to the system.
Such an “energy barrier”, as a matter of fact, exists for most of the forms of
energy. For the thermal form of energy it shows in a certain temperature
difference to be generated (e.g., a certain overheating or overcooling of one
of the phases to initiate phase transition); for mechanical motion of bodies –
in overcoming the static (rest) friction; for chemical and bio-chemical
processes – in supplying “activation energy”; for spontaneous fission of
uranium heavy nuclei – in introducing fast neutrons to enough amount; for
thermonuclear reactions – in raising heat up to tens million degrees, etc.
Availability of such criteria substantially extends the applicability of the
thermodynamic potentials method to bio-systems. Let us consider an
example of the inergy applied to
1.0 M Ech 0.5
develop the maturity criterion for 0.9
organisms in ontogenesis. Bio- M
0.8 0.4
systems are known to be 0.7 Ech
developing by two ways – 0.6 0.3
0.5
extensive and intensive (D. 0.4
εch
0.2
Lourier, H. Vagensberg, 1984). In 0.3
the first case, which is 0.2 0.1
characteristic, e.g., for plankton 0.1
t
0
populations, the inergy flow the 0 0.2 0.4 0.6 0.8 1.0
bio-mass consumes tends to a Lifespan
peak value the particular Fig.16.2. inergy as Maturity Criterion for
Bio-Organisms
environment conditions. In the
second case, which is
characteristic rather for living organisms, the unit consumption (by bio-
system mass or volume) of the free energy (inergy) from the environment
increases in the process of phylogenesis due to a more perfect structure of
the system and a more effective energy conversion in it. The working

317
capacity of an individual living organism adequately depends on its total
mass and specific working capacity. Since the mass M of a living organism
increases, whereas the specific working capacity (per unit of mass)
decreases with its age due to senescence, the total working capacity of a
particular living organism as the production of the above two parameters
most exactly corresponds to the concept of its maturity in ontogenesis (E.
Laifut, 1977, E. Rudakov, 1978) 1).
This statement is illustrated by curves on Fig.16.2 showing the mass M
of a living organism, its specific inergy үх and its total inergy Еch = Мεch
related to max values and plotted against the relative lifespan. The mass of a
living organism is known to be increasing with its development, reaching a
maximum followed by gradual decreasing till its death due to dewatering of
living tissues (G. Gladyshev, 1988). As for the specific inergy, this value,
like the Gibbs specific free energy, steadily decreases in full compliance
with the second law of thermodynamics due to the composition
modification of bio-tissues. Because of such a behavior of the above curves
the total inergy of the system Ech reaches its maximum much earlier than the
mass. The last fact has been verified in multiple observations and may be
considered as a strictly ascertained fact.

16.5. Unity of Technical and Biochemical Energy Converters

The investigation of muscular chemical-into-mechanical energy

converters in a number of guidelines on bioenergetics is based on the TIP
provisions formally applied to bio-systems (S. Keplen, E. Essig, 1968; A.
Rubin, 1984). In this connection the kinetic equations of metabolism and
energy conversion processes in muscular movers are expressed in the form
of Onsager’s linear phenomenological laws (5.1.6) wherein the flows Ji and
forces Xj are derived from the expression for the entropy generation rate
(see 4.1.4). This looks quite logic until the investigator encounters the
efficiency ambiguity for such converters showing in the fact that according
to (4.1.4) the entropy generation may be resolved into the components JiXi
and JjXj voluntarily (S. De Groot, R. Mazur, 1964). In fact, the only TIP

1)
It is relevant to note that using the concept of inergy is very promising for estimate of
“buffering capacity” and “ecosystem development degree”.

318
requirement when choosing flows and forces is that the total entropy
generation be maintained (invariance of dissipation function (5.3.3)
characterizing the dissipation power). Doubts in the TIP applicability grow
even stronger when the different-kind forces Xi and Xj relate to the same
(even or odd) time functions. In this case the Onsarer’s (but nor Cazimir’s)
symmetry conditions must be observed, for which the requirement the
matrix of phenomenological coefficients to be positively defined is
expressed as the restriction (Rij+ Rji)2 < 4RiiRjj. In this case, as will be shown
in Chapter 11, the efficiency of any (including biological) energy converters
can not be in excess of about 17.5%, which does not naturally occur in
reality. The point is that in the energy conversion processes, unlike the
energy transfer, the different-kind flows Ji and Jj are interrelated. This
relation can be revealed based on the law of energy conservation in the form
of (2.6.9), wherefrom it follows that at stationary conditions with reversible
conversion of the ith form of energy into the jth form the power at the
converter input and output is the same:

Хi·Ji = – Хj·Jj . (16.5.1)

The said interrelation complies with the differential reciprocity

relationships which for energy conversion processes are of invariably anti-
symmetrical character:

∂Ji/∂Хj = – ∂Jj/∂Хi . (16.5.2)

Such a character of the reciprocity relationships is caused by opposite

sign of the work done by the forces Хi and Хj and does not depend on
whether they are even or odd time functions (i.e. on the anti-symmetry
conditions Cazimir stated proceeding from the static-mechanical
considerations). This anti-symmetry is of principal importance as resulting
in opposite sign of the terms in the phenomenological laws of energy
conversion, which under linear approximation become:

Ji = Lij Хi – Lij Хj . (16.5.3)

Jj = Lji Хi – Lji Хj . (16.5.4)

319
 Unlike the Onsager’s phenomenological laws, these equations
express decreasing the flow Ji of primary energy carrier with
approaching the “no- load” operation (i.e. with increasing the force Хj to
be overcome). Therefore these are the equations which, with regard to the
energy conversion, should be termed phenomenological (i.e. really
experience-based) rather than Onsager-postulated laws (5.1.6).
Let us consider for example the fibril as a muscle element converting
chemical energy into mechanical one. In this case the generalized rate Ji of
the primary process is construed as the flow Jr of the substances taking part
in this chemical reaction, while the generalized rate Jj of the related process
– as the muscle traction rate. Each of these flows according to (16.5.3)–
(16.5.4) depends on both forces, the first of which, Хi, is now construed as
the negative gradient –grad(Arλr) of chemical reaction affinity, while the
other – the fibril contraction Хj. Note that when the chemical reaction runs
thru to the end (ξr = 1), the force Хi is equal to the chemical reaction affinity
Ar.
Laws (16.5.3) and (16.5.4) may be more conveniently represented in
dimensionless form not containing phenomenological coefficients. To do so,
let us consider the so-called “isometric muscle traction” condition (analog
of the engine/welding transformer no-load operation). At this condition the
force Хj reaches its peak value Хjo (no-load strength) when Jj = 0. This
condition is opposed by the so-called “unloaded muscle traction” condition
(analog of the transformer secondary winding short circuit) when Хj = 0,
while Jj is equal to the short-circuit current Jjk. Expressing Хjо and Jjk
according to (16.5.3) and (16.5.4) in terms of the phenomenological
coefficients gives at Хi being constant:

Хj/Хjо + Jj/Jjk = 1. (16.5.5)

Such a form of the phenomenological laws still more stresses the unity
of energy conversions in any forms. In a more general case of the
phenomenological coefficients as some power functions of the relating
forces Lii = Lii(Хj/Хjо)а and Ljj = Ljj(Jj/Jjk)а , where a is an experimental value,
these equations become non-linear:
(Хj /Хjо)1+n = (1 – Jj /Jjk)1+n . (16.5.6)

320
 The character of such semi-empirical laws with a various value of the
exponent (1+a) equal to 1.0, 1.15, 1.3 and 1.5 is plotted on Fig.16.3 in the
form of curves 1, 2, 3 and 4, respectively. As follows from the figure, at a
=0, the relationship Хj/Хjо vs. Jj/Jjk is linear (straight line 1). The higher the
exponent “a” is, i.e. the stronger the dependence of phenomenological
coefficients on forces is, the more non-linear the curve becomes. However,
this non-linearity is caused by exclusively the non-linearity of laws
(16.5.3)–(16.5.4) proper and has nothing to do with relating the rates of both
processes, i.e. the flows Ji and Jj as sometimes imagined (Keplen, Essig,
1968). Let us, for comparison, consider the Hill’s empirical equation which
accurately enough characterizes the muscles for different kinds of animals:
(Хj /Хjо)1+n = (1 – Jj /Jjk)1+n . (16.5.6)

where æо is a value Hill named the

Jj /Jjk “efficiency index” for muscle. It is
1.0 connected with the value qch =
0.8
1 2 3 4 Lii/(LiiLjj)0,5 he named the “flow
relation factor” by the relationship
0.6 qch= 1/(1 + æо)0,5, i.e. is a function
0.4
of the quality factor æо = Ф – 1 we
earlier introduced. The Hill’s curves
0.2 for various qx well correspond to
Xj /Xjo those plotted on Fig.16.3, i.e. the
0 0.2 0.4 0.6 0.8 1.0
higher the flow relation factor qch is,
the more non-linear the curve
Fig.16.3. Load semi-empirical becomes. At the same time the
characteristics of the fibril curves of (16.5.6) basically differ
from the Hill’s curves in the
interpretation of what causes the non-linearity of the phenomenological
laws. By Hill the linear relationship Хj/Хjо vs. Jj/Jjk corresponds to the value
Lii/(LiiLjj)0,5 = 0 (or Ф = ∞), i.e. to the trivial case of Lii = –Lji = 0. However,
the fact the relation factor becomes zero means no interrelation between the
flows Ji and Jj, i.e. no energy conversion itself! So the non-linearity of
phenomenological laws has appeared to be connected with the muscle
movers’ efficiency. This fact once again confirms that TIP can not be
formally exported to the processes of useful conversion of energy (V. Etkin,
1995).

321
 Conclusions to Part 5

Even more significant are the results obtained from thermokinetic

regarding the theory of transfer irreversible processes. The consistently
thermodynamic (“non-hypothetic”) substantiation of its fundamentals
allows extending this theory to the nonlinear region and states lying far from
equilibrium. Such a generalization has been reached both due to applying
more general (differential) relationships of reciprocity and due to reducing
the transfer laws to a simpler diagonal form not demanding the application
of these relations. This has made it possible to implement a new method of
analyzing superposition effects, which replaces the stationarity conditions
by the partial equilibrium conditions and allows further reduction of the
number of kinetic coefficients in the transfer equations from n(n+1) to n.
This has made it possible to set a number of new relationships between
physical effects and to provide a theoretical substantiation of a number of
known empirical laws.

322
 KEY SIMBOLS

U, u – total (J) and specific (J/kg) energy of a system.

E, ε – total (J) and specific (J/kg) ordered energy (inergy).
Ek, Ep – kinetic and potential ordered energy (J).
Ū, u – total (J) and specific (J/kg) unordered energy (anergy).
H, h – total (J) and specific (J/kg) enthalpy.
G, Γ – Gibbs and Helmholtz energy (J).
ФJ, ПX – local dissipation potentials, W
Θi, θi – total and specific generalized coordinate of the ith scalar
process
Ψi, ψi – generalized potential of the ith scalar process and its local
value; Ψi – the same in equilibrium state.
Fi , М i – resultant motive force (N) and torque (N·m) of the ith vector
process.
X i, xi – specific thermodynamic force of the ith vector process and
its local value.
Zi, zi – total and specific values of distribution moment Θi.
v, w – absolute and relative velocity, m/s.
Ri, Ri – displacement vector of Θi and its modulus.
J i , ji – generalized velocity of the ith vector process (flow) and its
density
Jiс, jiс – displacement flow of the ith energy carrier and its density.
Wi, wi – total (J) and specific (J/kg) work of the ith process.
W o,W u,W D – ordered, unordered and dissipative work, J.
N – process power (capacity), W.
Q, q – total (J) and specific (J/kg) process heat.
đQ, đW – elementary heat and work, J.
q*, w* – specific heat and transfer work of the kth substance, J/kg.
T, Tm – local and mass mean absolute temperature, K.
T1 , T2 – thermodynamic mean temperatures of heat supply and
abstraction, K.
М, Mk – mass of a system and the kth substance therein, kg.
р – absolute pressure, N/m2.
V, υ – total (m3) and specific (m3/kg) volume of a system.
S, s – total (J/K) and specific (J/kg⋅K) entropy of a system.
υk, υkо – partial molar volume of kth component and molar volume of

323
 the pure kth substance m3/mol.
sk, skо – partial molar entropy of the kth component and molar entropy
of the pure kth substance, J/mol⋅K.
сp, сv – isobaric and isochoric specific heat, J/kg⋅K
μk , μk – chemical potential of the kth component and its equilibrium
value, J/kg.
ζ k, ζ k ; – diffusive and osmotic potential of the kth substance in current
ςk, ςk and equilibrium status, respectively.
ck, ck – mass fraction of the kth component in its current and
equilibrium status, kg/kg.
φ – electric potential, V.
ρ – density of a system, kg/m3.
deΘi, duΘi – Θi-elementary variations conditioned by external energy
interchange and internal sources, respectively.
ηt, ηN – thermal and power efficiency.
Р, σs – dissipative function, W/K, and density of entropy internal
source, W/m3·K.
Rij – coefficients of resistance to the ith flow Ji from the jth forces.
Lij – phenomenological coefficients inverse to resistances Rij.
Ar – standard affinity (maximal work) of the rth chemical reaction.
Nkr , νkr – numbers of moles and stoichiometric coefficients of the kth
substances in the rth reaction.
ξr , wr – degree of completeness and rate of the rth chemical reaction.
B, Ф, – load and quality factors of a power installation.
E, Ае – electromotive force and vector potential.
H – magnetic field intensity, A/m.
E – electric field intensity, V/m.
P, D – polarization vector and electric displacement vector of
dielectric unit volume, C/m2.
ω, Ω – angular velocity of rotation and precession motion, 1/s.
ν, λ – radiation frequency and wavelength.
γ, t – Lorentz factor and time.
Ex,ex – total (J) and specific (J/kg) exergy of a system.
Ф – good quality criterion of machine.
В – load criterion of machine.
Gm – gas flow, kg/s.
Э – production output, W.
Зр – machine planned costs,$ .
œ, ĝ – variable component of unit cost and finished product unit
price, $/kg .
r – radius–vector of a field point.

324
 REFERENCES

Abragam A., Proctor W. //Phys. Rev., 1958.- V.109. – Р.1441

Adler C.G., Byrd J.W., Coulter B.L. The 2–nd law efficiency of solar
heating. //Solar Energy, 1981. – 226. – №6. – P.553…558.
Andresen B., Salamon P., Berry R.S. Thermodynamics in finite time:
extremals for imperfect heat engines.//J. Chem. Phys., 1977. – 266. – №4. –
P.1571…1577.
Andressen B., Berry R.S., Nitzan A., Salamon P. Thermodynamics in
finite time. //Phys. Rev. A., 1977. – 215. – №5. –P.2086…2093.
Abramenko T.N. // Ing.-phys. Zhurnal, 1982, V43–№5, p. 804.
Afanasieva-Erenfest T.A. Irreversibility, One-Sided Character and the
Second Law of Thermodynamics. // Zhurnal prikl. phys., 1928, V5, Issues
3-4, pp. 3-30 (in Russian).
Avramenko R.F., Nikolaeva V.I. Quantum Energy of Electron Bose
Condensate in the Environment. M., 1991.
Alemasov V.E. The Theory of Rocket Engines. M.: Oborongiz, 1962
(in Russian).
Andruschenko A.I. Fundamental Engineering Thermodynamics of
Real Processes.–M.: Vysshaya shkola, 1975 (in Russian).
Andruschenko A.I. Fundamental Thermodynamics of Cycles in Power
Plants.–M: Vysshaya shkola, 1985 (in Russian).
Arzelies H. La crise actuelle de la thermodynamique theorie // Nuovo
Cimento, 1966. – 41B. – P. 61.
Avramenko S., Avramenko K. Solid State Space-Energy Generator. //
New Energy News, 1994. – №8 (in Russian).
Avramenko S., Avramenko K. The Russian patent PCT/GB93 /00960,
– 10.05.1993 (in Russian).
Band Y.B., Kafri O., Salamon P. Optimisation of a model external
combustion engine. // J. Appl. Phys., 1982. –253.- №1. –P. 8…28.
Barrer M. Revue General de Thermique. //J. Appl. Phys., 1982, 253.
– №1.
Baur H. Einfuhrung in die Thermodynamik der irreversiblen
Processe. Darmstadt: Wiss. Buchgess, 1984.
Bazarov I.P. Nonequilibrium Thermodynamics and Physical Kinetics.
M.: MGU Publ., 1989, p. 140 (in Russian).
Bazarov I.P. Thermodynamics. Edn 4. M.: Vysshaya shkola, 1991 (in
Russian).

325
 Bakhareva I.F. Nonlinear Nonequilibrium Thermodynamics. Saratov:
Saratov University Publ., 1967, p. 140 (in Russian).
Bearden T.E. Gravitobiology // Library of Congress Catalog Card
USA,1991. –№ 86–50553.

Birjuk V.V. Gas Energy Separation Vorticity in Aeronautical

Engineering and Techniques. // Izv. Vuzov. Aviatechnika. Kazan: KAI.
1993.– №2, pp. 20-23 (in Russian).
Biot M.A. //Chemical Phys., 1977. – V.22. – pp.183…192.
Blumehfeld L.A. Problems of Biophysics.–M, 1977 (in Russian).
Boussinesq J. Théorie de l’écoulement tourbillant. – Paris, 1877 (in
Russian).
Boltzmann, L. Wissenschaftliche Abhandlungen, Vol. I, II, and
III, F. Hasenöhrl (ed.), Leipzig, 1909.
Born M. Critical Notes on Traditional Presentation of
Thermodynamics. // Modern Physics Development.–M.,Nauka, 1964, pp.
224-256 (in Russian).
Bourbaki N. The Architecture of Mathematics //The American
Mathematical Monthly, Vol. 57, No. 4. (1950), pp. 221-232
Brown L.C., Kirkaldy J.S.//Trans. AIME, 1969. –230. – p.223.
Brovkin L.A. On Measurable Heat Content Rise Effect for Solid
Materials. // Ing.-phys. zhurnal, 1960–№5; 1962–№6 (in Russian).
Broda E. Evolution of Bio-Energetic Processes. – Pergamon Press,
Oxford, 1975.
Brodyansky V.M. Perpetuum Mobile – Before and Now.–M.:
Energoatomizdat, 1989 (in Russian).
Bulatov N.K., Lundin A.B. Thermodynamics of Irreversible
Physicochemical Processes.–M.: Khimiya, 1984 (in Russian).
Burdakov V.P. Thermodynamics of Irreversible Processes.–M.: MAI,
1985 (in Russian)..
Caratheodory C. Untersuchungen uber die Grundlagen der
Thermodynamik. //Math. Ann., 1909. – 17. – №3. – P. 355.
Carnot. S. Reflexions sur la puissance motrice du feu et sur les
machines propres a developped cette puissance. Paris, 1824.
Cazimir H.B.G. // Rev. Mod. Rhys., 1945. –17.– P.343.
Chipman J., Bruch E.F. //Trans. AIME, 1968. –242. – P.35…41.
Clausius R. Die mechanische Warmethorie. Braunschweig, 1876.
Bd.1,2.
Cole D.C., Puthoff H.E. Extracting Energy and Heat from the
Vacuum. // Phys. Rev.E, 1993. –48. – №2. – РP. 76…88.
Curie P. // Oeuvres, Paris, 1904. – P.127.
Curson F.L., Ahlborn B. Efficiency of a Carnot engine at maximum
power output. //Amer.J.Phys.,1975. – 43. – P.22…24.

326
 Chambadal P. Development and Application of the Notion of
Entropy –Paris, 1963.
Cherkinsky Y.S. General Thermodynamics.–M.: Polieks, 1992 (in
Russian).
Daniels F.,Alberti R. Physical Chemistry.Wiley, New York, 1966.
Day W.A. Thermodynamics of Simple Media with Memory. Springer
Verlag, 1972.
De Donde Th, van Risselberg P. L’Affinité. – Paris, 1936.
Denbigh K. Thermodynamics of Stationary Reversible Processes.
Methuen, 1951
De Donde T., Van Risselberg P. The Thermodynamic Theory of
Affinity: Book of Principles, Paris, 1933.
De Groot S.R., Mazur R. Non-equilibrium Thermodynamics:
Amsterdam,1962.
Dircks Н. Perpetuum Mobile , 1870.
Gyarmati I. Non-equilibrium Thermodynamics. The Field Theory and
Variational Principles. Springer-Verlag, 1970
Dung M.H., Kozak J.J. Efficiency of light energy conversion in
photogalvanic cells and water cleavage systems. //J. Chem. Phys., 1982. –
277. – №6. – P.3246…3257.
Ebeling W. Structure Formation at Irreversible Processes.– Leipzig,
1976 .
Einstein A. Zur Elektrodynamik bewegter Körper"// Ann. der
Physik. 17, 1905. – S. 891–921.
Einstein A. The Kinetic Theory of Thermal Equilibrium and Second
Law of Thermodynamics.– Ibidem, V3, pp. 34-49.
Einstein A. The Theory of Thermodynamic Fundamentals.– Ibidem,
V3, pp. 50-66.
Einstein A. On a Heuristic Standpoint Regarding Light Generation
and Conversion.–Ibidem, V3, pp. 92-107.
Einstein A. The Planck’s Irradiation Theory and the Specific Heat
Theory.–Ibidem, V3, pp. 134-143.
Einstein A. Über die von der molekularkinetischen Theorie der
Wärme geforderte Bewegung von in ruhenden Flüssigkeiten suspendierten
Teilchen //Ann. der Physik, 1905. –17, s.549-560.
Einstein A. Über einen die Erzeugung und Verwandlung des Lichtes
betreffenden heuristischen Gesichtspunkt //Ann. der Physik, 1905.17, s.
132-148.
Einstein A. Autobiographical Notes (Centennial ed.). Chicago, 1979.
Einstein A., Infeld L. Evolution of Physics. – Cambridge University
Press, 1967.
Eigen M. Self–Organization of Matter and Evolution of Biological
Macromolecules.– Berlin, 1971.

327
 Etkin V.A. The choosing of condense-cooling arrangements′ optimal
characteristics for the power plants. // Thermal Engineering, 1974, 11, pp.
123-127.
Etkin V.A. On Thermodynamics of Reversible and Irreversible
Processes in Methodologically Unified Interpretation. // Collection of Sc.
and Meth. Articles. Heat Engineering.–M.: Vysshaya shkola. 1977, issue 2,
pp.56-60 (in Russian).
Etkin V.A. Axiomatic Problems in Modern Thermodynamics.–M.:
VNTITs., 1978, 106 p. (in Russian).
Etkin V.A. Phenomenological Derivation of Reciprocal Relations in
Thermodynamics of Irreversible Processes. //Chemical Thermodynamics
and Thermochemistry.–M.: Nauka, 1979, pp. 8-13 (in Russian).
Etkin V.A. On Maximum Efficiency of Non-Heat Engines. // Heat
Engineering.–M.: Vysshaya shkola, 1980, issue 3, pp. 43-51(in Russian).
Etkin V.A. Proof of differential reciprocal relation in non-linear
systems. // Russian Journal of Physical Chemistry, 1982, 56(5), pp. 345-
352.
Etkin V.A. Verification of Differential Relations in Nonlinear
Diffusion Processes. // Thermal Conductivity and Diffusion.–Riga, 1983,
issue 12, pp. 57-71 (in Russian).
Etkin V.A. The Method of Technical Work Coordinates Definition. //
Izv. Vuzov. Energetika, 1985–№6, pp. 86-95 (in Russian).
Etkin V.A. To the thermodynamic theory of non-linear irreversible
processes.// Russian Journal of Physical Chemistry, 1985, 59(3), pp. 2246-
2249.
Etkin V.A. To Solution to the Problem of Thermodynamic
Inequalities. // Izv. Sib. otd. AN SSSR. Ser. tehn. nauk, 1988, №15, issue 4,
pp. 34-39 Etkin V.A. Technical Working Ability of Nonequilibrium
Systems. // Sib. fiz.-tehn. zhurn., 1991, issue 6, pp. 72-76.
Etkin V.A. The fundamental equation of non-equilibrium thermody-
namics. // Russian Journal of Physical Chemistry, 1988, 62(8).- pp.1157-
1159.
Etkin V.A. Paradoxes of Thermodynamics.–M.: VTTITs, 1988.p. 90
(in russian).
Etkin V.A. Heat and Work in Irreversible Processes. // Izv. vuzov.
Energetika, 1988, No4, pp. 118-122 (in russian).
Etkin V.A. To the decision of problem of thermodynamic
inequalities. // Soviet. Journal of Appl. Physics, 1988, 4 (15), pp. 274-279.

328
 Etkin V.A. Origin of Onsager′s reciprocal relations. // Soviet. Journal
of Appl. Physics, 1989, 4, pp. 288-293.
Etkin V.A. Uniqueness of driving forces of irreversible processes.
// Russian Journal of Physical Chemistry, 1989, 63(6), pp. 918-919.
Etkin V.A. To the non-equilibrium thermodynamics of energy
transformation systems. // Soviet. Journal of Appl. Physics, 1990.- 6, 720-
725.
Etkin V.A. Thermodynamics of non-equilibrium processes of energy
transfer and energy conversion. - Saratov: State Univers. Press, 1991.-168 p.
(in Russian).
Etkin V.A. Method of studying linear and non-linear irreversible
processes. // Russian Journal of Physical Chemistry, 1991, 65(3), pp. 339-
343.
Etkin V.A. Technical Working Ability of Nonequilibrium Systems.
// Sib. fiz.-tehn. zhurn., 1991, issue 6, pp. 72-76 (in Russian).
Etkin V.A. Exergy as a criterion of the evolution, equilibrium, and
stability of thermodynamic systems. // Russian Journal of Physical
Chemistry, 1992, 66 (5), pp. 1205-1212.
Etkin V.A. Reciprocal relations of irreversible processes. // Soviet.
Journal of Appl. Physics, 1993, 1, pp.62-64.
Etkin V.A. The form of law of the many-component diffusion. //
Russian Journal of Physical Chemistry, 1994, 68(12), pp. 518-522.
Etkin V. To the similarity theory of power plants. //Atti del 49o
Congresso Nat. ATI.- Perugia, 1994. –V.4. – P.433…443.
Etkin V.A. To the non-equilibrium thermodynamics of biological
systems. //Biophysics, 1995, 40(3), pp. 661-669.
Etkin V.A. Synthesis and New Application of the Energy Transfer and
Conversion Theories: Thesis for Dr. Science (Technology) degree.–M.:
1998, 213 p.
Etkin V.A. Optimization of rocket engine specific corrective thrust.
//Russian Aeronautics, 1999, №1, pp. 105-107.
Etkin V.A. Thermokinetics (thermodynamics of non-equilibrium
processes of energy transfer and energy conversion). – Togliatti, 1999.- 228
p. ( in Russian).
Etkin V.A. Alternative form of the generalized transfer laws. //
Journal of Engin. Physics and Thermophysics, 2000, 72(4), pp. 748-754.
Etkin V.A. To the Thermodynamic Theory of Engineering Systems
Productivity. // Izv. AN SSSR. Energetika, 2000, №1, pp. 99-106 (in
Russian).
Etkin V.A. Maximum Power Conditions for Nuclear Power Plant
Cycles. // Teploenergetika, 2000, №3, pp. 48-51 (in Russian).
Etkin V.A. Fuel-Free Power Engineering: New Vistas. // Bulletin of
Scientific Center in Haifa, 2003, V1, pp. 67-71(in Russian).

329
 Etkin V.A. At the Interfaces between Natural Sciences. // Bulletin of
Scientific Center in Haifa, 2005, V5, pp. 42-43 (in Russian).
Etkin V.A. Energy and Anergy. // Bulletin of Scientific Center in
Haifa, 2006, V9, pp. 30-38 (in Russian).

Etkin V.A. To the Unified Theory of Real Processes. // Transactions of

Congress “Fundamental Problems of Natural Science and Engineering”, V1,
St.-Petersburg, 2006, pp. 577-587 (in Russian).
Etkin V.A. Nontrivial Consequences of System Approach to Physics.
// System Investigations and Control of Open Systems, 2006, issue 2, pp.
39-44 (in Russian).
Etkin V.A. On the Grounds of Quantum Mechanics. // Bulletin of
Scientific Center in Haifa, 2006, V10, pp. 19-27.
Etkin V.A. Many-Sided Entropy. // Bulletin of Scientific Center in
Haifa, 2007, V11, pp. 15-20.
Etkin V.A. Energodynamics (Synthesis of Energy Transfer and
Conversion Theories). SPb.: Nauka, 2008, 409 p. (in Russian).
Etkin V.A. To the uniform theory of a field. // Bulletin of Scientific
Center in Haifa, 2009, V19, pp. 17-23 (in Russian).
Ewans R.B., Watson . G.M., Truitt . J. //Journ. Appl. Phys., 1963. –
233(9). – P.2682…2688; 234(7). – P.2020…2026.
Faier V., Ross J. On the efficiency of termal engines with power
output. //J. Chem. Phys., 1981. – 275. – P.5485…5496.
Fairen V., Hatlee M., Ross J. Thermodinamic processes, times, scale,
and entropy production. //J.Phys. Chem., 1982. –286. –№1. – P.70…73.
Favorsky O.N., Fishgoit V.V., Yantovsky E.I. Theoretical Principles of
Space Electrojet Propulsion Units.–M.: Vysshaya shkola, 1970 (in Russian).
Feynman R., Leyton R., Sands M. Feynman’s Lectures on Physics
(2nd ed.).– Addison-Westley,1970
Feynman R. Character of Physical Laws.– London, 1965.
Fitts D.D. Nonequilibrium Thermodinamics. New York–Toronto–
London, 1962, 168 p.
Fourier J.B. Theorie analytique de la chaleur. Paris, 1822.
Fick A. Uber Diffusion. Berlin, 1855.
Ford R. Perpetuum Mobile Enigma, 1987.
Gauss K.F. Transactions, V5, Royal Scientific Society, Göttingen,
1967.
Gelfer J.M. History and Methodology of Thermodynamics and
Statistical Physics. Edn 2.–M, 1981 (in Russian).
Gerber P. // Z. Math. Phys., 1898. – 43. – Р. 93...104.
Geskin E., Spakovsky M.R. Towards the development of a generalized
energy equation. //Ecos’95, July 11-15, Istambul, А1–156. – Р.75…79
Gibbs J.W. //Trans. Connecticut Academy, 1875. –V.3. – P.108…248.

330
 Gibbs J.W. The Scientific Papers, New York, 1906. – V.1.
Gladyshev G.P. Thermodynamics and Macrokinetics of Natural
Hierarchical Processes.–M.: Nauka, 1988, 287 p. (in Russian).
Glensdorf P., Prigogine I. The Thermodynamic Theory of Structure,
Stability and Fluctuations, 1973, 280 p.
Gyarmati I. Introduction to Irreversible Thermodynamics. –
Budapest, 1960.
Gould R. A Book of Unexplained Facts, 1965.
Goldman M. Spin Temperature and NMR in Solids. Clarendon
Press, Oxford, 1970.
Graneou Р. Electromagnetic jet–populsion in the direction of carrent
flow. // Nature, 1982.– 295. 1982 №5847. – Р. 311…313.
Guhman A.A. To Grounds of Thermodynamics.–M.:
Energoatomizdat, 1986 (in Russian).
Gurov K.P. Phenomenological Thermodynamics of Irreversible
Processes.–M.: Nauka, 1978, p. 128 (in Russian).
Gutkowicz D., Procaccia I., Ross J. Efficiency of rate processes. //
Phys., 1978. –269. – № 9. – P.3898–3906.
James L.G. Apparatus for Heating Fluids», U.S. Patent №5.188.090
Joule J. The Scientific Papers, London, 1884. –V.1.
Haken H. Synergy.– Springer-Verlag, 1982
Hamilton W. On General Method in Dynamics (Variational Principle
of Mechanics) // British Association Report, 1834, pp. 513–518.
Helmholtz H. On the Theory of Electrodynamics. // J. für
Mathematik, 1870.-Part I, V. 72 ; 1873.-Part II, V. 75; 1874.-Part III,
V. 78.
Helmholtz. H. Preface to “Principles of Mechanics Stated in New
Respect” by H Hertz.–AN SSSR, 1959, p. 296.
Hertz H. On Fundamental Equations in Electrodynamics of Moving
Bodies.- Frankfurt, 1890.
Hill T. L. Free energy transduction in biology. //N.Y. Acad. press,
1977, 299 p.
Hilbert D. Die Grundlagen der Physik.- Götting. Nachrichten, 1915. –
V.1.
Hilsch R. Die Expansion von gasen im Zentrifugalfeld als Kаlten-
process. // Z.Naturforsch., 1946. –V.1. – S.206.
Hugoniot H. Complex Rendus de l’ Acad. Des Sciences, Paris, 1880.
Katchalsky A., Curran P.F. Nonequilibrium Thermodynamics in
Biophysics. Harvard Univ. Press, Cambridge, 1967.
Kafri O., Band Y.B., Levine R.D. Is work output optimized in a
reversible operation. //Chem. Phys. Lett., 1981. –277. – №3. – P.441.
Kalafati D.D. Nuclear Power Plant Thermodynamic Cycles.–M.:
Gosenergoizdat, 1963 (in Russian).

331
 Kaganovich B.M., Filippov S.P., Equilibrium Thermodynamics and
Mathematical Programming.// Novosibirsk: Nauka, 1995 (in Russian).
Kanarev F.M. Fundamentals of Microcosm Physical Chemistry.
Krasnodar, 2004 (in Russian).
Kedrov B.M. Gibbs’ Paradox.–M.: Nauka, 1969 (in Ruaaian)
Keplen S.R., Essig E. Bioenergetics and Linear Thermodynamics of
Irreversible Processes, 1966.
Keren E., Kafri O., Band Y.B. Optimisation of the energy output in
pulsed lasers. //J. Appl. Phys., 1982. – 253.- №3. – P.1373…1380.
Kirkaldy J.S., Purdy G.R.//Canad. J. Phys, 1962. – 40(2). – P.208.
Kirillin V., Sychev V.V., Sheindlin A.E. Engineering
Thermodynamics.–M.: Nauka, 1979 (in Russian).
Klimantovich Y.L. Entropy Decrease in the Process of Self-
Organization. S-Theorem. Letters to Zhurnal tehnicheskoi phiziki, 1983,
V8, 1412 p. (in Russian).
Klimantovich Y.L. The Statistical Theory of Open Systems, V1, 2, 3;
M.: Yanus, 1995, 1998, 2002 (in Russian).
Kosinov N., Garbaruk V. Vacuum origin of Electron and Positron”.
Physical Vacuum and Nature, 2000 –№4 (in Russian).
Kosinov N.V. Physical Vacuum Electrodynamics.// Physical Vacuum
and Nature, 1999, №1, pp. 24-59 (in Russian).
Kotin A.F. On the Second Form of the First Law of Thermodynamics
// Collected Sc.-Meth. Articles. Thermodynamics.–M.: Vysshaya shkola,
1976, issue 1, pp. 50-55 (in Russian).
Krichevsky R.I. Notions and Grounds of Thermodynamics.–M.:
Khimiya, 1970.
Krishtal M.A., Volkov A.I. Multi-Component Diffusion in Crystalls.–
M.: Metallurgiya, 1985.
Krishtal M.M Evolution of the Temperature Field and Deformation at
Faltering Fluidity.// Metallurgical science and thermal processing of
metals, 2003, №4 (in Russian).
Kubo R. Thermodynamics. Amsterdam, 1968..
Kvasnikov I.A. Thermodynamics and the Statistical Theory of
Nonequilibrium Systems.–M.: MGU Publ., 1987, p. 559 (in Russian).
Lagrange J. Mecanique Analytique, Paris,1811.
Landau L.D., Lifshitz L.M. Mechanics, Elsevier, Amsterdam ...
Tokyo, 1975.
Lavenda B.H. Thermodynamics of irreversible processes. – London,
1978, 182 p.
Lebon G., Jou D. //J.Cont.Phys. – 33.- №1. – P.41…51.
Li J.C.M.//Journ.Chem. Phys., 1958. –28. – P.747…750.
Lighfoot E. Transfer Phenomena in Living Systems.– Elsevier, 1967.
Linden C. at al. A new developments of the thermodinamics in finite
times. //Nature J., 1992. – 215. – №4. – P.249…253.

332
 Lykov A.B., Mikhailov Y.A. Heat-Mass Transfer Theory.–M.: GEI,
1963 (in Russian).
London H. //Nature, 1938 – 142. – P.612; Proc.Roy. Soc. London A,
1939. – 171. – P.484.
Lorenz G.A. Versuch einer Theorie der elektrischen und optischen
Erscheinungen in bewegten Korpern. Leiden, 1895.
Lorenz G.A. Lectures on Thermodynamics.–London, 1927.
Mason E.A., Wendt R.P., Bresler E.H. //J. Chem. Farad. Trans., 1972.
– 268(11). –P.1938–1950.
Mamontov M.A. Basic Thermodynamics of Variable-Mass Bodies.–
Tula, 1970 (in Russian)..
Maxwell J.C. Treatise on Electricity and Magnetism., vol,.2. Oxford,
1873.
Meixner I. Thermodynamik der irreversiblen Processe. – Aachen,
1954, 178 s.
Merkulov A.P. The Vorticity Effect and its Application to
Technique.–M.: Mashinostroenie, 1969 (in Russian).
Millis R . Grand Unified Theory of Classical Quantum Mechanics,
1997. 557 p.
Millis M. Challenge to Create the Space Drive, Journal of Propulsion
and Power, 1997. – №13. – Р.577…582.
Muchnik G.F., Rubashov I.B. The Methods of the Heat Exchange
Theory.–M.: Vysshaya shkola, 1974, V.1(3) (in Russian).
Munster A. Thermodynamic des processus irreversibles. – Paris,
1966, 268 p.
Mozurkewich H. Berry R.S. Optimization of a heat engine, based on a
dissipative system. //J. Appl. Phys., 1983. – 254. – №7. – P.3651.
Moray T.H., Moray J.E.The Sea of Energy / Cospray Research
Institute, 1978 .
Navier L.M.A. // Mem. Acad. Sciences, 1822. – 389. 389–№.1.
Nefedov E.I., Protopopov A.A., Yashin A.A. A System of
Electrodynamic Equations for Transverse and Longitudinal Waves.// Wave
Process Physics and Radio Engineering Systems, 1998, V1–No1
Nesterov B.V., Korovin N.V., Brodyansky V.M. On Exergic Efficiency
of Electrochemical Current Sources.// Electrokhimiya, 1980.–16, issue 6,
pp. 814-820 (in Russian)..
Nesterov S.B. Hydrino, Dehydrino and Energy Generation.// Energia,
1955, №6 (in Russian).
Nernst W. Theoretical and Experimental Grounds of the New Thermal
Law.–Stutgard, 1913.
Newton I. Mathematical Principles of Natural Philosophy (1678). 1st
American Edition, 1846.
Nicholls D. Bioenergetics. Academic Press, London-New York, 1979.

333
 Nicolaev G. On the longtudinal Electromagnetic waves. //Deutsche
Physik, 1993.– Vol. 2. – №8. – Р. 24…30.
Niper H. A. Revolution in Technik, Medizin, Gesellschaft.
//Gravitational Field Energy Research in Japan, 1983. – Р. 68…71.
Novikov I.I. Thermodynamics.–M.: Mashinostroenie, 1984 (in
Russian).
Onsager L. Reciprocal relations in irreversible processes. //Phys.
Rev., 1931. – 237(14). – P.405…426; 238(12). – P.2265…2279.
Ohm G.S. Die galvanishe Kette mathematisch bearbeitet. –Berlin,
1827.
Ondrechem M.J., Rubin M.H., Band Y.B. The generalized Carnot
cycle. //J. Chem. Phys., 1983. – 278. – №7. – P.4721…4727.
Ott H. //Zeitshr. Phys., 1963. – V.70. – S.75.
Periodic table of the elements. – VWR Scientific Products,
2000.
Petrov P.N., Brankov J. Present Problems of Thermodynamics.–Sofia,
1982.
Physical Encyclopedia.–M.: Sovetskaya Entsiklopedia, 1984 (in
Russian).
Postma P. Verlag Ak. K. Watenshap. – Amsterdam, 1911. – B.17. –
Р.339…351.
Planck M. Thermodynamics.–Berlin,1905.
Prandtl L. Gesammelte Abhandlungen zur angewandten Mechanik
und Hydro–aerodynamik. – Berlin, 1961. – T.12.
Prigogine I. Etude Thermodynamique des Phenomenes Irreversibles.
– Liege, 1947, 143 p.
Prigogine I. Order and Haos, Man’s new dialog with Nature. –
London, 1984.
Prigogine I. Introduction to Thermodynamics of Irreversible
Processes.– New York ,1961.
Prigogine,I., Glansdorff,P. Thermodynamics Theory of
Structure, Stability and Fluctuations. - London: Wiley, 1971.
Prigogine I., Nikolis J. Self-Organization in Non-Equilibrium
Systems.- Wiley,1977.
Prigogine I. Time, Structure and Fluctuations (Nobel lecture on
chemistry in 1977.
Prigogin I. From Being To Becoming. –Freeman, 1980.
Prigogine I; Stengers I. Order out of Chaos: Man's new
dialogue with nature.- Flamingo,1984
Prigogine I. From Existing to Appearing.– Wiley, 1985.
Puthoff H.E., Cole D.C. Extracting Energy and Heat from the
Vacuum. // Phys. Rev. E., 1993 – V.48 – № 2.
Pitzer K. // Pure Appl. Chem., 1961. – 2. – P.207…210.

334
 Ramsey N.F. Thermodynamics and Statistical mechanics by Negative
Absolute Temperature. // Phys. Rev. 1956. – V.103. – №1. – P. 279.
Rauscher E., Bise W. Harnessing the Earth–Ionosphere Resonant
Cavity. // Harnessing the Wheelwork of Nature: Tesla’s Science of Energy,
2002. –233. №З.
Risselberghe P. // Journ. Chem. Phys., 1962. – 36.- P.1329...1332.
Rysselberghe P. Thermodynamics of irreversible Processes. – Paris –
New York – Toronto – London, 1963, 112 p.
Roschin V.V., Godin S.M.. An Experimental investigation of the
Physical Effects in a Dynamic Magnetic System. // New Energy
Technologies, 2001. №1.
Rozonoer L.I., Tsirlin A.M. Optimal Control of Thermodynamic
Processes.// Avtomatika i telemehanika, 1983, No1–pp. 70-79.; No2–pp. 88-
101; No2–pp. 50-64 9 (in Russian).
Rubin A.B. Thermodynamics of Biological Processes.–M.: MGU,
1984 (in Russian).
Rubin M.H. Optimal configuration of a class of irreversible heat
engines. //Phys. Rev. A., 1979. – 219. –№3. – P.1272…1276; 1980. – 222.
– №4. – P. 1741…1752.
Rudenko A.P. The Self-Development Theory for Open Catalytic
Systems.–M.: Nauka, 1969 (in Russian).
Rudenko A.B., Orlov V.N. Finite Time Thermodynamics (review).//
Teploenergetika, 1984, №9, pp. 68-71 (in Russian).
Salamon P., Nitzan A., Andressen B., Berry R.S. Minimum entropy
production and the optimization of heat engines. //Phys. Rev. A., 1980. –
221. – №6. – P.2115…2129.
Salamon P., Nitzan A. Finite time optimizations of a Newton law
Carnot cycle.//J. Chem. Phys., 1981. – 274. – №5. – P. 3546…3560.
Sedov L.I. Continuum Mechanics.–M.: Nauka, 1971, V.1 (in
Russian).
Shenk H., Kais H.//Arc. Eisenhutten, 1960.– 31(4). – S.227.
Shipov G.I. The Theory of Physical Vacuum.–M.: Nauka, 1997 (in
Russian).
Sychev V.V. Complex Thermodynamic Systems.–M.: Energia, 1970
(in Russian).
Sienintyez S., Salamon P. Finite–Time Tremodynamics and
Thermoeconomics, 1990, 308 p.
Shoulders K. R. «Energy Conversion Using High Charge Density»,
U.S. Patent №5.018.180, May 21, 1991.
Stratonovich R.L. Nonlinear Nonequilibrium Thermodynamics,–M.:
Nauka, 1985 (in Russian).
Strebkov D.S., Avramenko S.V., Nekrasov A.I., Roschin O.A. On the
Possibility of Single-Line Power Transmission.// Tehnika v selskom
hozyaistve, 2004, №4, pp. 35-36 (in Russian).

335
 Szargut J., Petela R. Exergy.–Varshava, 1963.
Tamm I.E. A. Einstein and Modern Physics. In collection “Einstein
and Modern Physics”.–M.: 1956, p. 92 (in Russian).
Terentieva Z.A., Kobozev N.I. On Possible Relation between
Informational Theory Entropy and Thermodynamic Entropy.// Zhurn. Fiz.
Khimii, 1976, V50–№4, pp. 877-881 (in Russian).
Tesla N. Improvements relating to the transmission of Electrical
Energy, 1906. (US Patent number №8,200).
The Heat Exchange Theory. Terms.–M.: Nauka, 1980 (in Russian)
Thermodynamics. Terms.–M.: Nauka, 1973 (in Russian)
Tolman R. Relativity, Thermodynamics and Cosmology.–
Oxford,1969.
Tomson W. Mathematical and physical papers. Cambridge, 1882. –
V.1.
Tribus M. Thermostatics and Thermodynamics.– New York,1961
Tribus M., Evans R.B., Crellеn G.L. Thermoeconomics. /Principles of
desalination. – New York: Acad. Press, 1966. – Vol.23. – P.77…101.
Truesdell, C. Rational Thermodynamics, 2nd ed.; Springer-
Verlag: New York, 1984.
Тurrel H.J. Diffusion and Heat flou in Liquids. – L.: Willey, 1961.
Umov A.I. Selected Works.–M., L.: 1950, 203 p.
Ulam S. Unsolved Mathematical Problems.– New York 1960.
Van-der-Waals J., Konstamm F. Course of Thermostatics, V1, 1936.
Vant Hoff J.H. Die Chemishe Dynamik. – Braubnschweig, 1901.
Veinik A.I. Thermodynamics of Irreversible Processes.–Minsk:
Vysheishaya shkola, 1966, 359 p. (in Russian).
Veinik A.I. Thermodynamics of Real Processes.–Minsk, 1991. (in
Russian).
Vukalovich M.P., Novikov I.I. Engineering Thermodynamics.–M.:
Energia, 1968 (in Russian).
Weber W. Werke. – Berlin: Springer, 1894. – Vol. 4. – S. 247...299.
Zaev N.E., Avramenko S.V., Lisin V.N. Measurement of Conduction
Current Induced by Polarization Current.// Zhurnal russkoi fizicheskoi
mysli, 1991.– №2.– PP. 18-224 1994.– №10.– PP. 8-9 (in Russian).
Zhuravlev V.A. Thermodynamics of Irreversible Processes in
Problems and Solutions.–M.: Nauka, 1979 (in Russian).
Zhabotinsky A.M. Concentration Self-Oscillations.–M.: Nauka, 1974,
178 p.
Zeldovich J.B. The Vacuum Theory, May Be, Solves the Cosmology
Problem.// UFN, 1981.–V133, issue 3 (in Russian).
Zemansky, M. (1981). Temperatures Very Low and Very High.-
Dover, 1957.
Ziegler H. Extreme Principles of Thermodynamics of Irreversible
Processes. Continuum Mechanics.– Berlin, 1966.

336
 Contents

Introduction………………………………………………………………………...3

Part 1
THERMODYNAMICS AS THEORETICAL
BASIS OF HEAT ENGINEERING

Chapter 1. Methodological Principles of the Thermokinetics…………..10

1.1. Exclusion of Hypotheses and Postulates from Theory Grounds………………12

1.2. System Approach to Objects of Investigation………………………………..15
1.3. Negation of Process and System Idealization outside the Framework of
Uniqueness Conditions………………………………………………………...17
1.4. Number-of-Degrees-of-Freedom Theorem as a Basis of Real Processes
Classification ………………………………………………………………….24
1.5. Extension of Variables Space with Introduction of Spatial Heterogeneity
Parameters for System as a Whole ……………………………………………26
1.6. Coordinates of Non-Equilibrium Redistribution and Reorientation
Processes... .………...........................................................................................30

Chapter 2. Specificity of Thermokinetics in Application

to Spatially Heterogeneous Systems…………………………..31

2.1. Substantiation of Total Energy Conservation Law…………………………….33

2.2. Imparting Identity Character to Fundamental Relationship of
Thermokinetics…………………………………………………………………35
2.3. Introducing Process Time and Capacity into Equations
of Thermodynamics………...............................................................................38
2.4. Elimination of Dissipative Terms from Energy Balance Equations…………..41
2.5. Discrimination between Convertible and Inconvertible Parts of Energy.
Inergy and Anergy, Their Properties as Characteristic Functions……..………43
2.6. Definition of General Form for Equations of State and Transfer……………...52

Conclusions to Part 1.................................................................................................56

337
 Part 2.
HEAT ENGINEERING FUNDAMENTALS

Chapter 3. Classic Thermodynamics………………………………………...58

3.1. Principle of Excluded Perpetual Motion of the 1st Kind

(First Law of Thermodynamics) ……………………………………….............59
3.2. Principle of Entropy Existence (Second Law for Reversible
Processes) …………………………………………………………………….64
3.3. Principle of Entropy Rise (Second Law for Irreversible
Processes) ………………………………………………………………..........69
3.4. Principle of Excluded Perpetual Motion of the 2nd Kind …………………….72
3.5. Principle of Unattainable Absolute Temperature Zero
(Third Law of Thermodynamics) …………………………..............................75
3.6. Principle of Entropy Maximum as Equilibrium Condition……………………78

Chapter 4. Thermodynamics of Irreversible Processes………………….83

4.1. Theory of Dissipation Process Rate in the Making.......................................... 83

4.2. Motive Forces and Generalized Rates of Transfer Processes............................87
4.3. Entropy Balance Equation and Dissipation Rate...............................................88
4.4. Phenomenological Laws of Transfer Processes….............................................90
4.5. Reciprocal Relations in Transfer Processes.......................................................93
4.6. Law of Minimum Entropy Production…..........................................................97

Chapter 5. Heat-Mass Transfer Theory…………………………………..100

5.1. Elimination of Strange Delimitation of Thermodynamics

and the Heat Exchange Theory......................................................................100
5.2. Conductive Heat Exchange (Heat Conductivity)..........................................101
5.3. Convective Heat Exchange...........................................................................103
5.4. Radiant Heat Exchange….............................................................................104
5.5. Heat Exchange and Mass Exchange in Open Systems…..............................105

Conclusions to Part 2…...........................................................................................110

Part 3.
ELIMINATION OF NEGATIVE CONSEQUENCIES
OF THERMODYNAMICS EXTRAPOLATION

Chapter 6. Correction of Thermodynamics of the Open and

Polyvariant systems ………………………………………..111

6.1. Distinction between Ordered and Unordered Works………………………..111

338
6.2. Substantiation of Second Law applicability to
Non-Heat Engines……………………...........................................................117
6.3. Specification of Material Equilibrium Conditions………………………….123
6.4. Definition of Potential State Functions for Open Systems.............................127
6.5. Solution to the Problem of Thermodynamic Inequalities...............................128
6.6. Possible Use of Environmental Heat in Non-Heat Engines ………………...133

Chapter 7. Thermodynamic Resolution of Gibbs Paradox…………..136

7.1. Origin and Nature of Gibbs Paradox…............................................................136

7.2. Thermodynamic Inadmissibility of the Gibbs’ Paradox…..............................139
7.3. Entropy Reference Point Shift in Mixing Process as Entropy
“Jump” Reason….........................................................................................143
7.4. Thermodynamic Theory of Mixing Processes ……………………………....146

Chapter 8. Denial of the Theory of “Heat Death” of the Universe…...149

8.1. Reasons of Appearing Concept of “Heat Death” of the Universe …………..149

8.2. Discrimination between Thermodynamic and Statistical-Informational
Entropies …………………………………………………………………….151
8.3. Correction of Entropy–Dissipation and Entropy–Irreversibility Relations …155
8.4. Inconsistence of Entropy Rise Principle…………………………………….156
8.5. Inadmissibility of Entropy Rise Principle as Extrapolated to the Universe…159
8.6. Processes in the Universe as Reciprocal Antipodes…………………………162

Chapter 9. Paralogism of Negative Absolute Temperatures……….166

9.1. Negative Absolute Temperature Notion Origin...............................................166

9.2. “Inversion” of the Second Law in Negative Absolute
Temperature Range…………………………………………………………..168
9.3. Non-Thermal Nature of Spin-Spin Interaction…...........................................171
9.4. Description of Spin Systems from the Positions of Thermokinetics..............172

Chapter 10. Paralogisms of Relativistic Thermodynamics…………...174

10.1. Ambiguity of Relativistic Transformations of Thermodynamic Values.......177

10.2. Non-Invariance of Expression for Efficiency of Relativistic
Carnot Cycle………………………………………………………………..178
10.3 Relativistic Carnot Engine as Compound Engine.........................................179
10.4. Inapplicability of Relativity to Absolute Values………………………......180

Conclusions to Part 3...................................................................................182

339
 Part 4
FURHER GENERALIZATION OF HEAT ENGINE THEORY

Chapter 11. Theory of Similarity of Energy Conversion Processes….183

11.1. Unity of Energy Transfer and Conversion Processes...................................184

11.2. Universal Index of Energy Converters’ Perfection.....................................187
11.3. Kinetic Equations of Energy Conversion Processes.
Anti-Asymmetry of Relationships Therein................................................191
11.4. Similarity Criteria for Energy Converting Systems....................................196
11.5. Universal Load Characteristics of Energy Converters.................................198
11.6. Similarity of Load Characteristics for Real Machines................................201

Chapter 12. Theory of Productivity of Engineering Systems……...207

12.1. Synthesis of Thermokinetics with Thermoeconomics..................................207

12.2. Nominal Operating Conditions for Power Units..........................................210
12.3. Optimal Overload Degree of Power and Processing Plants........................212
12.4. “Cruising Speed” Operation of Transport Vehicles....................................214
12.5. Peak Power Reach Conditions for Heat Power Plant..................................216

Chapter 13. Converters of Field Forms of Energy…………………...223

13.1. Theoretical Possibilities of Creating Alternators.........................................223

13.2. Gravitational Energy Converters..................................................................227
13.3. Generators Using Electrical Field Energy....................................................232
13.4. Current Generators Using Magnetic Field Energy.......................................237
13.5. Converters of Radiation Field Energy.........................................................247

Conclusions to Part 4.............................................................................................254

Part 5
EXTENSION OF ENERGY TRANSFER THEORY APPLICABILITY

Chapter 14. Theory of Nonlinear Transfer Processes………………..256

14.1. Proof of Existing Motive Forces’ Resultant in Independent

Transfer Processes........................................................................................257
14.2. Superposition Effects as Found from Partial
Equilibrium Conditions................................................................................261
14.3. Value of Superposition Effects as Predicted from Known

340
 Thermodynamic Parameters......................................................................263
14.4. Simplification of Phenomenological Laws without Onsager’s
Reciprocal Relations Applied....................................................................267
14.5. Investigation of Membrane Processes with the Reciprocal
Relations Violated........................................................................................274

Chapter 15. Generalization of the Theory to Systems Standing

Far Away from Equilibrium..............................................279

15.1. Reciprocal Relations for Chemical Reactions.............................................279

15.2. Investigation of Irreversible Processes in Systems Standing
Far Away from Equilibrium.........................................................................283
15.3. Simplification of Transport Laws Based on Differential
Reciprocal Relations...................................................................................289
15.4. Further Reduction of the Number of Kinetic Factors
in Transport Equations..................................................................................293
15.5. Setting Additional Constraints between Superposition Effects...................297

Chapter 16. Specificity of Transfer Processes in Bio-Systems……….301

16.1. Vector Character of Reversible Biochemical Reactions…………… …….302

162. Interrelation between Active and Passive Transport Processes
in Bio-Systems……………………………………………………………..304
16.3. The Role of Ordered Energy in Vital Activity of Bio-Systems……….......307
16.4. Maximum of Inergy as Maturity Criterion for Bio-Organisms…………...313
16.5. Unity of Technical and Biochemical Energy Converters……………….. ..318

Conclusions to Part 5...............................................................................................322

Key Symbols................................................................................................323
References....................................................................................................325
Contents.......................................................................................................337

341