Donatello Annaratone - Transient Heat Transfer PDF
Donatello Annaratone - Transient Heat Transfer PDF
Donatello Annaratone - Transient Heat Transfer PDF
1 Introduction
Let us consider Fig. 1 where the warm side of the wall is at temperature t1 while
the cold side is at temperature t2; due to this temperature difference we know that a
certain amount of heat transfers from the warm to the cold side.
Clearly, the heat entering the warm side is produced by an external source
which transfers it to the wall; then the same heat is transferred from the cold wall
to an external source.
Let us assume that the external source corresponding to the cold side suddenly
transfers heat to it, thus increasing its temperature. For sake of simplicity, we assume
that the temperature reaches t1, i.e., the identical temperature of the other side.
Clearly, at this point the heat transfer through the wall stops because the
temperature difference between the two sides that caused it no longer exists.
The temperature on the inside of the wall is lower compared to that of both
sides, thus creating a heat flux towards the inside from both sides and progressively
increasing the inside temperatures until the latter stabilize, depending on certain
physical characteristics of the material. The example shows how the entire wall
reaches temperature t1, and the heat transfer through the wall stops.
Naturally the phenomenon takes place in the same way, even though to a
different degree, even when the temperature increase of the cold side is lower than
the one assumed earlier. In that case, when balance is reached the heat transfer
through the wall corresponds to the new difference in temperature between the
sides.
Based on our assumptions and as a result of the unbalances in temperature,
there is an increase in heat content of the wall.
Of course, the opposite takes place if the temperature of the warm wall suddenly
drops. In that case there is a slow decrease in temperature inside the wall at the
expense of a heat flux moving towards both sides until balance is reached, and the
heat transfer through the wall is in agreement with the new temperatures of the sides.
t
1
t
2
Let us consider the cubic element shown in Fig. 2 and assume that the element is
crossed by heat only in the direction x.
If conditions are unsteady, heat dQ1 enters the cubic element and heat dQ2 exits
the element. The latter differs from dQ1 (under steady conditions both types of heat
are, of course, identical); the heat dQ is stored in the cubic element and is equal to
dQ ¼ dQ1 dQ2 : ð3Þ
If the left side of the element registers the thermal gradient qt/qx (the derivative
is partial given that t also depends on time), based on Fourier’s well-known law,
within time dh and given that the surface crossed by the heat is equal to dydz, we
have
ot
dQ1 ¼ kdydz dh: ð4Þ
ox
If we move to the other side of the cubic element, i.e., at distance dx from the
previous one, the thermal gradient is given by
ot o ot ot o2 t
þ dx ¼ þ 2 dx: ð5Þ
ox ox ox ox ox
Therefore,
ot o2 t
dQ2 ¼ kdydzdh þ 2 dx : ð6Þ
dx ox
dy
dQ
2
dQ
1
dz x
y dx
4 Transient Heat Transfer
o2 t
dQ ¼ kdydzdh dx: ð7Þ
ox2
This heat increases the temperature of the cube; within time dh this increase is
equal to
ot
dh: ð8Þ
oh
The heat dQ is equal to the volumetric specific heat of the material (given, as
we already saw, by the product of the specific heat referred to mass c by the
density of material q) multiplied by the volume of the cubic element and by the
noted increase in temperature. Therefore,
ot
dQ ¼ cqdxdydz dh: ð9Þ
oh
A comparison between (9) and (7) leads to
ot k o2 t
¼ : ð10Þ
oh cq ox2
Equation 10 represents Fourier’s general law of thermal conductivity.
As we can see, the partial derivative of temperature with respect to time goes up
in sync with an increase in thermal diffusivity. Consequently, variations in tem-
perature are faster and the duration of the unsteady state is reduced.
The integration of this differential equation makes it possible in principle to
compute the development of temperature t over time for every point of the wall.
As shown, (10) was derived assuming that the heat transfer occurs only in one
direction (x); this is the most frequent case, and from now on we shall always refer
to this scenario. Please, note that if there is heat flux also in directions y and z the
general law of thermal conduction changes as follows:
k o2 t o2 t o2 t
ot
¼ þ þ : ð11Þ
oh cq ox2 oy2 oz2
Based on (10) we obtain rather complicated equations for the computation of
t on which we shall not dwell; it suffices to notice that they are valid in principle
but during an attempt to apply them to a specific case of interest, it is established
that some of them do not satisfy the borderline conditions of space and time.
Therefore, it is required to use the equation or the equations that satisfy these
conditions.
The analysis of the equations in question clearly shows that the approach to
solve the issues of unsteady state is anything but easy.
Only few physically elementary situations can be studied in a relatively simple
way.
2 General Law of Thermal Conduction 5
1.0
0.9
0.8
0.7
0.6
(2 xaθ )
0.5
0.4
f
0.3
0.2
0.1
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0
x
2 aθ
Fig. 3 Function f pxffiffiffiffi
2 ah
is none other than Gauss’ error integral; its values may be taken from Table 1 or
from Fig. 3; it is zero for g = 0 and it is equal to 1 for g ¼ 1.
3 Heat Diffusion in Plate of Infinite Thickness 7
At the beginning of the phenomenon, thus for h = 0, given that h is time, and
for x = 0, i.e., on the surface of the wall the value of g is zero, this means that the
integral in question is zero.
Recalling (12) and the fact that the superficial temperature is increased to t1 we
have
t1 ¼ A: ð14Þ
Considering the generic section of the wall with abscissa x in the beginning of
the phenomenon, i.e., for h = 0; this leads to g ¼ 1, and as a result the integral is
equal to one.
Based on (12) and recalling that the wall is at temperature t0, we obtain
t ¼ t1 þ Bx þ C ¼ t0 : ð15Þ
Equation 15 is in contrast with the fact that the entire wall is at temperature t0;
therefore, it is necessary to set B = 0; this leads to
C ¼ t0 t1 : ð16Þ
Thus, based on (12)
x
t ¼ t1 þ ðt0 t1 Þf pffiffiffiffiffi ð17Þ
2 ah
where f( ) indicates the error integral shown in Fig. 3; note that a stands for
thermal diffusivity.
Thus, Eq. 17 makes it possible to compute the temperature in every section of
the wall identified by the abscissa x and for any time indicated by h.
It is interesting to establish the existing relationship between the variation in
temperature occurring after a certain time in any location of the plate and the
difference in temperature ðt1 t0 Þ between the temperature t1 of the side and the
initial one of the plate.
Based on (17)
t t0 x
u¼ ¼ 1 f pffiffiffiffiffi : ð18Þ
t1 t0 2 ah
As far as the phenomenon we are examining, Fourier’s dimensionless number
(Fo) is crucial; with reference to the generic position of the plate (thus considering
distance x from the side brought to temperature t1), it is given by
kh ah
Fo ¼ ¼ ð19Þ
cqx2 x2
where a as usual represents thermal diffusivity.
From (18) and based on (19) we obtain
1
u¼1 pffiffiffiffiffiffi: ð20Þ
2 Fo
8 Transient Heat Transfer
In the beginning of the previous section we pointed out that Eq. 17 refers to a plate
of infinite thickness; nonetheless, the above is valid also for a plate of finite
thickness provided, of course, that the value of x is less than the thickness xw of the
plate, and that the heat crossing the other side is zero or irrelevant.
4 Heat Transfer in Plate of Finite Thickness 9
4.1 Examples
(1) Let us consider a steel plate at a given initial temperature; one of the sides is
brought and kept at a higher temperature; for this type of analysis the value of
these two temperatures is unimportant.
For steel we may assume that c = 500 J/kg K; q = 7850 kg/m3; k = 45 W/m K.
The goal is to determine after how much time the heat passing at a distance of
60 mm from the heated side is equal to 60% of the heat hitting the plate.
Based on Table 2 for v = 0.6 we obtain Fo = 0.4894.
Thus, the required time is given by h = 0.4894 9 500 9 7850 9 0.062/
45 = 153 s, i.e., about 2.5 min.
In order for the procedure to be valid (26) must be satisfied; so the thickness xw of
qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
45153
the plate must be xw 4 5007850 ¼ 0:167 m ¼ 167 mm.
10 Transient Heat Transfer
1.1
100 1.0
0.9
80 0.8
0.7
60 0.6
0.5
χ
40 0.4
0.3
20 0.2
0.1
0 0.0
0 18 36 54 72 90 108 126 144 162 180
x (mm)
5 Plate Immersed in Fluid 11
t (°C)
1.1
100 1.0
0.9
80 0.8
0.7
60 0.6
0.5
χ
40 0.4
0.3
20 0.2
0.1
0 0.0
0 24 48 72 96 120 144 168 192 216 240
x (mm)
As stated in the introduction, in this case as well as in the following ones, we shall
use a finite differences program.
Let us consider a flat plate (e.g., a steel plate), at a given initial temperature,
which is immersed in a fluid at a different temperature that either increases or
decreases its temperature; we know that the plate reaches the temperature of the
fluid in a theoretically infinite time.
On the other hand, if we conventionally assume that this process may practi-
cally be considered complete when the difference in temperature between fluid and
plate is reduced to 5% of its initial one (in other words, the increase or decrease in
temperature is equal to 95% of the initial difference in temperature) it is possible to
identify the required time.
The time in question indicated by h, is a function of various quantities,
specifically: the thickness xw of the plate, the specific heat c, the density q, the
thermal conductivity k of the material of the plate, and finally the heat transfer
coefficient a of the fluid surrounding the plate.
We know that for the transient phenomena we are discussing the dimensionless
number of Fourier is fundamental. By indicating it with Fo, it is given by
k
Fo ¼ h: ð29Þ
cqx2w
By analyzing the condition described above, i.e., the increase or decrease in
temperature equal to 95%, we see how the various quantities in question influence
the value of both h and Fo.
12 Transient Heat Transfer
We establish that the value of h is proportional to the value of c and q; the specific
heat and the density of the material do not impact the value of Fourier’s number.
A decrease of the value of k increases the value of h but this increase is
generally modest; in any case it is far from proportionality between h and the
reciprocal of k; thus, the value of Fo decreases with a decrease of the value of k.
If the thickness of the plate increases the value of h goes up but the increase is
not such to lead to proportionality between h and the square of xw; so if xw
increases, Fo decreases.
Finally, if we consider the heat transfer coefficient a of the surrounding fluid
which is not included in Fourier’s number, it is obvious that an increase of a causes
the value of h (and consequently of Fo) to decrease; in fact, the increase of a
facilitates the heat transfer from the fluid to the plate and viceversa.
Therefore, we establish that Fo decreases if a and xw increase, and if
k decreases.
At this point it becomes natural to consider Biot’s dimensionless number
indicated by Bi, and given by
axw
Bi ¼ : ð30Þ
k
Please, note that Biot’s number is formally identical to Nusselt’s number; the
difference consists in the fact that Nusselt’s number considers the values of a and
k in reference to a fluid and considers the diameter of the tube where the fluid flows
instead of the thickness xw; this is why this number is fundamental in the analysis
of heat transfer by convection from a fluid to the wall of a tube; on the other hand,
while Biot’s number considers the heat transfer coefficient a of the fluid, it also
analyzes the thermal conductivity of the material of the solid and its thickness;
these are in fact the quantities of interest to us.
By analyzing the values of h and Fo in reference to variations in the quantities
in question, we establish that the value of Fo univocally depends on the value of
Bi; in other words:
Fo ¼ f ðBiÞ: ð31Þ
The values of Fo as a function of Bi for increase or decrease in temperature of
the plate of 95% are listed in the last column in Table 3.
Note that the increase or decrease in temperature of the plate may also be
partial; in particular, it is possible to use a fluid at a higher temperature compared
to the final temperature desired for the plate.
Thus, it is interesting to be able to have the values of Fo as a function of Bi even
in these cases.
Table 3 shows the values of Fo for a temperature increase or decrease equal to
30, 40, 50, 60, 70, 80 and 90% of the initial temperature difference between fluid
and plate.
Note that the final temperature of the plate is the average one across the
thickness; it is important not to forget that the temperature of the plate varies
across the thickness depending on the circumstances; in the case of temperature
5 Plate Immersed in Fluid 13
increase of the plate the temperature in the middle is lower compared to that on the
sides of the plate; in the case of decrease of temperature the temperature in the
middle is higher than the one on the sides.
At this point the analysis must go further.
We believe, in fact, that instead of referring to Fourier’s number it is more
appropriate to refer to a new dimensionless number.
It is the product of Fourier’s number by Biot’s number.
By indicating it with X, it is therefore equal to
ah
X ¼ Fo Bi ¼ : ð32Þ
cqxw
14 Transient Heat Transfer
Please note that in comparison with Fo this dimensionless number does not
include k; a is included instead, and the thickness xw is no longer squared.
The values of X as a function of Bi and of the percentage of temperature
increase or decrease are shown in Table 4.
The analysis of the values of X leads to interesting considerations.
Note that for values of Bi ranging from 0.01 and 0.1 X is practically
independent from Bi; this means that the influence of k on h is irrelevant and
that the time h is about inversely proportional to a and about proportional to
thickness xw.
For higher values of Bi and specifically for extremely high values of Bi the
impact of k on h starts to be noticeable, the impact of a on h decreases while the
impact of xw on h creases.
In fact, if we consider the values of X shown in the table for Bi [ 1:
X ¼ X1 Bib ð33Þ
where X1 represents the value of X for Bi = 1 and exponent b varies for all the
examined instances between 0.16 and 0.38.
Therefore,
ah ax b
w
¼ X1 ð34Þ
cqxw k
then
x1þb
w
h : ð35Þ
a1 b k b
This establishes that the exponent of xw is greater than one, whereas the
exponent of a is smaller than one, to the contrary of what happens for values of Bi
ranging from 0.01 and 0.1; finally, this highlights the impact of k on h.
This greater influence of k on h is due to the fact that, as we shall see later on,
for the high values of Bi the difference between the temperature of the sides of the
plate and the mean value of the temperature across the thickness is more high-
lighted, difference which is obviously influenced by the value of k.
Once the value of X is known through Table 4 based on the value of Bi, the time
h required to obtain the desired result is computed through (32), given that
cqxw
h¼X : ð36Þ
a
Clearly, the computation is particularly easy and it is possible to avoid using a
finite differences program each time that would be necessary to solve this task.
At this point let us look at a few examples.
5.1 Examples
(1) Let us consider a steel plate with a thickness of 45 mm, initially at a tem-
perature of 300°C which is to be cooled through air to a temperature of 20°C.
Considering that c = 500 J/kg K; q = 7850 kg/m3; k = 45 W/m K; a = 20
W/m2 K.
Biot’s number is therefore equal to 20 9 0.045/45 = 0.02.
If we consider a decrease in temperature equal to 95%, i.e., a final temperature
of the metal plate equal to 34°C, based on Table 4 we obtain X = 1.4538.
Based on (36) we have h = 1.4538 9 500 9 7850 9 0.045/20 = 12839 s
corresponding to 214 min., thus about 3 and a half hours.
16 Transient Heat Transfer
140
120
t (°C)
100
5502 s
80
60
9171 s
40
12839 s
20
0 5 10 15 20 25 30 35 40 45
x (mm)
Moreover, note that based on the values of X, the steel plate goes from 300 to
216°C in 1548 s, to 160°C in 3010 s and to 104°C in 5228 s.
The behavior of the temperatures through the wall and depending on time
variations is shown in Fig. 6.
(2) Let us now consider a steel plate with a thickness of 90 mm at an initial
temperature of 20°C to be heated to 300°C through a fluid at 420°C.
Considering that c = 500 J/kg K; q = 7850 kg/m3; k = 45 W/m K;
a = 200 W/m2 K.
Biot’s number is equal to 200 9 0.09 9 45 = 0.4.
The heating is equal to 280/400 = 0.7 = 70%.
Based on Table 4 X = 0.6304.
Therefore, based on (36) h = 0.6304 9 500 9 7850 9 0.09/200 = 1113 s,
corresponding to about 18 min.
In addition, note that based on the values of X, the plate goes from 20 to 140°C
in 328 s and to 220°C in 639 s.
If the plate were to be heated to 400°C, 2771 s would be required, and that
would correspond to about 46 min.
The behavior of the temperatures through the wall and depending on time
variations is shown in Fig. 7.
(3) Let us consider a plate made of refractory material with a thickness of 75 mm,
and a temperature of 500°C to be reduced through air to 20°C.
Considering that c = 1200 J/kg K; q = 2000 kg/m3; k = 1.5 W/m K;
a = 20 W/m2 K.
Biot’s number is equal to 20 9 0.075/1.5 = 1.
Let us also consider cooling equal to 95% so that the plate reaches a
temperature of 44°C.
5 Plate Immersed in Fluid 17
260
240 795 s
220
t (°C)
200
180
477 s
160
140
120
239 s
100
0 9 18 27 36 45 54 63 72 81 90
x (mm)
160
6654 s
140
120
100
80
11089 s
60
40
15525 s
20
0.0 7.5 15.0 22.5 30.0 37.5 45.0 52.5 60.0 67.5 75.0
x (mm)
(4) Assuming at this point that the same plate of the previous example is to be
heated from 20 to 160°C with a fluid at 300°C with a heat transfer coefficient
of 100 W/m2 K.
Heating is equal to 140/280 = 0.5 = 50%.
Biot’s number is equal to 100 9 0.075/1.5 = 5.
Based on Table 4, X = 0.605.
Thus, h = 0.605 9 1200 9 2000 9 0.075/100 = 1089 s, equal to about
18 min.
The behavior of the temperatures through the wall and depending on time
variations is shown in Fig. 9.
Even in this case the analysis of the impact of the various quantities on the time
required to obtain a certain result shows that the value of Fourier’s number solely
depends on Biot’s number.
In view of the previous section we will not compute Fourier’s number as a
function of Bi and analyze the values of the dimensionless number X instead, given
that we consider it more significant for our study.
Let us examine a plate licked by a fluid on just one side either receiving heat
from the fluid or transferring heat to it assuming either no heat loss through the
other side or a negligible one.
Under these assumptions it was possible to build Table 5 listing the values of
dimensionless number X as a function of Biot’s number (Bi).
180
mean value after 1089 s
160
140
t (°C)
1089 s
120
100 778 s
80
60 467 s
40
233 s
20
0.0 7.5 15.0 22.5 30.0 37.5 45.0 52.5 60.0 67.5 75.0
x (mm)
6 Plate Heated or Cooled on One Side 19
The analysis of the values of X in this area establishes that with reference to
(35) the exponent b is never greater than 0.026; the impact of k on h remains
modest.
If we examine the values of X for Bi = 0.1–1 the exponent b varies from 0.034
to 0.113.
Finally, if we look at the values of X for Bi = 1–10 the exponent b varies from
0.18 to 0.5.
Even in this case it is best to cite some examples.
For a useful comparison with the plate immersed in fluid we consider two
examples already discussed in the previous section.
6.1 Examples
(1) Let us consider a steel plate with a thickness of 90 mm, initially at 20°C to be
heated to 300°C through a fluid at 420°C licking only one side.
Considering that c = 500 J/kg K; q = 7850 kg/m3; k = 45 W/m K;
a = 200 W/m2 K.
Biot’s number is equal to 200 9 0.09/45 = 0.4.
The heating is equal to 280/400 = 0.7 = 70%.
Based on Table 5, X = 1.3384.
Thus, h = 1.3384 9 500 9 7850 9 0.09/200 = 2364 s, corresponding to
about 39 min.
Comparing this result with that of the plate immersed in fluid we establish that
with the fluid licking the plate only on one side the required time more than
doubles.
The behavior of the temperatures for various amounts of time including, of
course, 2364 s, is shown in Fig. 10.
(2) Let us assume a plate made of refractory material with a thickness of 75 mm,
initially at 20°C to be heated to 160°C through a fluid at 300°C licking only on
one side.
Considering that c = 1200 J/kg K; q = 2000 kg/m3; k = 1.5 W/m K;
a = 100 W/m2 K.
Biot’s number is equal to 100 9 0.075/1.5 = 5.
Heating is equal to 140/280 = 0.5 = 50%.
Based on Table 5, X = 1.7145.
Therefore, h = 1.7145 9 1200 9 2000 9 0.075/100 = 3086 s, correspond-
ing to about 51 min.
A comparison of the value obtained with the one corresponding to a plate
immersed in fluid shows that the required time is almost triple.
The behavior of the temperatures for various amounts of time, such as 3086 s,
is shown in Fig. 11.
7 Tube Heated or Cooled on Outside Surface 21
260
240
1689 s
220
t (°C)
200
180
160 1013 s
140
120
507 s
100
80
0 9 18 27 36 45 54 63 72 81 90
x (mm)
200
180
mean value after 3086 s
160
t (°C)
140
120 3086 s
100
2204 s
80
60
1323 s
40
661 s
20
0.0 7.5 15.0 22.5 30.0 37.5 45.0 52.5 60.0 67.5 75.0
x (mm)
In this section we examine a tube heated or cooled through a fluid licking it on the
outside surface, while the heat transfer to the internal fluid through the inside
surface either does not take place or is negligible.
This case is similar to the one discussed in the previous section regarding a
plate licked by a fluid on just one side.
22 Transient Heat Transfer
Regardless of their similarity the two cases differ due to the curvature of the
wall of the tube; thus, given equal thickness and equal values for the other
quantities they produce different values of the dimensionless number X.
As we know, time h which is necessary to cause a certain event, depends on the
value of X; therefore, given equal conditions, timing for the tube is different
compared to the plate.
By first approximation (but in almost all instances the obtained results are very
close to reality) it is possible to use the following equation:
xw
X ¼ Xp 1 : ð37Þ
Do
In (37) Xp is the value of X relative to the plate (Table 5), xw is the thickness of
the wall of the tube, and Do is the outside diameter of the tube.
In the case of a bar instead, where xw/Do = 0.5, based on (37) we
obtain X = 0.5Xp; in one of the following examples we will discuss this case, as
well.
To facilitate reading Table 6 lists the values of the dimensionless number X for
xw/Do = 0.1, and Table 7 lists the values for X per xw/Do = 0.2.
Table 6 Dimensionless number X for tube heated or cooled on outside surface (xw/Do = 0.1)
Bi Tube heating or cooling (%)
20 30 40 50 60 70 80
0.010 0.1980 0.3165 0.4509 0.6116 0.8021 1.0522 1.3847
0.015 0.1984 0.3171 0.4541 0.6154 0.8104 1.0558 1.4112
0.020 0.1986 0.3176 0.4549 0.6171 0.8150 1.0666 1.4125
0.025 0.1991 0.3183 0.4550 0.6174 0.8162 1.0670 1.4263
0.030 0.1994 0.3188 0.4566 0.6184 0.8175 1.0742 1.4287
0.040 0.2000 0.3198 0.4582 0.6205 0.8202 1.0777 1.4409
0.050 0.2007 0.3209 0.4596 0.6225 0.8246 1.0814 1.4456
0.060 0.2013 0.3219 0.4612 0.6245 0.8273 1.0870 1.4504
0.070 0.2023 0.3230 0.4626 0.6278 0.8300 1.0907 1.4552
0.080 0.2030 0.3240 0.4641 0.6299 0.8327 1.0942 1.4599
0.10 0.2042 0.3259 0.4671 0.6339 0.8381 1.1014 1.4724
0.12 0.2054 0.3287 0.4700 0.6380 0.8435 1.1085 1.4821
0.15 0.2077 0.3317 0.4744 0.6440 0.8516 1.1193 1.4964
0.20 0.2111 0.3367 0.4828 0.6542 0.8651 1.1372 1.5205
0.25 0.2144 0.3423 0.4900 0.6656 0.8786 1.1550 1.5445
0.30 0.2177 0.3479 0.4983 0.6756 0.8921 1.1729 1.5687
0.40 0.2237 0.3576 0.5128 0.6957 0.9209 1.2112 1.6171
0.50 0.2294 0.3678 0.5281 0.7171 0.9478 1.2471 1.6690
0.60 0.2355 0.3778 0.5422 0.7370 0.9767 1.2831 1.7178
0.80 0.2461 0.3974 0.5723 0.7780 1.0304 1.3578 1.8157
1.0 0.2563 0.4157 0.5997 0.8187 1.0861 1.4300 1.9178
7 Tube Heated or Cooled on Outside Surface 23
Table 7 Dimensionless number X for tube heated or cooled on outside surface (xw/Do = 0.2)
Bi Tube heating or cooling (%)
20 30 40 50 60 70 80
0.010 0.1760 0.2810 0.4023 0.5435 0.7120 0.9359 1.2295
0.015 0.1763 0.2817 0.4035 0.5469 0.7193 0.9456 1.2531
0.020 0.1769 0.2822 0.4042 0.5478 0.7241 0.9468 1.2582
0.025 0.1772 0.2829 0.4048 0.5488 0.7254 0.9531 1.2680
0.030 0.1773 0.2834 0.4055 0.5502 0.7274 0.9548 1.2702
0.040 0.1780 0.2845 0.4068 0.5521 0.7298 0.9589 1.2810
0.050 0.1789 0.2855 0.4085 0.5538 0.7322 0.9621 1.2862
0.060 0.1795 0.2864 0.4099 0.5557 0.7346 0.9652 1.2894
0.070 0.1804 0.2876 0.4112 0.5580 0.7370 0.9684 1.2938
0.080 0.1810 0.2885 0.4129 0.5598 0.7394 0.9715 1.2992
0.10 0.1824 0.2906 0.4156 0.5634 0.7449 0.9780 1.3077
0.12 0.1839 0.2927 0.4186 0.5676 0.7498 0.9853 1.3163
0.15 0.1859 0.2960 0.4229 0.5730 0.7577 0.9948 1.3291
0.20 0.1893 0.3011 0.4304 0.5832 0.7697 1.0107 1.3518
0.25 0.1927 0.3061 0.4373 0.5928 0.7825 1.0276 1.3728
0.30 0.1960 0.3114 0.4446 0.6023 0.7953 1.0446 1.3957
0.40 0.2023 0.3214 0.4590 0.6215 0.8201 1.0776 1.4387
0.50 0.2083 0.3312 0.4727 0.6405 0.8458 1.1107 1.4834
0.60 0.2142 0.3406 0.4868 0.6597 0.8707 1.1439 1.5283
0.80 0.2252 0.3593 0.5143 0.6971 0.9213 1.2105 1.6171
1.0 0.2355 0.3773 0.5411 0.7343 0.9712 1.2774 1.7080
7.1 Examples
(1) Let us consider a steel tube with an inside radius ri = 40 mm and an outside
radius ro = 50 mm.
Thus xw/Do = 10/100 = 0.1.
Moreover, c = 500 J/kg K; q = 7850 kg/m3; k = 45 W/m K; a = 100 W/m2 K.
The initial temperature of the tube is equal to 20°C; in order to increase it to 220°C
through the external fluid which is at 520°C.
Heating is equal to (220 - 20)/(520 - 20) = 0.4 = 40%.
Biot’s number is equal to 100 9 0.01/45 = 0.02222.
Based on Table 6, X = 0.4549.
The required time is therefore equal to h = 0.4549 9 500 9 7850 9 0.01/
100 = 178.5 s.
Figure 12 shows the behavior of the temperatures in the tube for various time
periods, including of course h = 178.5 s; the abscissa shows radius r.
(2) Let us consider a steel tube with an inside radius ri = 30 mm and an outside
radius ro = 50 mm.
Thus, xw/Do = 20/100 = 0.2.
In addition, c = 500 J/kg K; q = 7850 kg/m3; k = 45 W/m K; a =
100 W/m2 K.
24 Transient Heat Transfer
200
180 128 s
160
t (°C)
140
77 s
120
100
80 38 s
60
40 41 42 43 44 45 46 47 48 49 50
r (mm)
200 539 s
180
t (°C)
160
323 s
140
120
100
162 s
80
30 32 34 36 38 40 42 44 46 48 50
r (mm)
200
858 s
180
160
t (°C)
140
515 s
120
100
80 257 s
60
0 5 10 15 20 25 30 35 40 45 50
r (mm)
(3) Let us now consider a steel bar with the same characteristics of the previous
example but, of course, with ri = 0 and xw = 50 mm. The temperatures in
question are the same.
Thus, we obtain Bi = 100 9 0.05/45 = 0.1111.
Based on Table 5, Xp = 1.2238 and therefore based on (37), X = 0.5 9
1.2238 = 0.6119.
Therefore, h = 0.6119 9 500 9 7850 9 0.05/100 = 1201 s.
Figure 14 shows the behavior of temperatures for different time periods
including, of course, h = 1201 s. We establish that after this time the mean
temperature of the bar is very close to the expected 230°C.
We are interested in studying the time required to obtain a certain heating of the
plate through radiation on both sides from a heat source at a given temperature; we
assume it is represented by two walls facing each other on the sides of the plate.
Two operations are required in order to be able to use what was established for
the plate immersed in fluid.
The first consists of identifying an ideal heat transfer coefficient for radiation
which allows us to obtain a value of Biot’s number to enter Table 4 and compute
from value of X the value of h.
The second consists of identifying the existing ratio between the surface tem-
perature of the plate which is of interest for radiation, and the mean temperature of
the wall we always referred to in previous computations.
26 Transient Heat Transfer
We also observe that, given that the surface temperature of the plate varies over
time, as far as radiation we will consider the mean temperature between the start
and the end of the process.
Finally, the computation will have to be conducted by a trial and error process,
as we shall see later on.
Let us start by considering the first problem.
As usual, if a indicates the heat transfer coefficient which is an ideal value for
radiation in this case and recalling what is known about radiation:
a Tf Ts ¼ Bm r Tf4 Ts4
ð38Þ
then
a ¼ Bm r Tf þ Ts Tf2 þ Ts2 :
ð39Þ
Table 8 Factor Y
tf ts
80 100 120 140 160 180 200 220 240 260
400 33.64 35.15 36.75 38.44 40.21 42.07
420 35.91 37.48 39.13 40.86 42.68 44.59 46.59
440 38.30 39.92 41.61 43.39 45.27 47.23 49.28
460 40.80 42.47 44.21 46.05 47.97 49.98 52.09 54.30
480 43.42 45.13 46.93 48.82 50.79 52.86 55.03 57.29
500 46.15 47.92 49.77 51.71 53.74 55.86 58.08 60.40 62.82
520 49.01 50.83 52.73 54.72 56.81 58.99 61.26 63.64 66.12
540 51.99 53.86 55.82 57.86 60.00 62.24 64.57 67.01 69.55 72.20
560 55.10 57.02 59.03 61.14 63.33 65.63 68.02 70.51 73.11 75.82
580 58.33 60.31 62.38 64.54 66.79 69.14 71.60 74.15 76.81 79.58
600 61.70 63.74 65.86 68.08 70.39 72.80 75.31 77.93 80.65 83.49
620 65.21 67.30 69.48 71.75 74.12 76.60 79.17 81.85 84.63 87.53
640 68.85 71.00 73.23 75.57 78.00 80.53 83.17 85.91 88.76 91.72
660 72.63 74.83 77.13 79.53 82.02 84.62 87.31 90.12 93.03 96.06
680 76.55 78.82 81.18 83.63 86.19 88.85 91.61 94.48 97.46 100.55
700 80.62 82.95 85.37 87.89 90.50 93.23 96.05 98.99 102.04 105.20
720 84.84 87.23 89.71 92.29 94.97 97.76 100.65 103.65 106.77 110.00
740 89.21 91.66 94.21 96.85 99.60 102.45 105.41 108.48 111.66 114.96
760 93.74 96.25 98.86 101.57 104.38 107.30 110.32 113.46 116.71 120.08
780 98.42 100.99 103.67 106.44 109.32 112.30 115.40 118.61 121.93 125.37
800 103.26 105.90 108.64 111.48 114.42 117.48 120.64 123.92 127.31 130.82
820 108.26 110.96 113.77 116.68 119.69 122.82 126.05 129.40 132.86 136.45
840 113.43 116.20 119.07 122.05 125.13 128.33 131.63 135.05 138.59 142.25
860 118.76 121.60 124.54 127.59 130.74 134.01 137.38 140.88 144.49 148.22
880 124.27 127.17 130.18 133.30 136.53 139.86 143.31 146.88 150.57 154.37
900 129.94 132.92 136.00 139.19 142.49 145.90 149.42 153.06 156.82 160.71
920 135.80 138.84 142.00 145.26 148.63 152.11 155.71 159.42 163.26 167.23
940 141.83 144.95 148.17 151.50 154.95 158.50 162.18 165.97 169.89 173.93
960 148.04 151.23 154.53 157.94 161.45 165.09 168.84 172.71 176.70 180.82
980 154.44 157.70 161.07 164.55 168.15 171.86 175.69 179.64 183.71 187.91
1000 161.06 164.36 167.80 171.36 175.03 178.82 182.73 186.75 190.91 195.19
tf ts
280 300 320 340 360 380 400 420 440 460
580 82.46
600 86.43
620 90.54 93.67
640 94.80 97.99
660 99.20 102.46 105.84
680 103.76 107.09 110.54
700 108.47 111.87 115.39 119.03
720 113.34 116.81 120.40 124.12
740 118.38 121.91 125.58 129.37 133.28
(continued)
28 Transient Heat Transfer
Table 8 (continued)
tf ts
280 300 320 340 360 380 400 420 440 460
760 123.57 127.18 130.91 134.78 138.77
780 128.93 132.61 136.42 140.36 144.43 148.63
800 134.46 138.21 142.10 146.11 150.26 154.54
820 140.16 143.99 147.95 152.04 156.26 160.62 165.11
840 146.03 149.94 153.97 158.14 162.44 166.88 171.46
860 152.08 156.06 160.18 164.42 168.80 173.32 177.98 182.78
880 158.31 162.37 166.56 170.89 175.35 179.94 184.69 189.57
900 164.72 168.86 173.13 177.53 182.08 186.76 191.58 196.55 201.66
920 171.32 175.53 179.89 184.37 188.99 193.76 198.66 203.72 208.92
940 178.10 182.40 186.83 191.40 196.10 200.95 205.94 211.08 216.37 221.81
960 185.07 189.45 193.97 198.62 203.41 208.34 213.42 218.64 224.02 229.54
980 192.24 196.70 201.30 206.03 210.91 215.93 221.09 226.40 231.86 237.48
1000 199.60 204.15 208.83 213.65 218.61 223.71 228.96 234.36 239.92 245.62
Note that Biot’s number is unknown until the value of a is known, and this
depends, as we saw, on Y, i.e., on ts; the latter value also depends on Biot’s
number. This is why proceeding by trial and error is required, as we shall see
through the examples. Generally, though, it suffices to proceed through two steps
to obtain values of the different parameters which further steps would not
significantly vary.
Once Biot’s number is defined and the heating percentage is known, the value
of the dimensionless number X, as well as time h, can be obtained through Table 4.
The result is approximate; in fact, if it is true that a certain value of a and Bi
lead to a certain value of X and h through Table 4, we should not forget the fact
that the value of a is obtained based on the assumption that the temperature of
reference of the surface of the plate is the mean value between the temperature at
the beginning of the process and the one at the end of it; this is a conventional
temperature which may not correspond to the real phenomenon, given that this is
radiation; nonetheless, such solution is at least the simplest one and other solutions
would be difficult to identify.
At this point examples are the best way to clarify the required procedure.
8.1 Examples
(1) Let us increase the temperature of a steel plate with a thickness of 60 mm from
20 to 460°C through radiation on both sides.
The heat source is assumed to be at 900°C and the overall black level Bm to be
equal to 0.6.
Moreover, c = 500 J/kg K; q = 7850 kg/m3; k = 45 W/m K.
8 Plate Radiated on Both Sides 29
420
755 s
380
t (°C)
340
582 s
300
260
220
330 s
180
0 6 12 18 24 30 36 42 48 54 60
x (mm)
The mean surface temperature between beginning and end of the heating is
therefore not equal to 240°C, as assumed earlier, but equal to (20 ? 470)/
2 = 245°C.
Consequently, the value of Y is equal to 157.8 and a = 94.68 W/m2 K.
Biot’s number turns out to be equal to 0.126, and at this point the process by
trial and error may be considered complete; further passages would not lead to
significant variations in the various parameters.
Based on the value of Bi and on the heating percentage from Table 4, we
obtain X = 0.3467.
Therefore, h = 0.3467 9 500 9 7850 9 0.06/94.68 = 862 s, and that
corresponds to about 14 min.
The behavior of the temperatures for various timings including, of course,
h = 862 s, is shown in Fig. 15.
(2) Let us assume a steel plate with a thickness of 100 mm radiated on both sides;
the goal is to increase its temperature from 20 to 300°C through a heat source
at 800°C; the overall black level is equal to 0.7.
In addition, as usual c = 500 J/kg K; q = 7850 kg/m3; k = 45 W/m K.
By ignoring the difference between surface temperature and mean temperature
of the plate the mean surface temperature from beginning and end of the
process is equal to (20 ? 300)/2 = 160°C.
With tf = 800°C and ts = 160°C through Table 8 we obtain Y = 114.42; thus
a = 0.7 9 114.42 = 80.1 W/m2 K and Bi = 80.1 9 0.1/45 = 0.178.
The heating process is equal to 280/780 = 0.359 = 35.9%.
Based on Table 9: ts/tm = 1.0538 and at the end of the process the surface
temperature of the plate is equal to 1.0538 9 300 = 316°C.
8 Plate Radiated on Both Sides 31
1094 s
280
260
965 s
240
t (°C)
220
764 s
200
180
160
140
427 s
120
0 10 20 30 40 50 60 70 80 90 100
x (mm)
For the plate radiated on one side and assuming no heat waste or a negligible one
through the other side, the same procedure required for the plate radiated on both
sides must be set in place.
Of course, the values of the ratio ts/tm differ from those in Table 9 for radiation
on both sides; they are listed in Table 10.
As expected, the values in Table 10 are higher than those in Table 9, given the
greater difficulty for the heat to penetrate the plate.
Of course, the value of Biot’s number has been computed and the heating
percentage is known, the value of the dimensionless number X can be obtained
through Table 5; after establishing the value of X time h may be computed, as
already seen earlier.
32 Transient Heat Transfer
9.1 Examples
360
340
401 s
320
t (°C)
300
280
323 s
260
240
220 245 s
200
0 3 6 9 12 15 18 21 24 27 30
x (mm)
34 Transient Heat Transfer
The heat transfer from a heating fluid to a heated fluid through a plate can be
accomplished in a variety of different ways.
Considering transient phenomena which are of interest in this case, it is
therefore impossible to elaborate a general purpose computation method to
identify time h which is required to obtain a certain result.
Nonetheless, after setting certain conditions it is possible to do interesting
research to be, as shall be outlined next. Subsequently, we will introduce a few
criteria to take into account to use the results of the research even in situations
beyond the scheme at the basis of the research itself.
Let us consider a plate at an initial conventional temperature of 20°C.
The plate is licked on the sides by two fluids at different temperature, and in any
case at a temperature higher than 20°C.
Initially, both fluids transfer heat to the plate, and the latter is heated up; but
when the temperature of the plate surface in contact with the cooler fluid reaches
10 Transient Heat Transfer Through a Plane Wall 35
440
400
360
t (°C)
280
240
200 2644 s
160
2060 s
120
1490 s
80
950 s
40
0
0 8 16 24 32 40 48 56 64 72 80
x (mm)
the temperature of the fluid, the direction of the thermal flux is reversed, and now it
is the plate transferring heat to the fluid at the expense of the heat it receives from
the warmer fluid.
This leads to an almost steady condition where the distribution of the tem-
peratures through the plate is stabilized and all the heat transferred from the
warmer fluid to the plate is transferred to the cooler one.
We investigated how much time is required for 95% of the heat transferred
from the warmer fluid to the plate to be transferred to the cooler one; we do not
have steady condition but we are very close, i.e., only 5% of the heat transferred
from the warmer fluid to the plate is still required to complete the heating process
of the plate and the stabilization of temperatures.
For research to deliver simple criteria for the computation of h it is necessary to
adopt the same heat transfer coefficients a of both fluids to have just one number of
Biot.
By adopting such conditions we establish that the dimensionless number X,
used as usual to compute h, essentially depends on the number of Biot as well as
on the ratio between the temperatures of the two fluids indicated by t0 for the
heating fluid and by t00 for the heated fluid.
Thus, it was possible to create Table 11 where X is a function of Bi and of t0 /t00 .
In fact, with equal values of the ratio t0 /t00 , the temperature of the heating fluid
(and consequently that of the heated fluid) has a certain impact on the value of X,
even though this impact is modest; for instance, as far as temperature t0 , going
from 300 to 1000°C the increase of the value of X varies between 1 and 2%, thus
making it negligible.
36 Transient Heat Transfer
Table 11 Dimensionless number X for heat transfer through a plane wall with a0 = a00 (warmer
fluid transfers 95% of heat to colder fluid)
Bi t0 /t00
1.2 1.3 1.4 1.6 1.8 2.0 2.5 3 4 5
0.010 2.859 2.688 2.587 2.431 2.329 2.256 2.135 2.061 1.973 1.922
0.015 2.879 2.707 2.590 2.435 2.333 2.259 2.138 2.063 1.976 1.925
0.020 2.883 2.711 2.593 2.437 2.334 2.301 2.177 2.102 2.012 1.961
0.025 2.903 2.760 2.641 2.482 2.379 2.303 2.180 2.105 2.015 1.963
0.030 2.941 2.764 2.645 2.486 2.381 2.306 2.183 2.107 2.018 1.966
0.040 2.947 2.771 2.650 2.513 2.407 2.331 2.207 2.130 2.040 1.987
0.050 2.978 2.800 2.680 2.519 2.413 2.337 2.212 2.136 2.055 2.003
0.060 2.985 2.807 2.686 2.525 2.432 2.354 2.229 2.152 2.061 2.008
0.070 2.992 2.828 2.705 2.544 2.437 2.360 2.234 2.157 2.066 2.013
0.080 3.014 2.834 2.712 2.550 2.442 2.366 2.240 2.163 2.071 2.018
0.10 3.027 2.847 2.724 2.561 2.454 2.377 2.251 2.173 2.081 2.028
0.12 3.041 2.860 2.737 2.573 2.466 2.388 2.261 2.184 2.091 2.038
0.15 3.061 2.879 2.755 2.590 2.483 2.405 2.278 2.200 2.107 2.053
0.20 3.095 2.910 2.786 2.620 2.511 2.433 2.304 2.225 2.131 2.077
0.25 3.128 2.943 2.817 2.649 2.540 2.461 2.331 2.251 2.156 2.102
0.30 3.161 2.975 2.847 2.678 2.569 2.488 2.358 2.278 2.183 2.127
0.40 3.227 3.038 2.909 2.738 2.625 2.544 2.411 2.329 2.232 2.176
0.50 3.294 3.102 2.970 2.795 2.680 2.598 2.463 2.380 2.282 2.225
0.60 3.360 3.165 3.031 2.854 2.738 2.654 2.516 2.432 2.331 2.273
0.80 3.492 3.290 3.153 2.970 2.849 2.763 2.620 2.534 2.430 2.371
1.0 3.625 3.416 3.274 3.084 2.962 2.871 2.725 2.635 2.529 2.467
1.2 3.756 3.541 3.395 3.200 3.073 2.980 2.830 2.736 2.627 2.563
1.4 3.889 3.665 3.515 3.315 3.183 3.087 2.933 2.837 2.725 2.659
1.7 4.087 3.855 3.696 3.487 3.349 3.250 3.089 2.988 2.869 2.802
2.0 4.285 4.042 3.878 3.659 3.516 3.412 3.243 3.138 3.016 2.944
Please, note that the values of X shown in Table 11 were obtained by adopting
t0 = 1000°C, and are therefore conservative.
In addition, the condition that the heat transfer coefficients of both fluids be
the same as the basis of the research leading to Table 11 is actually out or the
ordinary.
A more general analysis of this phenomenon must include the reasonable
assumption that the heat transfer coefficients of the two fluid are different; thus, it
is necessary to analyze how this condition impacts the value of X.
To that extent we start by adopting the heat transfer coefficient as a reference
for the computation of the number of Biot, to be considered as far as Table 11, the
average between the heat transfer coefficients of the two fluids.
The value of X obtained this way through Table 11 is indicated by X*;
At this point we introduce a corrective factor K so that the actual value of X is
given by
X ¼ KX : ð42Þ
10 Transient Heat Transfer Through a Plane Wall 37
Now we must compute the values of K; these values combined with the values
of X* listed in Table 11, make it possible to compute the actual values of X.
By indicating the heat transfer coefficient of the heating fluid with a0 and the
heat transfer coefficient of the heated fluid with a00 , two distinctions are required;
the first concerns the presence of condition a0 [ a00 and the second the presence of
condition a00 [ a0 .
In the first case we obtain Table 12 and we establish the following: the values
of K are always greater than one; they increase with an increase of a0 /a00 , of Bi and
of t0 /t00 ; this leads to an increase of X which amounts to a few percentage points if
the value of a0 /a00 is not considerably greater than one but may reach or exceed
50% in certain extreme cases.
In the second case we obtain Table 13 and we establish the following: the
values of K increase with an increase of a00 /a0 and of Bi but except for a few
isolated instances they decrease with an increase of t0 /t00 ; in some cases the value of
K is even lower than one; the values of K are always lower than those in Table 12.
Tables 12 and 13, as well as Table 11, make it possible to compute the value of
X even when the heat transfer coefficients of the two fluids are different, as we
shall see in the examples.
In conclusion, we would like to point out that the assumption was an initial
temperature equal to 20°C; if it is higher instead, the value of X naturally
decreases; generally this is a decrease of a few percentage points and may reach
38 Transient Heat Transfer
10% if the initial temperature of the plate is equal to half the temperature of the
heated fluid.
We shall not focus on this phenomenon because the number of alternatives
would simply be infinite, and also because if the phenomenon is present and we
ignore it we are being conservative by calculating a slightly greater time than the
actual one.
10.1 Examples
The ratio between the heat transfer coefficients is equal to 6 with a0 [ a00 ; therefore,
we must examine Table 12.
Based on the values of a0 /a00 , of Bi and t0 /t00 we may assume that K = 1.2; then
X = 1.2 9 2.36 = 2.832.
Thus, time is equal to h = 2.832 9 500 9 7850 9 0.03/105 = 3175 s, corre-
sponding to about 53 min, or almost 1 h.
Figure 19 shows the heat transferred from the two fluids to the plate (Q0 and Q00 ) as
a function of time, as well as the total heat stored by the plate (Qt) assuming
h0 = 3175 s, the ratio h/h0 is shown on the abscissa. Note that heat Q00 already
becomes negative for values of h slightly higher than 0.1 h0 which means that after
about 350 s the plate reached the temperature of the cooler fluid, i.e., equal to
200°C; from that moment on it the plate transfers heat to the fluid. Heat Q0
decreases considerably in value as the temperature of the plate increases; the heat
Qt stored by the plate decreases considerably, as well, and is reduced to 5% of the
heat transferred by the warmer fluid after 3175 s.
Direct computation with the finite differences program leads to h = 3169 s.
Clearly this value practically coincides with the one computed with the suggested
method; the small difference depends on the value assumed for the corrective
factor K, which is not exact.
32
28
24
(kW/m )
2
20
16
12
-4
-8
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
θ/θ
o
40 Transient Heat Transfer
If we now assume that the initial temperature of the plate to be equal to 60°C
instead of 20°C, the direct computation leads to h = 3100 s with a decrease of the
required time with respect to what was observed earlier to be 2.2%.
(2) Let us assume a steel plate with a thickness of 20 mm, licked on one side by a
fluid at 1000°C and on the other side by a fluid at 400°C; the heat transfer
coefficient of the fluid at 1000°C is assumed to be equal to 100 W/m2 K
whereas the heat transfer coefficient of the fluid at 400°C to be equal to
900 W/m2 K; the mean heat transfer coefficient is 500 W/m2 K; the values of
c, q and k are the usual ones for steel.
The number of Biot is therefore equal to 500 9 0.02/45 = 0.2222.
The ratio between the temperatures of the fluids is equal to 1000/400 = 2.5.
From Table 11 we obtain X* = 2.316.
The ratio between the heat transfer coefficients is 9 and is a00 [ a0 , so Table 13
must be used; based on it we may assume that K = 1.19, and that leads to
X = 1.19 9 2.316 = 2.756.
Then h = 2.756 9 500 9 7850 9 0.02/500 = 433 s, corresponding to about
7 min.
After 7 min 95% of the heat transferred by the warmer fluid to the wall is
transferred to the cooler fluid.
Figure 20 shows the same quantities of the previous example.
175
150
125
(kW/m )
2
100
75
50
25
-25
-50
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
θ/θ
o
10 Transient Heat Transfer Through a Plane Wall 41
Note that the value of Q0 decreases over time due to the increase in temperature
of the plate, even though the reduction is small given that the difference in
temperature between fluid and wall remains high. The value of Q00 , which is
initially very high due to the high value of the heat transfer coefficient,
decreases abruptly as the temperature of the plate gets close to the one of the
fluid; then the direction of the heat flux is reversed, and after about 0.4h0 the
heat moves from the plate to the fluid. The initially extremely high heat Qt
stored by the plate abruptly decreases to be reduced after 433 s to 5% of the
heat transferred by the warmer fluid.
Direct computation through the finite differences program leads to h = 431 s.
Clearly even in this case this value practically coincides with the one computed
with the suggested method.
If we now adopt an initial temperature of the plate equal to 100°C instead of
20°C, the direct computation leads to h = 414 s with a reduction of 4% with
respect to the value of 431 s computed before.
The typical case to be analyzed given its commonality is heat transfer through the
tube wall.
Clearly, the values of the dimensionless number X relative to the tube will differ
from those of the plane wall; as we shall see, such differences are not great in the
case where the heat transfer coefficients have the same value for the two fluids, but
they can be if these coefficients are different for the two fluids instead.
In order to allow the computation of X even for the tube we looked at the values
obtained by considering that the warmer fluid to be located outside the tube,
whereas the cooler one flows inside it: this is not the only possible condition but it
is the most likely.
In addition, out of infinite conditions we examined the one where the ratio
between the thickness of the tube and its outside diameter is equal to 0.2, i.e.,
xw/Do = 0.2; this may be considered restrictive but it is possible to establish that if
this ratio ranges from 0 and 0.2 the value of X can be computed by interpolating
between the value of X relative to the plane wall and the one relative to the tube
with xw/Do = 0.2.
Table 14 shows the values of X under these conditions and assuming that heat
transfer coefficients of both fluids are the same, i.e., a0 = a00 .
Naturally, as for the plane wall, the values of X make it possible to compute the
required time for 95% of the heat transferred from the warmer fluid to be trans-
ferred to the colder fluid while 5% is stored in the tube.
A comparison of the values in Table 14 with those in Table 11 (plane wall)
shows that the values of X relative to the tube are always higher than those referred
to the plane wall.
42 Transient Heat Transfer
Table 14 Dimensionless factor X for tube with xw/Do = 0.2 and a0 = a00 (warmer fluid transfers
95% of heat to colder fluid)
Bi t0 /t00
1.2 1.3 1.4 1.6 1.8 2.0 2.5 3 4 5
0.010 2.899 2.732 2.619 2.471 2.390 2.322 2.212 2.146 2.069 2.025
0.015 2.917 2.751 2.636 2.489 2.392 2.325 2.214 2.148 2.071 2.028
0.020 2.920 2.754 2.642 2.491 2.397 2.328 2.223 2.189 2.110 2.065
0.025 2.926 2.779 2.690 2.537 2.441 2.371 2.260 2.191 2.112 2.069
0.030 2.979 2.809 2.692 2.542 2.443 2.374 2.262 2.194 2.114 2.071
0.040 2.985 2.814 2.699 2.547 2.471 2.399 2.286 2.217 2.138 2.093
0.050 3.018 2.845 2.730 2.575 2.477 2.404 2.291 2.224 2.143 2.099
0.060 3.024 2.850 2.735 2.580 2.482 2.424 2.309 2.240 2.159 2.114
0.070 3.033 2.859 2.754 2.601 2.499 2.429 2.314 2.245 2.165 2.119
0.080 3.054 2.878 2.760 2.605 2.505 2.433 2.321 2.251 2.170 2.124
0.10 3.066 2.890 2.775 2.618 2.517 2.446 2.330 2.261 2.180 2.133
0.12 3.082 2.906 2.785 2.631 2.530 2.456 2.341 2.272 2.189 2.143
0.15 3.100 2.923 2.805 2.647 2.545 2.475 2.358 2.288 2.206 2.158
0.20 3.135 2.956 2.836 2.676 2.575 2.502 2.383 2.313 2.232 2.184
0.25 3.169 2.989 2.865 2.708 2.605 2.530 2.412 2.339 2.258 2.210
0.30 3.202 3.021 2.898 2.738 2.632 2.557 2.437 2.367 2.281 2.235
0.40 3.270 3.084 2.959 2.795 2.689 2.614 2.492 2.419 2.332 2.285
0.50 3.336 3.149 3.021 2.855 2.745 2.670 2.545 2.470 2.383 2.333
0.60 3.403 3.212 3.084 2.914 2.803 2.724 2.597 2.522 2.435 2.382
0.80 3.538 3.339 3.207 3.030 2.917 2.836 2.706 2.626 2.535 2.482
1.00 3.670 3.467 3.329 3.146 3.030 2.946 2.809 2.728 2.635 2.580
If the heat transfer coefficients of the two fluids are different, the mean arith-
metic value of the two values must be used to compute the number of Biot; a
corrective factor K must be introduced by using (42), as was already done for the
plane wall.
In the present case X* is the value of X obtained through Table 14 while the
value of K is obtained through Tables 15 and 16; Table 15 refers to the case where
a0 [ a00 , whereas Table 16 is to be used when a00 [ a0 .
As you see, the values of K can be considerably higher than one and demon-
strate the impact of the ratio between the two heat transfer coefficients on the value
of X and consequently on time h.
A comparison of the values of K in Table 15 with those in Table 12 (a0 [ a00 )
establishes that the values relative to the tube are always lower than those relative
to the plane wall; viceversa, a comparison of the values of K in Table 16 with
those in Table 13 (a00 [ a0 ) establishes that the values relative to the tube are
always higher than those relative to the plane wall.
As already pointed out, if xw =Do 6¼ 0:2 the value of X for the plane wall must
be computed through (42) by using Tables 11, 12 and 13; then still through (42),
the value of X for the tube must be computed with xw/Do = 0.2 by using
Tables 14, 15 and 16.
11 Transient Heat Transfer Through a Curved Wall (Tube) 43
Table 15 Corrective factor K for tube with xw/Do = 0.2 and a0 [ a00
a0 /a00 Bi t0 /t00 a0 /a00 Bi t0 /t00
1.2 1.8 2.5 3.5 5 1.2 1.8 2.5 3.5 5
2 0.01 0.968 0.987 0.999 1.006 1.016 6 0.01 1.000 1.043 1.075 1.093 1.113
0.1 0.974 0.992 1.003 1.009 1.020 0.1 1.018 1.065 1.091 1.109 1.129
0.3 0.983 0.999 1.010 1.015 1.025 0.3 1.053 1.098 1.124 1.141 1.160
0.6 0.995 1.009 1.019 1.023 1.033 0.6 1.099 1.141 1.166 1.182 1.201
1.0 1.007 1.019 1.027 1.030 1.038 1.0 1.148 1.186 1.210 1.224 1.241
3 0.01 0.969 1.000 1.017 1.028 1.043 8 0.01 1.020 1.071 1.106 1.128 1.150
0.1 0.979 1.008 1.025 1.036 1.050 0.1 1.041 1.096 1.127 1.148 1.170
0.3 0.998 1.024 1.040 1.050 1.063 0.3 1.083 1.138 1.168 1.186 1.208
0.6 1.021 1.045 1.059 1.067 1.081 0.6 1.140 1.190 1.219 1.238 1.260
1.0 1.045 1.065 1.078 1.085 1.097 1.0 1.201 1.247 1.275 1.292 1.312
4 0.01 0.978 1.016 1.037 1.052 1.069 10 0.01 1.034 1.094 1.128 1.157 1.180
0.1 0.991 1.028 1.049 1.062 1.080 0.1 1.063 1.123 1.157 1.179 1.203
0.3 1.017 1.051 1.071 1.084 1.100 0.3 1.109 1.170 1.204 1.225 1.249
0.6 1.050 1.081 1.099 1.110 1.125 0.6 1.173 1.231 1.264 1.284 1.309
1.0 1.083 1.110 1.127 1.137 1.150 1.0 1.243 1.297 1.328 1.348 1.371
5 0.01 0.989 1.027 1.057 1.073 1.092 12 0.01 1.047 1.116 1.151 1.175 1.207
0.1 1.005 1.047 1.071 1.087 1.106 0.1 1.081 1.147 1.183 1.207 1.232
0.3 1.035 1.076 1.100 1.114 1.131 0.3 1.132 1.198 1.234 1.257 1.281
0.6 1.076 1.113 1.134 1.148 1.166 0.6 1.201 1.265 1.300 1.323 1.349
1.0 1.118 1.150 1.170 1.183 1.199 1.0 1.279 1.339 1.374 1.394 1.420
Table 16 Corrective factor K for tubes with xw/Do = 0.2 and a00 [ a0
a00 /a0 Bi t0 /t00 a00 /a0 Bi t0 /t00
1.2 1.8 2.5 3.5 5 1.2 1.8 2.5 3.5 5
2 0.01 1.068 1.054 1.046 1.030 1.022 6 0.01 1.261 1.234 1.199 1.161 1.130
0.1 1.079 1.061 1.048 1.033 1.026 0.1 1.286 1.255 1.220 1.180 1.146
0.3 1.080 1.064 1.052 1.038 1.031 0.3 1.313 1.281 1.249 1.208 1.176
0.6 1.080 1.066 1.056 1.043 1.038 0.6 1.346 1.313 1.281 1.244 1.216
1.0 1.080 1.068 1.060 1.048 1.044 1.0 1.376 1.342 1.313 1.279 1.253
3 0.01 1.131 1.109 1.093 1.068 1.051 8 0.01 1.308 1.284 1.256 1.209 1.173
0.1 1.146 1.120 1.099 1.075 1.059 0.1 1.344 1.319 1.281 1.235 1.193
0.3 1.153 1.130 1.110 1.088 1.072 0.3 1.389 1.355 1.317 1.270 1.230
0.6 1.162 1.140 1.122 1.102 1.090 0.6 1.434 1.398 1.361 1.315 1.280
1.0 1.168 1.148 1.133 1.114 1.104 1.0 1.477 1.440 1.404 1.362 1.328
4 0.01 1.182 1.156 1.129 1.103 1.080 10 0.01 1.356 1.334 1.304 1.252 1.209
0.1 1.201 1.171 1.144 1.113 1.091 0.1 1.396 1.373 1.332 1.281 1.233
0.3 1.215 1.188 1.162 1.132 1.111 0.3 1.447 1.415 1.375 1.322 1.276
0.6 1.232 1.205 1.182 1.154 1.137 0.6 1.504 1.467 1.427 1.376 1.333
1.0 1.246 1.221 1.200 1.176 1.160 1.0 1.558 1.519 1.479 1.430 1.390
5 0.01 1.225 1.198 1.165 1.136 1.105 12 0.01 1.395 1.376 1.345 1.291 1.237
0.1 1.247 1.216 1.185 1.149 1.120 0.1 1.438 1.418 1.375 1.322 1.269
0.3 1.269 1.237 1.207 1.173 1.146 0.3 1.497 1.466 1.424 1.367 1.315
0.6 1.293 1.262 1.235 1.202 1.179 0.6 1.561 1.526 1.483 1.427 1.378
1.0 1.315 1.285 1.260 1.231 1.210 1.0 1.627 1.587 1.542 1.488 1.442
44 Transient Heat Transfer
11.1 Examples
6
(kW/m)
-1
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
θ/θ
o
11 Transient Heat Transfer Through a Curved Wall (Tube) 45
12
10
8
(kW/m)
-2
-4
-6
-8
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
θ/θ
o
46 Transient Heat Transfer