03 - Continuously Variable Tractor Transmissions

This is not a peer-reviewed article.
Karl Th. Renius and Rainer Resch. 2005. Continuously Variable Tractor Transmissions.
ASAE Distinguished Lecture No. 29, pp. 1-37.
Agricultural Equipment Technology Conference, 14-16 February 2005, Louisville, Kentucky, USA.
Copyright 2005 American Society of Agricultural Engineers.
ASAE Publication Number 913C0305.
CONTINUOUSLY VARIABLE
TRACTOR TRANSMISSIONS
Karl Th. Renius
Prof. Dr.-Ing. Dr. h.c.

Retired Head of Department of Agricultural Machinery
Technical University of Munich (TUM)
Munich-Garching, Germany
Rainer Resch
Dr.-Ing.
Technical Manager
Imagine Engineering
Munich, Germany
For presentation at the 2005 Agricultural Equipment Technology Conference

Louisville, Kentucky, USA
14-16 February 2005
Published by
ASAE – the Society for engineering in agricultural, food, and biological systems
2950 Niles Road, St. Joseph, MI 49085-9659 USA
Lecture Series has been developed by the Power and Machinery Division Tractor Committee (PM-47)
of ASAE to provide in-depth design resource information for engineers in the agricultural industry.
Topics shall be related to the power plant, power train, hydraulic system, and chassis components such
as operator environment, tires, and electrical equipment for agricultural or industrial tractors or
self-propelled agricultural equipment.
ASAE is grateful to Deere & Co for sponsoring the ASAE Distinguished Lecture Series.
Table of Contents
INTRODUCTION AND HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

The tractor transmission system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
“CVT” and other used abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
History I: The early phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
History II: The intermediate phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
History III: The modern phase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Global demands: technology levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
MARKET DEMANDS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Tractor families . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Basic drive transmission requirements, tractor speeds . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Basic PTO requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Closed loop control and drive line management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Comfort, safety, service . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Transmission life . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Load and speed spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Relation of load spectra to speed ranges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
PRINCIPLES OF CVTs AND THEIR INTEGRATION . . . . . . . . . . . . . . . . . 14
Survey on physical CVT principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Mechanical CVTs and their integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Hydrostatic CVTs and their integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
PRINCIPLES OF POWER SPLIT CVTs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Basic functions of power split CVTs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Efficiency versus speed ratio range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Internal and external power split configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Power split CVTs with input- and output-coupled planetaries . . . . . . . . . . . . . . . . . . . . . . 24
Conclusions for practical applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
EXAMPLES OF POWER SPLIT CVTs FOR TRACTORS . . . . . . . . . . . . . . 27
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
The Claas-Jarchow “HM-8” and “HM-II” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
The Fendt “Vario” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
The Steyr “S-Matic” for Steyr, Case, New Holland and others . . . . . . . . . . . . . . . . . . . . . 29
The ZF “Eccom” for Deutz-Fahr, John Deere, Claas and others . . . . . . . . . . . . . . . . . . . . 31
John Deere “Autopowr ” (IVT-USA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
EXAMPLE OF A POWER SPLIT CVT FOR A RICE TRANSPLANTER . . 33
CONCLUSIONS AND OUTLOOK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
CONTINUOUSLY VARIABLE TRACTOR TRANSMISSIONS
Karl Th. Renius
Prof. Dr.-Ing. Dr. h.c., Retired Head of Department of Agricultural Machinery

Technical University of Munich (TUM), Munich-Garching, Germany
Rainer Resch
Dr.-Ing.
Technical Manager, Imagine Engineering, Munich, Germany
Abstract. The paper starts with a longer analysis of the colorful CVT history including typical early pro-
blems. Market demands for tractor transmission systems with CVTs address gross traction, speeds, effi-
ciencies, typical control strategies, expected load spectra and others. Physical CVT principles are
described in a survey, focusing on tractor drives but considering also the automobile sector. Hydrodyna-
mic torque converters have no chances for tractors, while mechanical, hydrostatic and electric CVTs
have. A single unit is however not sufficient, a system is needed. A combination with power shifted ranges
can realize the required wide band of tractor speeds. Hydrostatic and electric CVTs (driven by Diesel en-
gines) are in most applications however not able to meet the efficiency targets in a direct mode. The fuel
cell could improve this situation for electric drives, but is not yet economically available. Mechanical
chain variators offer outstanding efficiencies, but need at least two ranges with a power split for active
zero speed control. The power split technology thus has the potential for adequate CVT transmission sys-
tems. Performances of basic systems are analyzed. Commercially available tractor CVTs with hydrosta-
tic-mechanical power split are described as examples.
Keywords. CVT, Efficiency, Electric CVT, Hydrostatic CVT, HST, IVT, Load Spectra, Mechanical
CVT, Planetary, Power split, Power train, Toroid drive, Tractor, Tractor transmission, Transmission.
INTRODUCTION AND HISTORY have been introduced in Europe 1996 for standard
tractors opening a new era of power train design
Agricultural mechanization is going to continue principles.
with its importance for the world`s nutrition chang-
ing to high sophisticated mechatronic systems in
the high developed countries as addressed by The tractor transmission system
Schueller and Stout (1995). The tractor remains The complete tractor transmission (also called
still the most important machine and its transmis- transaxle) is usually defined as demonstrated by
sion system is a key component representing about Fig. 1: A combination of the vehicle speed change
25-30 % of the total tractor first cost. Stepped trans- gearbox, the rear axle with brakes, the power take
missions have been used since the beginning but in- off (PTO) and - if required - arrangements for the
finitely variable drives with automatic controls front axle drive and for the drive of auxiliary units
ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 5

Flange for auxiliary drive
Gearbox for vehicle ever created the demand of a shiftable gear box.
speed change PTO gear box
Some other very early and at the same time high
sophisticated CVTs should be mentioned. Fig. 2
shows the famous idea of Renault (Renault 1907) to
Input shaft use the principle of internal hydrostatic-mechanical
power split for a passenger car gearbox. The input
Front drive shaft PTO shaft shaft (left) from the IC engine is driving the cylin-
Final drive, brakes and rear axle der block of the left axial piston pump, while the
Fig. 1 Tractor transmission system, side view right cylinder block is rigidly fixed with the chassis.
(Renius, 1999). The power thus is transferred by both, the hydro-
static path and the housing. The higher the vehicle
(mainly hydraulic pumps). The functions can also speed, the lower the hydrostatic portion and the
be addressed by a power flow diagram as used by E. higher the efficiency. The pump is controlled man-
Paul Browning in his well known and still very use- ually while the hydrostatic motor is automatically
ful ASAE Lecture No. 4 (Browning 1978). controlled by the system pressure. Realization of
this outstanding concept failed because of problems
with the required tolerances, materials, skills and
“CVT” and other used abbreviations production costs (Molly 1966). Early developments
CVT is an abbreviation for “Continuously Variable of hydrostatic drives for vehicles in the US have
Transmission”. It has initially often been used to been addressed by Mortenson (1960).
address continuously variable variators for car At the same time agricultural engineers in-
transmissions with steel belt friction drives. vented a friction drive CVT for a self propelled
IVT is an abbreviation for “Infinitely Variable plough of the German company Stock, Fig. 3. The
Transmission”. It is used like the proposed general vehicle speed is controlled by the radii of the fric-
meaning of “CVT”, expresses however little better tion contacts. This vehicle had two main traction
the infinite ratio of nominal speeds down to an “ac- wheels carrying almost the whole vehicle weight
tive” zero speed. and vertical ploughing force. The CVT was in this
case also used to offer a power steering with a sim-
HST is an abbreviation for a (direct) Hydro-
ple steering mechanism, which enabled different
static Transmission (without power split), often
radii for both friction contacts. The complete ma-
used in Japanese publications.
chine is on display at the “Deutsches Museum”
HMT stands for “Hydro-Mechanical Transmis- (German Museum of Technology), Munich. This
sion”. It addresses the combination of hydrostatic invention has not been realized, probably because
and mechanical power transfer, in particular for of the poor power density and missing durability of
power split concepts. the friction drive.
Modern hydrostatic vehicle drives with and
History I: The early phase
CVTs are the dream of every progres-
sive vehicle transmission engineer
since a long time. About 100 years ago,
battery-driven electrical drives al-
lowed already an infinitely variable
speed control and low noise levels. The
main problem was the poor capacity
for stored energy. This problem could
be solved by the introduction of inter-
nal combustion engines, which how- Fig. 2 Passenger car CVT with internal hydrostatic power
split (Renault 1907).
6 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

History II: The intermediate phase
The so called “swash plate axial piston unit”
was first developed in the US since 1942 by
Friction drive CVT Denison and could also gain commercial im-
(Steel - solid rubber) portance in the first stage mainly in the US.
The two mentioned axial piston units
Steering came in competition in the US from about
system 1950. In Europe this happened much later
CVT control from about 1970, by Sundstrand, Mannes-
mann Rexroth, Linde and others (Schunder
1995). The friction between piston and cylin-
der allowed in the first stage tilting angles
only around 18 degrees (bent axis unit much
higher) and limited applications for motor
use in case of low rpm. Therefore early re-
Fig. 3 Early design of a friction type CVT prototype of a search had been carried out investigating the
Stock motorplow 1907 (Renius 1974a). friction mechanics and possible improve-
without power split mostly use so called “axial pis- ments (Renius 1974b).
ton units” for both, pump and motor, Fig. 4. The Although the bent axis principle has little higher
efficiencies and better start up behavior in motor
applications, the strong competition between both
was decided more and more in favor of the swash
plate unit due to the lower specific production
costs, the “drive through shaft”, the compact design
and other advantages.
The first broad research on hydrostatic CVTs
for tractors had been carried out in the former Na-
Fig. 4 Bent axis axial piston unit (left) and swash tional Institute of Agricultural Engineering (NIAE)
plate axial piston unit (right). Silsoe, UK, developing a hydrostatic driven proto-
type (Hamblin 1952), which was presented 1954,
first concept, which gained importance, was the so Fig. 5. The variable displacement axial piston
called “Thoma” axial piston unit. Hans Thoma pump has been driven by the IC engine working di-
(1887-1973) is called to be the father of the today so
called “bent axis axial piston unit”. This promising Variable displacement axial piston pump
Frame with pipes
development was first produced in series by Radial piston motor
Vickers, USA, for aircraft hydraulics according to a
license contract 1935-1941 (Schunder 1995).
Pittler, Demag and other companies started in
Germany with license productions from 1937.
Around 1940, series production was mainly used
for military applications such as aircrafts, tanks and
warships with peak pressures up to 250 bar (about
3600 psi). After World War II the concept was im-
proved and introduced in series production for civil
use by several companies. Improvements came also
from ideas of H. Molly at this time (Schunder Fig. 5 Silsoe research tractor with hydrostatic
CVT, presented 1954 by the National Institute of
1995).
Agricultural Engineering (NIAE), Silsoe, UK.

rectly by tubes on large radial Hydrostatic CVT with swash plate Range
piston motors being inte- axial piston units in a shift
grated with the driving back-to-back arrangement H L
wheels (NN 1954). Several
tractor companies of the US
have also developed proto-
types at this time or little later Double pump
for charging PTO
(Meile 1961). One of the first CVT circuit
commercially available vehi- and operating
cles with a hydrostatic drive control cylin- Pump for main
was the Linde “hydrocar” ders tractor hydraulics
(Hannover fair 1955), which
became the basis of the today
well known Linde forklifters
with hydrostatic Linde trans-
missions. Fig. 6 First in series produced hydrostatic drive system for tractors, In-
ternational Harvester , USA 1967 (Renius 1969).
A first commercial intro-
duction of a hydrostatic CVT for tractors was tried with Sundstrand, USA. The transmission system
by the German tractor company “Eicher”, launch- works with two ranges, Fig. 6. This leads to a better
ing the new “HR” tractor with the hydrostatic overall efficiency for small and high tractor speeds.
Dowty “Taurodyne” unit in 1965. Two bent axis The system had been applied for several tractor
units with variable displacement had been linked types of IHC but the production was finally termi-
together using the connecting oil pipes as nated after “some 10.000 units produced” (personal
four-bar-links, enabling automatically a suitable estimation). The design was in the opinion of many
combination of tilting angles with low pressure experts well done, Fig. 7. But the used hydrostatic
losses (Worn and Walker 1966). The transmission swash plate units created high losses and high first
system was working without ranges. Eicher could costs in comparison with the relatively simple con-
produce and sell only a small number because of ventional stepped transmissions of the 1960s. Fig. 8
several reasons. At the same time the German com- gives an impression of the full load efficiency of the
pany Kopat tried to introduce a hydrostatic CVT complete tractor transmission system according to
with internal power split,
Variable displacement pump Variable displacement motor
system “Allgaier”, for trac-
Valve block
tors and construction ma- Charge pump
chinery – finally also without
Output shaft
success.
Input
More important became
shaft
the decision of International
Harvester, to start 1967
series production of agricul-
tural tractors with the 656 PTO
(engine power about 60 kW )
using two swash plate units
“back-to-back”. Both were
variable in displacement and Charge
used pressure supported con- pipe
trol (Morris 1967, Asmus
and Borghoff 1968). The de- Control cylinders
velopment had been carried Fig. 7 Design of the hydrostatic section of the IHC transmission system
out in a close co-operation plotted in Fig. 6. Swash plate units for pump and motor (Morris 1967).

100
synchronous speeds. The promising concept was
% not commercially applied to tractors. At the same
Hydrostatic unit time Deutz tried the introduction of a hydrostatic
80 drive for its new tractor concept “INTRAC”
(Breuer 1974), but the project was also terminated
Efficiency
70 after a low number of units produced. Other compa-

60 nies such as Versatile worked on intelligent power
hhydr × hrange L × haxle hhydr × haxle split systems for tractors, finally also without suc-
50 cess. A flood of patents were crated regarding hy-
drostatic power split systems, but no system could
40 achieve a series production for tractors, nor for
other vehicles. One reason may be, that the compet-
30
0 6 12 18 24 km/h 32 ing conventional transmissions had been relatively
Vehicle speed cheap and efficient at that time. Secondly comfort
and automation was not yet a strong demand in the
Fig. 8 Full load efficiencies estimated for the 1970s and electronics had not been available.
transmission system of Fig. 6 (Renius 2003).
Hydrostatic drives could however gain impor-
tance for very small tractors (Witt 1965, Moyer
data from Sundstrand (Martin 1969) and own cal-
1966), becoming popular mainly in Japan and USA
culations. The best efficiency of the hydrostatic
and offering a high comfort level. The costs could
unit is 80.5 %, but the best total system efficiency is
be balanced by “simplified” components, improved
only 73 and 75 % because of the additional gear
production methods and high production volumes.
wheels including the final drive. These figures are
Energy losses and fuel economy had been consid-
still typical for similar structures.
ered to be less important in this market.
Best values for conventional transmissions are
however above 90 % for the gearbox and about
85-87 % for the total transmission system (Reiter History III: The modern phase
1990). It was well known at that time, that the en- The 1988 presented “Munich Research Tractor”
ergy losses could be reduced by the power split (Renius 1988, Kirste 1989) marked the beginning
principle (Jarchow 1964, Molly 1966, Kress 1968 of a series of new developments. Its transmission
and others). One important approach of this kind contains a mechanical pull type chain CVT, Fig. 9,
was the Cummins Sund-
strand “Responder” (Ross Speeds (log.):
1,9 9,5 km/h L
1972) with “dual mode”: The
L
combination of a first range 5 25 R H
with direct hydrostatic drive H
2,4 11,9 H
for low speeds with a second R
range working with power 3000 rpm
split. The project arrived 30 or 33 kW
however only at a pilot pro-
duction (Wadman 1973). An
T
other interesting power split L R
system was proposed 1973
by the Orshansky Transmis-
540
sion Corp., USA (Orshansky rpm
1972). This system used a
four-shaft planetary and of-
fered, for example, 4 power
Fig. 9 Continuously variable transmission of the Munich Research
split ranges with shifts at
Tractor 1988, CVT from P.I.V. (Kirste 1989). T torque sensor.

combined with two forward ranges Table 1 Basic transmission specifications of standard tractors by
and one reverse range. Hydraulically technology levels – word-wide view (Renius 2000, 2002).
controlled clamping forces using the
signal of a torque sensor enabled full Nomin. speeds, km/h No. of speeds PTO
Level
Shift
load efficiencies of the variator unit forward reverse forw./rev. speeds rpm
up to about 90 % including hydro-
static losses. The tractor, which had I 2–20(25) 3–8 6/2 to 8/2 SG, CS 540
been developed with broad support II 2–30 3–10 8/4 to 12/4 CS, SS 540/(1000)
of the German industry, is now on III (0.5)2–30(40) 3–15 12/4 to 16/8 SS, HL 540/1000
display at the German Museum of IV (0.3)2–40(50) 2–20 16/12 to 36/36 SS, PPS,
(or more) FPS 540/1000
Agriculture at University of (750/1250)
Hohenheim. A prototype for a V 0–50(60) 0–25 ¥ Automatic
60 kW tractor developed by Schlüter SG Sliding gear, CS Collar shift, SS Synchro shift, HL HiLo power shift, PPS
together with Hurth and Partial power shift (3 or more speeds), FPS Full power shift, ( ) options
P.I.V.-Reimers around 1990 was
the “Fordson” tractor could achieve in the 1920s a
promising but did not reach series production due to
world-wide market share above 50 % with only one
the decline of Schlüter. Fendt was also very inter-
model, this situation has completely changed be-
ested but the limited torque capacities of chain
cause of the wide-spread agricultural mechaniza-
variators did not allow their application for large
tion demands. The typical tractor in India, China or
tractors. Fendt and CLAAS started therefore the de-
South America has, for example, not many com-
velopment of hydrostatic power-split concepts in
mon components with the typical tractor in Europe,
the late 1980s. The CLAAS transmission with a
USA or Japan. Today`s global demands thus re-
four-shaft planetary was based on Jarchow patents
quire a complicated “multiple model strategy” re-
and could achieve outstanding high efficiencies,
garding both, tractor functions and tractor power.
but was finally produced only in a limited number
Five technology levels have been defined for global
of units. At that time some power split transmis-
product planning of tractors (Renius 2000, 2002) by
sions have been developed also for other mobile
typical component specifications. The transmission
machinery (Mitsuya et. al 1994, Schueller and
being the most complicated component is classified
Khan 1995). The Fendt "Vario" CVT was presented
and specified by Table 1. Transmissions using a
for tractors at Agritechnica 1995 and produced for a
CVT unit can be regarded to represent the highest
first large tractor model (926 Vario) from 1996
technology level V addressing markets with high
(Dziuba and Honzek 1997). The outstanding suc-
productivity and comfort demands.
cess motivated Fendt to apply the principle to more
and more models in the following years. Mean-
while ZF and Steyr followed with their own con- MARKET DEMANDS
cepts also based on hydrostatic units in power-split
arrangements. John Deere used in a first step the Tractor families
ZF-concept, but developed later on its own hydro-
static power split transmission for larger tractors. Table 2 shows a typical structure of a world-wide
The systems of Fendt, Steyr, ZF and John Deere standard tractor program (model). It contains at
will be presented in a later chapter. least 4 families according to 4 basic engine con-
cepts. Every family consists of some models with a
large number of common parts. This reduces the to-
Global demands: technology levels tal number of parts and thus the production and lo-
Not every tractor on the world market needs a CVT. gistics costs (Welschof 1974, Jenkins 1997, Renius
Only a small section of the total volume is and will 2002).
be equipped in the future with such a “high tech” CVT transmission families should follow the
transmission. This shall be demonstrated by the tractor families. But family 1 is very sensitive in
principle of technology levels for tractors. While first costs and thus not yet equipped with CVTs.

Table 2 Production program for standard tractors with (at least) 4 trac- level little above. This very
tor families. Rated engine power for rated speed and ECE R24. Maxi- high load must be considered
mum engine power may be higher. for life calculations with a
limited time portion (see load
Tractor family 1 2 3 4 spectra).
Rated eng. Speed change: Infinitely vari-
30 – 55 60 – 90 90 – 125 135 – 210
power, kW able; if possible, under load
Rated engine for the full speed range. Short
2100 ... 2500 acceleration interruptions
speed, 1/min
Diesel engine 3 cylinder 4 cylinder 6 cylinder 6 cylinder
(range shift) accepted above
Displ. ~ 3l Displ. ~ 4l Displ. ~ 6l Displ. > 7l 8-10 km/h. Ability for auto-
matic control is a “must”.
Turbo charger & intercooler very common
Speed range forward: From
Volume of
moderate very large (options!) large zero to a maximum of 40, 50
functions
or 60 km/h (future versions in
Comfort level moderate high very high Europe).
Transmission efficiency for-
The following market demands are therefore listed
ward: Total drive line efficiency (engine to wheels)
for typical standard tractors above about 50 kW
for full load very near to those of stepped full power
rated engine power.
shift transmissions, see the target in Fig. 11.
Overdrive: Maximum vehicle speed should be pos-
Basic drive transmission requirements, sible at reduced engine speed to save fuel and re-
tractor speeds duce noise.
Basic drive diagram: A typical gross traction ver- Speed range reverse: From zero to about 25 km/h.
sus speed diagram is shown by Fig. 10 for a tractor Shift forward-reverse: Must be possible under load
with CVT and two mechanical ranges. Maximum (power shift).
gross traction on concrete is about equal to the Active zero speed control: Must be possible under
weight of the ballasted tractor resulting in the maxi- all practical load conditions.
mum axle output torques. In case of a hydrostatic
CVT, pressure is usually limited according to a Mechanical front wheel drive: An output shaft must
be available, integrated design with ability for auto-
matic control.
Gross traction of all driving wheels
Full load driveline efficiency
Tractor pulling with 90

(axle / gearbox input)
full ballast on concrete %

(e.g. Nebraska Test) 85 84
Tractor pulling off-road 82

80 Full load drive line efficiency 80
Without losses at rated engine speed should
With losses lie above the borderline
75
Range L 0
3 4 5 6 1012 20 km/h 40
Vehicle speed (at rated engine speed)
Range H
0 Fig. 11 Efficiency target for CVTs of larger trac-
Vehicle speed
tors above 100 kW (Renius 1993, 1994, 1999).
Fig. 10 Gross traction versus vehicle speed for a
CVT with two ranges L and H.

Basic PTO requirements Table 3 Manual and automatic control strategies
for practical operations of tractors with CVT.
Rear power take off for implements (PTO): Specifi-
cation according to relevant standards, but up to g Manual transmission control
four shaft speeds. If possible, CVT also for PTO. g Closed loop speed control for PTO
Other power take off functions: The PTO “live g Combined open loop control of engine
shaft” should be able to drive hydraulic pump(s)
and transmission (transports)
and other auxiliary components.
g Closed loop power control for heavy
PTO shift: Should be able for automatic control. operations (soil cultivation, combined
PTO efficiency (engine to shaft under full load, implements, transports)
tractor not moving): About 95 % (auxiliary drives g Closed loop cruise control with slip
such as oil pumps and idling elements of the drive compensation (distribution of chemicals
transmission excluded). and seeds)
g Fuel economy oriented combined open
Closed loop control and drive line manage- loop control of engine and transmission
ment (all partial load operations)
g Closed loop control of the total working
Closed loop controls are used for tractor-implement process of tractor and implements
systems because of
– increased performance and thus improved for system configuration should widely be self ex-
productivity, plaining for the driver. This is a very important
– better working accuracy and thus higher point as most drivers of CVT tractors quote the
working quality comfort to be the most important advantage. In case
– reduced energy input of a break down of the automated control system
– improved driver comfort, health and safety emerging systems must be available (Martinus and
– contribution to environment protection. Freimann 2002). The typical stepwise development
of safe mechatronic systems can be demonstrated
These advantages are closely connected with the re-
by the so-called V-model, Fig. 12. Initially it had
quirements of precision farming methods. Closed
been a development standard for government IT
loop controls are working considerably better with
systems (N. N. 2003) but is now going to become a
infinitely variable actuators than with graduated ac-
general scheme for the development of mechatronic
tuators. Typical control strategies for tractor trans-
systems such as tractor and tractor-implement sys-
missions are listed by Table 3. Electronic engine
tems, including embedded software.
control simplifies a complete drive line manage-
ment (Scarlet 1993, Harms and Jaufmann 1996). It Test Documentation
may be mentioned, that fuel saving strategies for Specification of Specification Validation/
overall system System test
tractors with CVTs have been already proposed in
1964 (Howson 1964).
Specification of Integration
Electronic control strategies and related design subsystems Tests
directives must consider the new standard
ISO 11 783 (Goering et al 2003).
Module design Module testing
Comfort, safety, service
Comfort and safety Software-implementation,
-coding, -test
Applications of electronics for semi-automatic or
fully automatic control strategies require skillful Fig. 12 V-model for the development process of
solutions for the man-machine-interface. Settings mechatronic systems.

The process starts with the definition of the basic Load and speed spectra
product specifications as outlined before, specify-
ing on this basis the subsystems. Actual implemen- Transmission speed collectives can be picked up
tations of the functional structures take place in the from the local requirements. A generalized estima-
lower part of the V-model while the right hand leg tion for Europe can be found in the CIGR Hand-
indicates all verification activities beginning with book III, page 142 (Renius 1999).
the modules and ending with the complete system. The torque load spectra for the input shaft of
Tools like FMEA (Failure-Mode and -Effects CVTs are stronger than those for conventional
Analysis) or model-based developments of elec- stepped transmissions due to the popular control
tronic systems can be adopted from the automobile strategy “power control for heavy operations”,
sector for tractor mechatronics. which increases the use of the installed power con-
siderably. This is demonstrated by Fig. 13.
Maintenance and repair The torque is counted in time portions and re-
Requirements of service and repair should be as lated to the rated engine torque at rated engine
low as possible (e.g. long oil change intervals or re- speed (not maximum torque!) in order to generalize
placement of filters only “on demand”). All trans- the diagram. It shows as a basis the “Renius” load
mission elements (including clutches and brakes) spectrum, which has been applied in Europe for
should be dimensioned for life. For diagnostic pur- many conventional transmissions as a first ap-
poses an electronic interface of the transmission proach. This curve could be modified (Vahlensieck
system must be on hand. Thereby networked archi- 1999) and was used for CVT developments.
tectures enable central access to system data using In addition, load spectra for the transmission
standardized protocols. output are necessary. Fig. 14 shows a standard load
spectrum, which was published for 2WD (Renius
1976) and later confirmed for 4WD and typical Eu-
Transmission life ropean conditions (e.g. Rempfer 2003). These sta-
Tractor transmissions have to be designed and tistical fundamentals have been considered as a
tested for the total tractor life. Nominal tractor life guideline for calculation and tests of several
should always be defined together with a statistical stepped transmissions and CVT developments in
probability. The authors prefer the B10 definition: Europe. High torque reserves above about 25 %
10 % can fail, 90 % last longer. Tractor B10 life need little stronger load spectra for the transmission
should be about 8.000 to 12.000 hours (real time), input. The experience shows, that the transmission
increasing with the tractor power. input load spectra are mainly determined by the en-
gine characteristics and the transmission control
3.0
Fig. 13 Total torque load spectra for trac-
Transmission input torque load
tor transmission input shafts. Probability

2.5
I: stepped transmissions about 90 %.
Curve I Standard load spectrum “Renius”
2.0 for the input shaft of stepped transmissions
with dry master clutch, moderate torque
1.5 reserve and no automatic power control
II: CVTs (Renius 1976).
1.0 1.0 = rated input torque Curve II Typical deviation for CVT trans-
missions with closed loop power control
and moderate torque reserve (torque
0.5 back-up) (Vahlensieck 1999).
0
0.001 0.01 0.1 1 10 % 100
Cumulative frequency of time portions

2.0
Fig. 14 Total torque load spectra for trac-
Sum of gross tyre traction forces
tor drive line output. Torque calculation
related to tractor net weight
with effective tire radii for normal inflation
1.5
pressures of about 1.2 bar; probability for
European conditions and 4WD about 95 %
(Renius 1976, Rempfer 2003).
1.0
0.5
0
0.001 0.01 0.1 1 2 5 10 % 50 100
Cumulative frequency of time portions
strategies (but scarcely by local conditions) while running this part of the mentioned tests should be
the output load spectra are mainly determined by added to the curve of Fig. 14 creating a certain
the local traction conditions and the tractor hump in the left strong part of the spectrum.
weight. A typical dimensioning of a pair of gear wheels
meeting a load spectrum with the lowest possible
Relation of load spectra to speed ranges dimensions has been presented first in ASAE paper
76-1526 (Renius 1976).
A complete outline would exceed the purpose of
this paper. But a simplified rule may be given. In
principle, the transmission output load spectrum
must be balanced with the input spectrum by con-
PRINCIPLES OF CVTs AND THEIR
sidering the limited power of the engine including
INTEGRATION
dynamic overloads.
This leads for typical weight-to-power ratios of Survey on physical CVT principles
standard tractors to a characteristic vehicle speed of Four typical groups of continuously variable vehi-
about 5 to 7 km/h. Below this speed, the transmis- cle drives can be called to be important, Table 4.
sion loads must be derived from the traction dic- Their physical principles are completely different,
tated load spectrum of Fig. 14. Above this speed, also their ratio controls and typical fields of appli-
transmission loads should be derived from the input cation.
load spectra of Fig. 13. A modification is recom-
mended to increase the curve a little within the typi- The “hydrodynamic torque converter” is
cal working range for heavy duty operations up to achieving by far the highest production volume for
about 12 km/h and to decrease the curve a little for vehicle transmissions and offers in the same time
the resting speed range, which is mainly used for the lowest production cost, but has two major weak
light field operations, transports and tractor move- points regarding the use for tractors:
ments farm-field-farm. 1. The maximum efficiency is not poor but only
Regarding the transmission output load spec- available within a very limited band of ratios.
trum, the special conditions of Nebraska Test and 2. The ratio cannot be controlled in a closed loop,
OECD Test are not included. Extremely high pull- as the ratio results automatically from the load.
ing forces are generated for speeds much below the A solution to diminish weak point 1 is to add a free
mentioned 5 to 7 km/h due to cleaned but well grip- wheeling element or even a clutch blocking the
ping concrete with optimized tire conditions and unit. But this reduces the CVT effect and thus re-
high ballast. Thus the loads during preparing and quires a high number of additional conventional

Table 4 Important physical principles of continuously variable transmissions for vehicle drives.
Principle of energy
Type of CVT Ratio control Application Efficiency
transmission
Important for
Usually
Mass forces at pump pass. cars and
1 Hydrodynamic automatically poor
and turbine construction
by load
machinery
Traction forces within Radius of Important for

2 Mechanical excellent
friction contacts traction force passenger cars
Important for
Hydrostatic forces at Displacement
3 Hydrostatic mobile moderate
pump(s) and motor(s) of the units
machinery
Frequency of
Electro-magnetic
current or
4 Electrical forces at generator(s) Upcoming moderate
electric flux or
and motor(s)
load
ranges. Weak point 2 could be diminished only Modern traction drive CVTs are lubricated by oil.
very little by adjustable vanes of the reaction wheel.The maximum usable friction coefficients
Thus all developments trying an introduction (steel/steel) are about 0.06 to 0.12 and are mainly
for tractors finally failed. As the authors do not seeinfluenced by the type of fluid: high values with
a principal change of this situation for the near fu- special “traction fluid”, low values for example
ture, the torque converter will not be included in thewith rape seed oils. There is also some research on
following considerations. small dry friction drives resulting in higher friction
coefficients.
Those CVTs, which are most considered for
tractor transmissions, are listed little more in detail Best efficiency of the traction contact results
by Table 5. As noted by Table 4, the CVT princi- from a compromise between slip and rolling resis-
ples No. 3 and 4 are not excellent in efficiency. tance losses – similar to traction mechanics of
They are not able to meet the target of Fig. 11, if pneumatic tractor tires. The rated Hertz pressures
they are applied in a direct mode. A power split sys- can arrive at levels of 3000 N/mm2 for toroidal
tem can increase system efficiency over unit effi- CVTs (Machida et al 1995) and to about
ciency. Only mechanical CVT units can offer 800 N/mm2 for chain type friction drives. The high
excellent efficiencies in a direct mode, which is normal forces are usually generated by hydrostatic
their main advantage. pressures.
The following traction type CVTs are in pro-
duction for passenger cars:
Mechanical CVTs and their integration
– Push type chain variators (Van Doorne/Bosch,
used in millions of mostly smaller cars)
Survey
– Pull type chain variators (P.I.V./Luk/Audi, first
The ratio is controlled by variable effective radii, introduced in 2000 for cars)
see, for example, the historical prototypes of Fig. 3 – Toroidal CVTs (NSK Ltd./Nissan, half toroidal
and Fig. 9. version, first introduced for a larger car late
The torque T is in principle a function of normal 1999).
force FN , friction coefficient m and radius r:
T = m × FN × r

Table 5 Relevant CVT principles for tractor drives. Toroidal drive CVTs
pros`s con`s comments

Toroidal drives are classified in “full
Mechanical by High efficiency, No start from Push type CVT: toroidal” and “half toroidal” con-
friction “direct” low noise level zero, Japanese cars cepts, Fig. 15. A “full toroidal” CVT
low shift speeds, Pull type CVT: Audi- has been developed by the British
power limits LuK chain CVT (2000)
(chains) Toroid Nissan (1999)
company Torotrak Ltd., Fig. 16
Mechanical Start from Efficiency can be Prototypes known
(Greenwood and Soar 1991) and a
power split zero, lower than with under the name “half toroidal” CVT by NSK/Nissan
low noise level “direct” “geared neutral” (Machida et al. 1995). Double cavity
Hydrostatic Long Low efficiency Popular for small arrangements are recommended for
“direct” experience, high noise level tractors and crawlers both concepts improving power and
high shuttle
comfort efficiency (Machida et al. 1995).
Hydrostatic High efficiency Noise, CLAAS “HM-8” (1996) The “Torotrac” has been pro-
power split limited flexibility, Fendt “Vario” (1996) posed for various vehicles, also for a
shift can be “Steyr S-Matic” (2000)
complicated, ZF “Eccom 1.5” (2001) tractor transmission with power split
high R&D costs J. Deere/USA (2001) for LG Cable Ltd., South Korea (Kim
Electrical Low noise Expensive, Several research et al. 2000).
“direct” with level, high low efficiency activities on units and
Diesel engine shuttle comfort systems. A typical advantage of toroidal
Electrical Low noise Very expensive, Long term potential, CVTs is the potential for high torque
“direct” with level, high extra motor for commercial use 2022? capacities and compact design due to
fuel cells shuttle comfort PTO, high R&D parallel power flow. The slip is
costs
higher than for chain drive CVTs,
Electrical High efficiency Expensive, high Several research
power split with low noise level R&D costs activities on units and
same for the losses born by drilling
Diesel engine systems. Commercial friction. Half-toroidal concepts are
use 2012? little better in efficiency than full-to-
roidal drives. Efficiency drops at
high loads and high speed reductions.
This is not so important for car transmissions, but is
a weak point for tractors.
Steel belt drive CVTs

Infinitely variable steel belt drive CVTs are classi-
Fig. 15 Half toroidal (left) and full toroidal (right)
traction drive CVT, shapes and reaction forces.
fied in push type and pull type concepts. Both have
been developed over several decades (Schön-
nenbeck and Wagner 2002).
Very first and today famous funda-
mentals for the pull type CVT have been
presented by Dittrich (1953). The push
type version was also subject of many pa-
pers; fundamentals see, for example,
Input Fushimi/Fujii/Kanehara (1996).
For tractor transmissions the pull type
steel belt variator has been preferred in the
Output P.I.V. version with many prototypes
Fig. 16 Traction drive CVT, full toroidal type in double cav- (Fig. 9). Fig. 17 represents a modern de-
ity arrangement. Position: High output speed (courtesy sign. The variator works in principle like
Torotrak Ltd.). an adjustable V-belt drive. There are no

This CVT type could catch a lot of interest due to its
introduction for cars with the “multitronic®” by
Audi/Luk in 2000 (Nowatschin et al. 2000). It is
called to be the worldwide “best in efficiency” au-
tomatic car transmission with some 100.000 units
produced until 2004.
A survey on the fundamentals (Sauer 1996,
PIV chain Srnik 1999) and several other publications confirm
Torque
sensor the outstanding high efficiency potential – probably
Torque the highest of all presently known CVT concepts. If
sensor the clamping forces are properly adjusted to the ac-
tual torque load, mechanical full load efficiencies
Fixed of about 95 % can be achieved within a wide band
pulley of speed ratios (Sue/Dittrich/Poll 2002). Practical
values are lower due to the losses of the hydraulic
Movable pulley/clamping cylinder (20-50 bar) control system. A simple system with a constant
displacement pump doubles the losses, resulting in
Fig. 17 Pull type chain CVT concept PIV (cour- best efficiencies of about 90 % . An improved hy-
tesy PIV Drives). draulic control system with a variable displacement
pump could achieve a measured best efficiency of
sliding losses within the steel chain due to the “roll- 92.5 %, Fig. 18 (Sauer 1995). This map has been
ing contact” joints (instead of sliding joints). The picked up for 1000 rpm, higher input speeds such as
joint pins are in traction contact with the pullies. A 3000 rpm are not reducing efficiencies. High
stiff design is recommended in many publications sophisticated hydraulic control systems have been
to prevent losses by radial slip. developed (Faust/Homm/Reuschel 2002, Mölle
2003).
2,2
0,895
0,79 0,78
2,0
0,9
1,8
0,82 0,81 ,8
0,905
0
0,83
Output/input speed
1,6
0,91
1,4
0,915
1,2
4
0,8
1,0 0,925
2
0,924 0,9
0,86
0,8 0,923
0,922
0,895
0,87
0,89
LD 6290 ã LTM
0,905
0,6
0,88
0,9
0,85
0,9
0,895
0,4
20 40 60 80 100 120 140 Nm 180
Input torque
CVT from P.I.V., Input speed 1000 /min, double-sensor system (8,9 Nm/bar, variable displacement vane pump, oil flow 7 l/min
Fig. 18 Measured efficiency map of a chain CVT (similar to Fig. 17). Variable displacement hydraulic
pump compensates changing input speeds (no flow surplus at high speeds) thus reducing the losses as com-
pared with a constant displacement pump (Sauer 1995).

The mechanics are complicated. It can, for exam- are solved: The enlargement of the transmission
ple, not be tolerated to assume equal clamping output speed ratio forward and a reverse range.
forces at both belt pulleys. The following correc- High tech markets will also require an active (pow-
tion factor z has been introduced by P.I.V. ered) zero speed control.
(Schönnenbeck and Wagner 1992): A simple combination of the CVT unit with a
clamping force driving pulley conventional power shifted range package has the
z=
clamping force driven pulley disadvantage, that the range shifts last long and that
an active zero speed control is not possible.
Practical values depend on ratio and torque load. If
An improved very early approach of solving
very low loads are excluded, z varies between
these problems is plotted in Fig. 19 based on a pat-
about 1.0 and 1.5 influenced by stiffness, output
ent of J.H. Kress/John Deere (1962). The system
speed and load (Sauer 1995). A typical z map has
works in mode L with power split thus realizing ac-
been measured by Westenthanner (2000). Actual
tive zero speed control by “geared neutral”. Modes
information on hydraulic low energy control sys-
R and H are “direct”. There is high relative parasitic
tems with adjusted flow rates can, for example, be
power circulating in mode L, but this is mitigated
picked up from Stöckl (2002). Interesting ap-
by the traction limit. Range shift between L and H is
proaches with speed modulated electro-driven con-
made in “synchronous points” which cuts adjust-
stant displacement pumps have been presented by
ment times of the CVT unit. An application of this
Koberger (2000), Bradley/Frank (2002) and others.
interesting idea for series production did not take
The live of steel belt CVTs is mainly dictated by place until now, but the principle of combining a
chain fatigue, only in the second order by surface power split range with one or more direct ranges is
wear and pittings. Internal chain dynamics have again under investigation now (Resch 2004), even
been measured by Sauer (1995) and confirmed by a for passenger car CVTs such as those of Audi-Luk
precise model of Srnik (1999). Fatigue modeling (Englisch et al. 2002).
with external load spectra and internal dynamics
Another approach of increasing the output
was first published by Vahlensieck (1999).
speed ratio of systems with chain variators is the
Friction drive CVTs can only be used for future so-called i2-principle, proposed for passenger car
tractor drive transmissions, if two basic problems transmissions. The variator is used in two ranges
with alternative opposite
power flow direction and
n1 synchronous shift between
both (Höhn/Pinnekamp
Power splitting 1994). Four shift elements
PTO are necessary. Prototypes at
HIGH TU Munich and ETH Zu-
1900 rpm 26 rich are working very well.
n2 20 The main disadvantage of
15 the i2-principle for tractors
P.I.V-Variator to Final REV. NEUT LOW 10 is a poor power density of
i = 0,5...2,0 H R L drive 2 6,5 the variator: High
6,5
0,5
4 torques/low speeds in the
CVT speed ratio n2/n1
4
6
2
low range and high
8 speeds/low torques in the
1 0
10 0,85 R high range. Full perfor-
R L H
1,5 mance of the variator
Travel speed would however need a
2 combination of high speeds
-10 -5 0 5 10 15 20 km/h 30
Fig. 19 System with steel belt CVT and 3 ranges (principle Kress 1962).
and high torques. A certain

load limitation in the low range may be possible by 11. They are however not used in general. If a blow
the traction limit. A reverse ratio as well as an ac- out takes place for a longer period of time, fluid
tive zero speed control needs additional efforts. temperature will exceed the limits and the CVT can
be destroyed. Modern systems prevent this danger
by a pressure limitation without blow out: The
Hydrostatic CVTs and their integration pump reduces its displacement automatically, if a
pressure signal exceeds a given limit. Such limits
Survey and basic functions are typically between 300 and 550 bar (high values
A hydrostatic CVT is formed by the combination of for high power in case of piston units).
at least one hydrostatic pump and at least one mo-
tor. At least one unit must have a continuously vari- Basic equations
able displacement.
The ideal “force balance” between mechanical and
The circuit of a simple typical hydrostatic CVT hydrostatic variables is described for pumps (P) and
unit is shown by Fig. 20. It usually works with one motors (M) by the the torque T, the displacement V
variable pump 1 and one constant or variable motor and the pressure difference p as follows:
2 in a closed circuit. A variable motor can improve p × VP
efficiency for high speeds. The charge pump 3 with TP =
safety valve 4 is always feeding the low pressure 2×p
pipe passing through filter 5 and check valve 6. Sur- p × VM
TM =
plus flow leaves the low pressure pipe automati- 2×p
cally by the flushing valve 7 arriving at the tank The ideal “speed balance” between rotational shaft
through pressure relief valve 8 and cooler 9. speed n and fluid flow Q is
The charging system (with, for example 20 bar) w ×V
is in addition often used for transmission control Q = n×V =
2×p
thus having in total the following duties:
with w as angular velocity. The ideal CVT torque
– to replace leakage ratio thus is
– to control fluid temperature
TP V
– to control fluid contamination = P
– to serve as auxiliary power TM VM
– to enable high pump speeds and the ideal CVT speed ratio
The pressure relief valves 10 are safety elements. If nP V
the charging system fails, an emergency re-filling = M
nM VP
of the circuit is possible by the suction check valves
These ideal formulas are influences in re-
ality by two groups of losses:
11 10 – leakage (reducing output speed)
5 6 – all types of friction plus charging
pump torque (reducing output torque)
7 A typical working point of a well de-
signed hydrostatic CVT with swash plate
4 8 units can have about 8 % speed loss and
1 3 2
10 10 % torque loss at a typical working
6 9 point (pressure 200-300 bar).
11
Low pressures result in low speed
losses but very high relative torque
losses. CVT working points with pres-
Fig. 20 Typical circuit of a hydrostatic CVT unit. Symbols
sures below about 100 bar should there-
ISO 1219-1. Legend see text (Renius 2003).

fore be avoided. This is one reason, to use variable Fig. 22 Pump
Displacement
motors. and motor dis-
Motor
placement ver-
Rules for high efficiency mp sus speed for the
Pu standard con-
The following rules can be recommended: cept of Fig. 21
0 Ground speed
– Pump and motor variable (Browning
rev. forw. 1978).
– Units with large tilting angles
– Actual heavy duty pressure about 200-300 bar
– At least the motor in bent axis concept ment control in case of variable motor is shown by
– Motor displacement above pump Fig. 22.
– Input speed not above Diesel engine speed The lower version of Fig. 21 has been intro-
– Controlled circuit pressure losses duced 2002 for Liebherr wheel loaders. The special
– Charging system not overdesigned gear box enables synchronous power shifts be-
– No major paddling losses tween three ranges (Renius 2003). Efficiencies can-
not compete with power split CVTs as introduced
for tractors but are on a high level for the "direct
Hydrostatic CVT integration to drive trains power flow" philosophy (Krauss and Ivantisynowa,
Power split configurations will be presented in sep- 2004) and clearly above those of transmissions with
arate chapters. hydrodynamic torque converters.
The following information thus concentrates on Coming back to tractors: A high number of hy-
CVTs with “direct power flow”. They are only pop- drostatic CVTs with direct power flow is applied up
ular for small tractors. Their efficiency is relativelyto about 30 kW. A typical structure is presented by
poor but this is not so important in most Fig. 23, based on material which has been kindly
applications. Turf maintenance and garden tractors provided by Yanmar, Japan. The diesel engine
with rated engine power of about 10-20 kW (Betz drives the variable displacement swash plate pump
1991) require, for example, high comfort and working on a constant displacement swash plate
proper speed control and they often use not all the motor. Motor output shaft is combined with a con-
engine power for driving due to powered imple- ventional three range gear box in order to meet the
ments. tractor speed range demands.
There are several possibilities to combine a hy-
drostatic CVT unit with mechanical transmission Electrical CVTs and their integration
components. Fig. 21 shows two concepts. The up- Electrical CVTs have the following advantages:
per configuration is typical for most CVTs of
smaller tractors. The standard pattern of displace- – Low noise level,
– low maintenance/repair costs,
Stepped – low energy transfer losses
gear box – high ability for control systems
– low required energy for ratio control
– environmental friendly
Critical points are safety aspects for the high re-
Stepped
gear box
quired voltage, costs and efficiency.
An example may demonstrate the voltage de-
mand. If we assume 100 kW power and 200 A cur-
rent, a voltage of 500 V is needed.
Fig. 21 Combination of hydrostatic CVTs with
conventional gear boxes. Above: Standard con- Electrical drives could gain importance for mo-
cept for small tractors, below: new concept of a bile machinery drives because of the availability of
wheel loader. the following new technologies:

Tarasinski (2004),
Fig. 24. The efficiency
band has about the same
level as for hydrostatic
Rear axle
units, but the level of ro-
PTO tational speed is much
Hydrostatic
rear higher. The asynchron-
CVT or ous machine can
front achieve efficiency lev-
3 ranges or els above 85 % in a wide
Front axle drive both speed band, while the
PME concept with per-
Clutch for manent magnets has
sharp turns better peak values but
they are limited to low
speeds. The reluctance
machine can offer high
Fig. 23 Drive train of Yanmar tractors with hydrostatic CVT (swash plate efficiencies at high
units) for about 20-25 kW rated engine power. speeds. Speed levels
should be as high as pos-
– new power electronics for sible as the force density is rather poor. According
• AC/DC resp. DC/AC transformation to Tarasinski (2004), a circumferential force den-
• continuously variable voltage sity of only 20 kN/m2 between rotor and stator is
• continuously variable frequency typically for a fluid cooled asynchronous machine
• fast induction current control and 40 kN/m2 for an intensively fluid cooled syn-
– new magnetic materials and cooling principles chronous machine. The prices for power electronics
– digital overall control systems are still very high, but a decrease is expected by
As outlined earlier by Table 5, electrical CVTs their introduction for passenger car hybrid trans-
have a potential for tractor drive transmissions in missions.
power split systems, later on perhaps also in a direct Such a transmission has been introduced com-
mode with fuel cells. All possible solutions may re- mercially first by Toyota for its “Prius” in 1997
quire additional conventional transmission ele- (Killmann et al. 1999). It is since 2003 in the second
ments, such as ranges and gear wheels for speed generation on the market without changing its basic
reduction. structure Fig. 25. The engine drives the carrier of a
Efficiency trends of electrical generators and planetary, where the power is split into a CVT path
motors under typical loads have been collected by via sun gear/generator/controller/motor and a me-
chanical path to the ring gear, where merging of
100 both takes place. The system can be classified to be
% output coupled (see following chapter). It has sev-
80 eral additional functions such as starting the gaso-
Efficiency
PME Synchronous line engine, electrical braking (charging the

60 Asynchronous battery), charging the battery in case of power sur-
Reluctance plus, electrical boosting, reversing and driving slow
40 speeds electrically for example for stop and go traf-
fic.
20 The Prius reinforced broad research on hybrid
Speed CVTs in the car industry. But it may also encourage
Fig. 24 Typical efficiencies for three concepts of research for electrical tractor drives, which is car-
electrical machines (Tarasinski 2004). ried out since some years for example at the Univer-

Controller Battery Ni-MH The basic idea of the power split principle is dem-
onstrated by Fig. 27.
Gasoline engine The input power is split into
Merging
(special – a constant ratio mechanical path and
design)
– a variable ratio path using a CVT unit
On the output side of the transmission both power
flows are merged again. Six basic power split con-
figurations with planetaries like Fig. 26 have been
analyzed by Jarchow (1964). Total efficiency is
typically above that of a CVT “direct” due to the
high efficiency of the mechanical path. This shall
Electr. synchronous unit Power split point Electric
be demonstrated by a typical example with realistic
(4 planets) synchr. unit values for full power:
We assume for the structure of Fig. 27 a split ra-
Fig. 25 First commercial hybrid car CVT, working
with an electro-mechanical power split system:
tio of “70 % mechanical” and “30 % with CVT
Toyota “Prius” 1997. unit” with a related mechanical efficiency of 95 %
and a CVT efficiency of 80 % (hydrostatic, swash
sity of Dresden (Barucki 2000). Also tractor plate units). The resulting total efficiency can be
companies, mainly Agco-Fendt and John Deere, in- calculated as
dicated activities.
0.7 × 0.95 + 0.3 × 0.80 = 0.905 = 90.5 %
PRINCIPLES OF POWER SPLIT The efficiency benefit increases with decreasing
CVTs CVT power portion. Best total efficiency can there-
fore usually be expected with zero power in the
CVT path, which is possible with hydrostatic units.
Basic functions of power split CVTs Total efficiency would be for such a lockup point in
The previous chapters indicate, that the efficiency this example above 90.5 %, but below 95 % due to
demands for tractor CVTs above about 50 kW can- the idling losses of the hydrostatic unit and its leak-
not be met by “direct” hydrostatic or “direct” elec- age (it is still loaded, but delivering no power). An-
trical transmissions other not so favorable phenomena with power split
CVTs is the possibility of circulating power. Re-
A special transmission system is needed, called
garding again the system of Fig. 27, we can assume,
“power split” system in order to increase effi-
that the variator output speed to the sun gear can be
ciency. It often needs a planetary for splitting or
reversed in case of a hydrostatic or electrical CVT
merging the power.
unit. This means, that its power flow direction is re-
Fig. 26 shows the
Ring Planet
three-shaft standard plane- Mechanical Power
tary set. Input path merging
Three planetary gear Sun
wheels are typical as they Carrier 3
offer outstanding self cen- Power 1
tering of the sun without a Fig. 26 Standard split 4 Output
bearing and optimal power planetary gear set.
distribution. Speed rela- 2
tions can be briefly analyzed by plotting the straight
lines of circumferential speeds for all elements. Variable path
Further fundamentals see Browning (1978).
Fig. 27 Principle of a CVT with external power
split and two power transmitting pathes.

versed - in Fig. 27 now from right to left. This number of the teeth at ring gear
i0 = - , here
power is superimposed to the system input power, it number of the teeth at sun gear
must be transferred by the mechanical path again to i 0 = -3
the right and it is thus called “circulating”. If its
With these equations, the ratio of system input to
value is low compared to the input power, total sys-
output speed can be found as
tem efficiency can still remain above that of a direct
CVT. This strategy is therefore often applied in hy- w in speed input shaft
i total = = =
drostatic power split tractor transmissions in order w carrier speed output shaft
to increase output speed range.
= 0
(i - 1) × i 12 × i 34 × i var
i 0 × i 12 × i var - i 34
Efficiency versus speed ratio range
The variator speed ratio range is defined as
An increased system efficiency results in a reduced
i var,max
speed ratio range and an enlarged system speed ra- S Var =
tio range creates efficiency losses (both cases com- i var,min
pared with CVT mode “direct”). In order to where
demonstrate a typical speed and speed ratio model- i var,min minimum variator speed ratio
ing, the system of Fig. 27 shall be analyzed as an
i var,max maximum variator speed ratio
example in the following section.
In this example SVar is assumed to be 4, symmetri-
The rotational speed of the sun and the ring gear
in Fig. 27 can be determined by cally to ivar = 1 resulting in the values ivar,min = 0.5
and ivar,max = 2.0.
w in
w sun = If we assume i12 = i34 = –1, the speed ratio range of
i 12 × i Var the total transmission system Stotal gets to
where max( i total ) 1.14
w in S total = = = 1.429
rotational speed of input shaft min( i total ) 0.8
speed shaft 1
i 12 = This means that the total transmission ratio range is
speed shaft 2
1.429 and thus only about 35 % of the variator
speed shaft 2
i Var = speed ratio range.
speed sun wheel
The speed of the ring gear can be expressed as fol- Internal and external power split configu-
lows:
ration
w
w ring = in Power split CVT concepts can be divided in the two
i 34 groups
where – CVTs with internal power split
speed shaft 3 – CVTs with external power split
i 34 =
speed shaft 4 The transmission of Renault described already in
The shaft speeds at a planetary gear set, Fig. 26, can Fig.2 is an example for the first group. Fig. 28
be expressed by the equation of Willis: shows a similar concept.
The swash plate axial piston units are placed in
w sun - w ring × i 0 = w carrier × (1 - i 0 )
a compact back-to-back design. The cylinder
where block 1 at the primary unit is driven by the input
w ring rotational speed of ring gear speed n1, the cylinder block of the secondary unit is
w carrier rotational speed of carrier fixed in the surrounding transmission housing. The
unit capsule and both swash plates revolve with the
output speed n2. Thus the fluid displacement at the

power is transmitted fully mechanically,
but the hydrostatic unit is of course still
under pressure (leakage).
To the knowledge of the authors this
concept has been realized only in proto-
types (Molly 1966) and not introduced
commercially for agricultural machinery.
As explained in the following section, this
system is similar in basic mechanics to out-
put-coupled external power split concepts.
Fig. 28 Continuously variable transmission with internal

power split (n rotational speed, T torque, P power, 1 input, Power split CVTs with input- and
2 output). output-coupled planetaries
primary unit is determined by the speed difference The following considerations concentrate
n1 – n2 and at the secondary unit by the output speed on external power split systems, as they seem to be
n2. The input torque of the pump is working on the more important type for tractors and other mo-
both, the fluid and via the swash plate on the revolv- bile machinery. A huge number of solutions has
ing capsule. Therefore the power split ratio is con- been proposed (also by many patents): sometimes
trolled by the speeds: difficult to analyze. An excellent early classifica-
tion and mathematical analysis has been presented
Phydr = 2 × p × T1 × ( n1 - n 2 ) by Kress (1968). For CVTs with external power
split and three shaft standard planetaries (Fig. 26),
Pmech = 2 × p × T1 × n 2
Kress formed two typical groups, Fig. 29:
where
A Systems with input-coupled planetaries
w = 2 × p× n
B Systems with output-coupled planetaries
T1 torque on input shaft
“Coupling” means to cover the surplus grade of
n1 primary rotational speed freedom of the planetary. This classification is very
n2 secondary rotational speed useful and was confirmed by many other authors
Pmech power input to mechanical path (Browning 1978, Ziebart et al. 1983, Renius 1999
and others). There are further modifications possi-
Phydr power input to hydraulic path ble within the groups A and B by other planetary ar-
Power merging takes place at the swash plate of rangements (Jarchow 1964).
unit 2.
In neutral position of swash plate the primary
unit produces no oil flow and
the output speed is zero. Driv- A Planetary input coupled B Planetary output coupled
ing through the forward speed
range, the primary swash plate
Input Mech. path Output Input Mech. path Output
is first swivelled to the maxi-
mum tilt angle, while the sec- CVT path CVT path
ondary swash plate keeps its
maximum position. Then the
displacement of the secondary Power split,
Power split, Power merging, Power merging,
unit is reduced to zero. At start
fixed speed ratio fixed torque ratio fixed torque ratio fixed speed ratio
up the hydraulic power is
100 % in the very first moment. Fig. 29 Two basic concepts of continuously variable transmissions us-
If both speeds are equal, the ing the external power split principle (classification of Kress 1968).

The structures A and B are mirrored at each other, The target of Fig. 11 would require for the system
but their operational behavior is completely differ- of Fig. 30 efficiencies of about 90 %, as Fig. 11 in-
ent, as shown in the following. cludes the final drive. This limits the usable band-
Analysis of power split concepts should address width of output speeds with the consequence of
the expected efficiency characteristic, the portion additional ranges, as realized by Claas, Steyr, ZF
of the power through the variable path and, the cir- and John Deere.
culating power and the rotational speeds of trans- If backward operations with small speeds are
mission elements. important, a mechanical reverser should be consid-
The following analysis is done with the soft- ered to supply acceptable efficiencies.
ware Matlab, based on constant full power efficien- Fig. 31 shows for concept B (output coupled)
cies, which is a simplification but can aid as a first completely different characteristics. Starting up
useful approach (Resch 2004). The following sign from zero output speed the power in the variable
convention was used: path is first equal to the input power resulting in a
– Positive speeds have the direction of the input poor efficiency. The absolute losses are again miti-
speed. gated by the traction limit (Fig. 10). The power
– Output power is considered to be negative, transmitted by the mechanical path is increasing
whereas input power is positive. proportionally with output speed. The total effi-
ciency rises linearly as constant efficiencies are as-
– The sign of torque is determined according to
sumed for both power split paths. If the primary
the signs of speed and power.
shaft of the variator is locked up, the power is trans-
The diagrams in Fig. 30 and 31 use 97 % efficiency
mitted fully mechanically arriving to the best possi-
for each pair of gear wheels and 85 % for the
ble total efficiency, as described earlier. Also in this
variator. The variator value is based on hydrostatic
case, practical peaks are some % lower as explained
high angle bent axis units. System efficiency dia-
before. Once the primary shaft starts to turn in the
grams for other variator efficiencies follow later on.
opposite direction, negative circulating power
Fig. 30 analyzes concept A (input coupled) by arises and the efficiency decreases in a steeper de-
estimated efficiency plots and related power in the cline. Also this power split system requires addi-
variable path indicating in the same time the circu- tional ranges as realized by Fendt. For backward
lating power. This is all plotted versus transmission operation the rotational speed of the sun wheel
output speed (as a measure for vehicle speed). Typ- changes its direction and turns to be an input shaft
ical for this concept is the hyperbolic increase of the to the planetary gear set. Positive circulating power
power in the variable path towards zero output occurs creating poor efficiencies, however in this
speed. Near the start up point there is a huge amount case mainly for higher reverse speeds. If they are
of circulating power with the consequence of poor important, a mechanical reverser should be consid-
efficiencies although mitigated by the traction lim- ered.
its (Fig. 10). The circulating power for the forward
Fig. 32 and 33 demonstrate the influence of the
direction is called “negative” for the variator, as its
variator efficiency on the system efficiency. Sensi-
direction points from right to left – opposite to the
tivity is very high in working points with high
normal direction. The circulating power for the
variator power - in particular for reversed output
backwards regime is called “positive”, as the direc-
speeds.
tion is normal. In this case, the CVT unit has how-
ever to handle a higher power level – always above
input power – thus creating a very poor total effi-
ciency level. The best efficiencies are achieved in
the forward mode with low power portions in the
variable path with an optimum for the lockup point.
Practical peak values are some % lower due to the
idling friction losses and the leakage of the variator
being under load.

n T P n T P n T P
Ring - - + - + - + + +
Sun + - - + + + + + +
1.0
Total efficiency h
0.8
Shaft 2 locked-up
0.6
0.4
0.2
Related variator power PVar /PIn
8 Positive circu- Negative circu- 0

lating power lating power Power split
6
0
Fig. 30 Full power character-
istics and efficiencies for an in-
-2 put coupled CVT
fixed configuration (upper part: di-
-4 1 rection of torque T, speed n
Power fixed path and power P at the planetary
-6
variable 2 Power variable path gear set). Assumed efficien-
-8 cies: spur gears 97 %,
backward 0 forward
Speed output shaft Variator 85 %.
n T P n T P n T P
Ring + - - + - - - - +
Sun - - + + - - + - - 1.0
Total efficiency h
Power fixed path

0.9
Power variable path
0.8
0.7
Positive circu- 0.6

lating power Negative circu-
Related variator power PVar /PIn
3 Power split lating power
0
Fig. 31 Power characteristics
and efficiencies for an output
Shaft 2 locked-up
Shaft 1 locked-up
-1 coupled CVT configuration

fixed (upper part: direction of
-2 torque T, speed n and power P
1 at the planetary gear set). As-
-3 sumed efficiencies: spur gears
variable 2 97 %, Variator 85 %; maxi-
-4
backward 0 forward mum torque ratio input to out-
Speed output shaft put 50.

1.0
Conclusions for practical applica- 0.90
tions 0.85
0.8
Total efficiency h
1. Power split CVT systems are able to
increase system efficiency above 0.6
that of a “CVT direct” mode.
0.80
Shaft 2 locked-up
2. Both principles – power split sys- 0.4
tems with input and output coupled hVar=0.75
planetaries – are of interest and are 0.2
commercially used.
3. Power split CVT systems cannot 0.0
backward 0 forward
cover the full required tractor speed Speed output shaft
range, they need additional mechan-
ical ranges. The smaller the relative Fig. 32 Influence of the variator efficiency on the overall effi-
variator power, the better the system ciencies for the structure from Fig. 30.
efficiency, but the higher the
required number of additional 1.0
ranges.
0.90
4. Power split system engineering 0.9 0.85
Total efficiency h
should consider efficiency benefits 0.80

of the lockup point. It is favorable to hVar=0.75
apply this point for often used heavy 0.8
duty working speeds (for example
Shaft 2 locked-up
Shaft 1 locked-up
6-12 km/h) and high transport
speeds. 0.7
5. Computer aided analysis is recom-

mended with tools like
0.6
Matlab/Simulink, AMESim and backward 0 forward
others. Speed output shaft
6. Losses of the spur gear wheels have Fig. 33 Influence of the variator efficiency on the overall effi-
a higher influence on the system ef- ciencies for the structure from Fig. 31.
ficiency than often expected, in par-
ticular for working points with circulating
power or high rotational speeds. A proper ac-
EXAMPLES OF POWER SPLIT CVTs
tive lubricating system (oil pump) and oil level
FOR TRACTORS
management is recommended to prevent plung-
ing losses. Introduction
7. Input coupled systems are able to produce zero Five concepts of power split CVTs have been pre-
output speed also with mechanical variators, sented commercially for larger tractors until now as
such as chain variators or toroidal units mentioned in Table 5. Four of them could enter and
(“geared neutral”). The disadvantage of very stabilize a series production stage introducing
high circulating power portions are mitigated meanwhile most of the control strategies as listed
for tractors by traction limits (Fig. 10). by Table 3.
The Fendt “Vario” was the first (1996), fol-
lowed by Steyr S-Matic (2000), ZF “Eccom”

(2001) and the John Deere “AutoPowr” for the planetaries (Jarchow 1981, Renius and Koberger
7010 series (2001). All dates indicate use for trac- 2001, Linares 2003). The shifts between the ranges
tors produced in series. The Steyr transmission are done at synchronous speeds with dog clutches
business belongs since May 1, 2000 to ZF. The sta- and the power is handed over without any power in-
tus of 1998 (Claas, Fendt, Steyr and ZF) has been terruption in a very short time, in which the ratio is
reviewed by Lang et al (1998) collecting some ba- however kept constant (no acceleration/ decelera-
sic specifications (1998). All transmissions have tion of the vehicle). This principle is very difficult
been commented also in the German Yearbook Ag- to realize but doesn't need friction clutches (costs,
ricultural Engineering 1997-2003, always within idling losses). Safe, comfortable shifts require a
the chapter “engines and transmissions” (see, for high sophisticated electronic control system.
example, Renius and Koberger 2001). A very de- Measured Claas “HM-8” efficiencies have been
tailed mathematical analysis of all mentioned trans- found to be outstanding (diagrams see Renius and
missions has been presented recently (in Spanish) Vahlensieck 1996, Renius 2000), covering the tar-
by P. Linares (2003) referring to the fundamentals get of Fig. 11. Both Claas CVTs will not be dis-
of Kress (1968). cussed more in detail because of limited space in
The transmission systems are represented by this paper.
uniform own graphics using ISO 1219 symbols for
hydrostatics and own symbols for mechanical ele-
ments (Renius 1968 and 1976a).
The Fendt “Vario”
Fendt announced its new, infinitely variable
“Vario” transmission at Agritechnica 1995 for the
The Claas-Jarchow “HM-8” and “HM-II” largest new Fendt tractor “926 Vario” (191 kW),
A lot of pioneering work was carried out since the Fig. 34. It became the worldwide first in series pro-
late 1980s by Claas resulting in the concepts Claas duced power split transmission for standard tractors
“HM-8” (about 140 kW) and the later larger Claas (1996). Its principle was soon expanded to more
“HM-II”, also called “Traxion” (about 220 kW). and more other Fendt models forming meanwhile 3
Both have been developed for the Claas carrier trac- transmission families. At present (end of 2004) 4
tor “Xerion” and both are based on a patent of tractor families 400, 700, 800 and 900 are in pro-
Jarchow, a leading German spe-
MAN
cialist in power split systems Axial piston pump (bent axis)
(Jarchow 1981). The HM-8 had 7 191 kW 3
power split ranges in order to keep 2300 /min -30° to + 45°, max. 233 cm /rev.
the hydrostatic power portion low Power split planetary
and efficiency high. One creeper
range works “direct”. The concept
became public 1989 and was pro- PTO
duced since 1996, while the HM-II
(Traxion) with 5 power split Axial piston motors (bent axis)
3
ranges, one “direct” and a conven- Merger 0° to + 45°, max. 233 cm /rev
tional power shift reverser by shaft
clutches was presented at
Agritechnica 1999 (Fredriksen
1999). Both transmissions could L H
Rear axle
only arrive to a limited production
volume. to Front axle
The Claas-Jarchow approach is
characterized by an input coupled Fig. 34 First successful infinitely variable hydrostatic power split
four-shaft compound planetary transmission for standard tractors, Fendt “Vario” for the tractor
which consists of two standard “Favorit 926” (1996).

(axle output / gearbox input power)
90
duction with the “Vario” CVT as standard
Full load driveline efficiency

Measurements of Fendt Company,
equipment. It became recently also adopted % confirmed by DLG-tests
to MF tractor series 7400 (2003) and 8400 Range L Range H
85
(2004), called “Dyna-VT”. About 40.000
“Vario” transmissions have been produced
end of 2004. It could catch an outstanding
80
number of innovation awards and became
Target
the main driving force for the introduction (Renius 1993)
of power split transmissions for tractors.
75
The Fendt “Vario” is a power split CVT
with output coupled standard planetary 0
(compare with Fig. 29). The structure is 3 4 5 6 10 12 20 km/h 40
based on a principle of Molly (1966). Basic Vehicle speed (at rated engine speed)
development was done by the Fendt engi- Fig. 35 Full load drive line efficiencies (axle power by
neer H. Marschall (1936-1989), who got transmission input power) of the Fendt 926 “Vario” and its
several patents on it (Marschall 1973). In comparison with the target of Fig. 11.
the forward direction, power is split by the
planetary transmission and merged by the
Fendt started therefore an own development of
shaft of the two hydrostatic units (Dziuba and
45 degree variable bent axis units, in the final stage
Honzek 1997). The basic power characteristics fol-
with support of Sauer-Sundstrand (today
low Fig. 31. Hydrostatic power portion is 100 % in
Sauer-Danfoss). These units offer in their best
the starting point, but zero at the top speeds of the
points efficiencies of 95-96 % (swash plate units
ranges.
rather 89-92 %). The combination with a very low
The power split unit is combined with two con- number of mechanical transmission elements re-
ventional ranges. Range L covers speeds up to sulted in an excellent full load efficiency curve,
32 km/h, range H up to 50 km/h. These top speeds Fig. 35, which could well meet the target of Fig. 11
are near the lock-up points. The speed potential be- in spite of the high hydrostatic power portions. Due
yond the lock-up point is not used in all “Vario” to the high efficiency level it is even possible to re-
transmissions in order to prevent any circulating verse in power split mode in spite of regenerative
power in the forward mode. Hydrostatic power power.
flow is thus not reversed. Range shift is not needed
A serious problem of the Fendt philosophy was
in the field. Regarding on-road operations, an auto-
the high noise level of the large units. A break
matic shift on-the-go was introduced 2002/03. It al-
through in noise reduction could be achieved by
lows finger tip range selection within a speed band
isolating the CVT body noise by elastic suspen-
up to 20 km/h. The shift is done by active
sions (as usual for IC engines).
synchronization support of the CVT. This is, for ex-
ample, useful for heavy transports on hilly roads.
The small power interruption is mostly not realized The Steyr “S-Matic” for Steyr, Case, New
by the driver. For light duty transports, range H is Holland and others
sufficient for all speeds from zero to 50 km/h.
The Steyr “S-Matic” has been developed within the
The Fendt concept requires very high maximum
1990s by the Austrian company “Steyr
hydrostatic power and thus large units. All com-
Antriebstechnik (SAT)” which was taken over by
mercially available variable axial piston pumps and
Zahnradfabrik Passau GmbH (part of ZF) in 2000.
motors had been quoted to be too poor in efficiency
A first concept became already public in 1994
to meet the target of Fig. 11. “Wide angle” units
(Ulbrich 1994), series production was announced
with spherical pistons and piston rings had been
with a modified structure (Aitzetmüller 1999) and
early proved at TH Braunschweig to offer outstand-
realized early 2000 for the standard tractors Steyr
ing efficiencies (Hoffmann 1970), but had been
“CVT” and the Case “CVX” (both 88-125 kW).
available only in fixed displacement versions.

Speed K1 K2 K3 K4 The transmission
Swash plate axial piston units 1 0-8 km/h has been well ac-
2 8-14 km/h cepted by the mar-
3 14-30 km/h
ket and is
55 cm3 -55...0...55 cm3 4 30-50 km/h
engaged meanwhile also
Power split point PTO forward: KV used by the New
backward: KR Holland TVT trac-
tor series (2004).
McCormick has
announced at
EIMA 2004 to in-
troduce it to a new
tractor series VTX
K1 K2 K4 K3 KV KR to Rear axle in 2005/2006.
Merging The power split
planetaries only Dog clutches
CVT works with an
Fig. 36 Hydrostatic power split transmission Steyr “S-Matic” (2000). input coupled five
shaft compound
35 planetary, followed by two standard
km/h 5 m/s²
planetaries and a planetary for reverse,
30 Fig. 36. Four ranges are used to keep the hy-
drostatic power level low. It varies between
25
zero and 50 % (Lang et al. 1998). The hydro-
Tractor speed
Tractor speed
static unit is formed by swash plate units in
20
back-to-back configuration (compare with
Fig. 7) and can be dismounted/mounted eas-
15
Tractor acceleration ily at the transmission casing. The ranges are
10
shifted at synchronous speeds using the dog
clutches K1, K2, K3 and K4. The power is
5 handed over without power interruption, sim-
ilar to the Claas philosophy. The sophisti-
0 cated function was optimised by simulation.
0 1 2 3 4 5 s 6 The interruptions of acceleration could be
Time made very short as demonstrated by Fig. 37.
Fig. 37 Acceleration test of the Steyr “S-Matic” demon- Efficiencies of the S-Matic have been
strating range shift. 120 kW tractor on the road without published by Leitner et al. (2000) excluding
trailer. Data courtesy Steyr. axles, Fig. 38.
If axle efficiencies are included (for
100
Efficiency without axles
example 95 % for 6-12 km/h), the result

% is little below the target of Fig. 11. Ac-
80
cording to Leitner et al (2000), efficien-
cies increase 2-3 % for reduced input
70 speeds.
60 In order to solve noise problems,
50
critical actual frequencies have been
0 10 20 30 40 km/h 50 analyzed. If it was found, that they are
Vehicle speed amplified by resonance, a shifting up of
the Eigen frequencies of the relevant
Fig. 38 Full load transmission efficiency of the Steyr S-Matic.

Swash plate axial piston units
Power split point
K1 K3 B KV KR
K2 K4 PTO
110 kW to Rear
axle
2300/min
Merging planetaries
K1 K2 K3 K4 B forward: KV Fig. 39 Hydrostatic
0-6,4 km/h backward: KR
(same velocity) power split transmis-
6,4-13 km/h sion ZF “Eccom”
13-26 km/h to Front axle (2001).
26-51 km/h
housing sections was done by increased their stiff- Also other tractor companies are considering to use
ness. This method is very successful and can be the Eccom. About 12.000 units are in the market.
generally recommended (Kirste 1989). The four ranges concept allows it to keep the
Load spectra on the basis of Fig. 13 and 14 and hydrostatic power level low. It varies between
speed spectra on the basis of the CIGR Hand- about plus 40, zero and minus 35 % of total power,
book III (Renius 1999) have been applied for Fig. 40 (Pohlenz and Gruhle 2002) and uses – simi-
dimensioning the S-Matic (Leitner et al. 2000). lar to Steyr - a compact hydrostatic package of
swash plate units in back-to-back arrangement
(compare with Fig. 7). The power split system
The ZF “Eccom” for Deutz-Fahr, John works with an input coupled planetary. The vari-
Deere, Claas and others able path is starting at the split point via the PTO
Development of the ZF Eccom has been presented shaft arriving at the variator by the right three gear
first by ZF/ZP at Agritechnica ’97. It is at present wheels. Fig. 40 shows well the linear output speed
realized with the common basic structure of Fig. 39 variation of the hydrostatic unit between minus
by 3 transmission families (Pohlenz 2300 and plus 2300 rpm crossing at zero the lockup
and Gruhle 2002). Eccom 1.5 (max. 100 2300
110 kW) and Eccom 1.8 (max. Typical transmission input power (field)
% rpm
130 kW) are in production. A third
version Eccom 3.0 (max about 220 Speed
60 0
sun 1
kW) shall be available soon.
Power
40 -1150
Deutz-Fahr was the first tractor Hydrostatic
company using the Eccom (version 20 power portion -2300
Speed sun 1
1.5) in its new series “Agrotron TTV” 0

(92/103/110 kW) starting a pilot se-
ries in 2000 and series production in -20
Range
2001. John Deere applies the Eccom -40
since 2001 for its 6000 tractor line ad- 0 5 10 15 20 25 30 35 40 km/h 50
vertising it as “AutoPowr”. Claas has Vehicle speed
announced to use the Eccom 3.0 for Fig. 40 ZF “Eccom”: Hydrostatic power portions and speed of
its new Xerion carrier tractor family. first left sun gear within the 4 ranges (Pohlenz and Gruhle 2002).

point. Three combined standard planetaries (two of structure is completely different, as the concept is
them merging) form a five shaft compound plane- input-coupled and uses a 4-shaft compound plane-
tary. This could produce in principle 3 ranges using tary for the power split system (Mistry and Sparks
friction clutches K1, K2 and K3. They are shifted at 2002, Linares 2003). The engine drives both, the
synchronous speeds. A forth standard planetary variable axial piston unit and the left sun gear of the
(K4, B) enables enlargement of the system to four merging compound planetary. Its ring gear speed is
ranges. Forward-reverse shift is realized by two controlled by the fixed displacement hydrostatic
conventional reverser clutches (KV, KR). Clutch unit (both directions). Its carrier output is used for
dimensions are large due to the high possible slip- range L (KL, max. 15 km/h) and its output of the
ping speeds but they offer a very powerful revers- right sun gear used for range H (KH, max. 50 km/h).
ing. Shift between both is done at synchronous speeds.
Efficiencies of this transmission are published The transmission is prepared for top speeds up to 62
excluding the axles (Pohlenz and Gruhle 2002). If km/h. An extra planetary for reverse (max. 17
axles are included, they can cover the target of km/h) uses the carrier output of the merging plane-
Fig. 11 near the lockup points. Noise reduction has tary (as for L), but reverses output direction by
been achieved by measurements and analytical brake KR and generates output speed at its sun gear.
methods. The concept of the power shifted reverser is well
known from the John Deere “PowrQuad” transmis-
sion (CIGR Handbook III, Renius 1999).
John Deere “Autopowr ” (IVT-USA) At start-up the ring gear turns in opposite direc-
JD presented a newly designed transmission for its tion of the engine speed. Here regeneration occurs
7710 and 7810 tractors in 2001, Fig. 41. Rated en- as the variable unit acts as motor and the fixed unit
gine power was 118 and 129 kW but the engines of- as a pump. At the lock-up of the ring gear, in about
fered some extra power. The system works with the middle of the range, efficiency reaches its maxi-
two 45° bent-axis axial-piston units from mum value. When the ring gear speed has the same
Sauer-Danfoss (160 ccm/rev), similar in design to direction as the engine, the variable unit works as a
those which have been earlier installed in the Fendt pump and the fixed unit acts as motor. The trans-
“Vario” transmissions. However the transmission mission control is described in the paper of Mistry
and Sparks (2002) explain-
Power merging ing hydraulic circuit in de-
160 cm 3 KR tail and electrical circuit in
45° Shift L H synchr. concept.
Measured efficiencies
KL Towing away are not known by the au-
Power Split KH thors. But the 45° degrees
units and a special oil level
PTO management provide favor-
able conditions for good
values. The hydrostatic sys-
Park tem pressure relief valves
max. 160 cm 3 brake are set at 550 bar (compare
-45°...0...45° with Fig. 20), but typical
pressures under heavy load
Zero speed position (field 6-8 km/h) are about
to Front axle 250 bar. Charge pump de-
livers 40 l/min at rated en-
gine speed with flushing
Fig. 41 Hydrostatic power split transmission John Deere “AutoPowr” for output of 24 l/min at 18 bar
John Deere 7010 (and later 7020) tractor series. (Mistry and Sparks 2002).

Power split Power merging Fig. 42 Vehicle drive with power
split CVT of Yanmar rice
transplanter (2002).
Ranges
L, H R
13 kW
4 gear wheels for
speed change of
planting mechanism
H L
to Front axle
to Rear axle
EXAMPLE OF A POWER SPLIT CVT CONCLUSIONS AND OUTLOOK

FOR A RICE TRANSPLANTER
Power split CVTs with hydrostatic variator units
Yanmar Agricultural Equipment Co., Ltd., Japan, will continue to gain importance for tractors for
produces since Febr. 2002 a self propelled rice countries with high tech farm mechanization, and in
transplanter in series, which uses a power split CVT addition also for other mobile machinery such as
for the vehicle drive. This is – as far as the authors self propelled agricultural working machines or
know – the first series application of a power split construction machinery.
CVT for the ground drive of a self propelled agri- This trend is supported by new electronically
cultural working machine and could become a trend controlled Diesel engines and electrohydraulic
also for other self propelled agricultural machinery. CVT control strategies and leads to improvements
The Yanmar CVT was therefore included briefly in in productivity, precision of work, energy effi-
this paper. The information for developing the ciency, environment protection and driver comfort.
graphics of Fig. 42 has been kindly provided di- All these benefits are important, driver comfort is
rectly by Yanmar, Japan. sometimes most underlined by customers.
The 13 kW engine drives both, the variable hy- The progress by CVTs influenced meanwhile
drostatic axial piston unit and the carrier of the stan- also the development of stepped transmissions,
dard planetary. Power split takes place within the mainly those with partly or full power shift. If the
variable unit, power merging in the planetary with engine offers a high torque back up, the drive train
the sun connected to the second hydrostatic unit. gross traction hyperbola (Fig. 10) can be ap-
The system thus works with an input coupled plane- proached relatively well, if a certain engine speed
tary and delivers characteristics similar to that of variation is accepted. This is becoming a common
Fig. 30. Power is directed from the planetary ring strategy for pulling operations while there are some
gear to a conventional gear box, which contains for- limitations in engine speed variation for PTO oper-
ward-reverse range, two forward ranges L, H, a ations.
speed shift for the planting mechanism and output
These improvements balance however a little
shafts for rear and front axle. Measured efficiency
the market trend to CVTs.
curves are not available, but the system has been
well accepted by the market according to informa- Hydrostatic units are in the foreground building
tion from Yanmar. power split CVTs for tractor drives. Two axial pis-

ton concepts are common, swash plate and bent mechanization level than the highly developed in-
axis units, as demonstrated by the examples. Swash dustrialized countries. The low developed areas
plate pumps and motors are available in a large va- may not need high tech solutions because of low la-
riety by many producers worldwide while variable bor costs and other reasons, but they may need reli-
wide angle bent axis units are by far not so com- able, simple tractors, adopted to local conditions.
mon. Automation will however continue to penetrate the
For tractors below about 100 kW, mechanical upper technology levels (Table 1). After the presen-
CVTs like the pull type PIV variator remain inter- tation of the first commercially available autono-
esting because of their outstanding efficiencies. An mous guidance system by GEO TEC (Bittner 2000,
active zero speed control is possible by a power Freimann 2000) at Agritechnica 1999 (Gold
split configuration for low speeds (“geared neu- Medal), research and development has been rein-
tral”) which could be combined with a “direct” forced considerably on this subject. Master-slave
mode and a reverser similar to the famous proposals systems as well as small autonomous vehicles (plat-
of J.H. Kress (1962). Combinations of this kind are forms) are under research for further steps
also in discussion for car transmissions with fric- (Schueller 2002). They all will need mechatronic
tion drive CVTs. Range change can be made very drive systems, so that this business will remain im-
comfortable today due to model based control lay portant anyway. The authors see a certain shortage
outs. in skilled engineers in this field and would like to
More attention than in the past should be paid in encourage young professionals to develop their
the future to infinitely variable drives also for the skills in this fascinating field.
PTO in order to make its speed independent from
the engine speed. The required band width of out- Acknowledgments
put speeds is much smaller than for the ground
drive but efficiency must be extremely high, as The authors would like to thank all colleagues,
tractors are sold by PTO power in several countries. company representatives and other personalities
who supported development of this paper and they
A broad discussion and some research has
express also their gratitude to ASAE for the fruitful,
started on electrical units. They need a considerably
motivating co-operation.
higher speed level for adequate power density and
first costs, but are quoted to have potential for trac- The paper layout has been kindly performed by
tor drives in power split systems. co-author Dr. Rainer Resch.
The use of fuel cells is still far away in the future
but could make one day an electrical drive without
generator possible with high overall efficiencies. A
close combination of an electrical motor with each
individual tractor tire would offer a maximum flex-
ibility but seems to be not realistic because of the
wide band of speeds and torques for tractors
(Fig. 10 ). The authors rather expect a combination
of an electric motor with a range shift and final re-
duction drives. It could however be attractive to
drive and control at least each axle independent to
prevent circulating power and achieve better trac-
tion in turns (see, for example, the pull-turn-clutch
of Fig. 23 and the research results of Grad and
Brenninger at TU Munich).
The standard tractor will probably remain a key
machine worldwide for further decades. One reason
is, that about 90 % of all countries are on a lower

REFERENCES
Aitzetmüller, H. 1999. Steyr S-Matic - The future Continu- Greenwood, C.J. and G.B. Soar 1991. A practical CVT for a
ously Variable Transmission for all terrain vehicles. 13th mid-range high performance passenger car. In:
International Conference of the ISTVS, Munich Sept. VDI-Berichte No. 878: 597-610. Düsseldorf: VDI-Verlag.
14-17. Proceedings Vol. II: 463-470. Hamblin H.J. 1952. Hydraulic propulsion. Farm Mechaniza-
Asmus, R.W. and W.R. Borghoff 1968. Hydrostatic Trans- tion Vol. 4, No. 38: 229-230.
missions in Farm and Light Industrial Tractors. SAE paper Harms, H.H. and A. Jaufmann 1996. Tractor Management
No. 680570. Warrendale, PA: Society of Automotive En- System. AVEC '96 Intern. Symposium on Advanced Vehi-
gineers. cle Control Aachen. Proceedings: 1209-1225.
Barucki, T. 2000. Experiences with a Test-Bench for a Die- Höhn, B.R. and B. Pinnekamp 1994. The Autark Hybrid: A
sel-Electric Tractor Drive Train. AgEng 2000 Warwick Universal Power Train Concept for Passenger Cars. Inter-
July 2-7, paper 00-PM-25. Abstracts Vol. 2: 162-163. national Gearing Conference 7.-9.9.1994 Newcastle Upon
Betz, M. 1991. Hydrostatic Transmissions Application Con- Tyne. Proceedings: 253-258.
siderations for Commercial Turf Maintenance Vehicles. Hoffmann, D. 1970. Verlustleistungen und Wirkungsgrade
ASAE Distinguished Lecture Series Tractor Design No. eines Hydro-Axialkolbengetriebes mit sphärischen Kolben
16. St Joseph, MI: American Society of Agricultural Engi- (Energy losses and efficiencies of a hydrostatic axial piston
neers (see also ASAE paper 89-1649). transmission with spherical pistons). Landtechnische
Bittner, G. 2000. AGRO NAV Autonomous, off-road Vehicle Forschung Vol. 18, No. 3: 65-69.
Navigation and Implement Control System, using CDGPS Howson, D.F. 1964. The Case for Interconnected Engine and
and Inertial Backup. AgEng 2000 Warwick July 2-7, paper Pump Controls on a Tractor with Hydrostatic Transmis-
00-IE-007. Abstracts Vol. 1: 252-253. sion. J. of Agric. Engng. Res. Vol. 9, No. 3: 288-293.
Bradley, T.H. and A.A. Frank 2002. Servo-Pump Hydraulic Jarchow, F. 1964. Leistungsverzweigte Getriebe (Power split
Control System Performance and Evaluation for CVT transmissions). VDI-Z. Vol. 106, No. 6: 196-205.
Pressure and Ratio Control. In: VDI-Berichte 1709: 35-41. Jarchow, F. 1981. Hydrostatischmechanisches Stell-
Düsseldorf: VDI-Verlag. koppelgetriebe mit eingangsseitiger Leistungsver-
Breuer, B. 1974. INTRAC System 2000 - Vehicle Technique, zweigung (Infinitely variable hydromechanical couple
Development Methods, System Farm, and Industrial Ap- transmission with power split at input) German Patent No.
plications. SAE paper No. 740644. Warrendale, PA: Soci- 3 147 447 C2. Filed 1.12.1981, granted 14.06.1984.
ety of Automotive Engineers. Jenkins, G.W. 1997. Principles of Product Planning for
Browning, E.P. 1978. Design of agricultural tractor Worldwide Design. ASAE Distinguished Lecture Series
transmission elements. ASAE Distinguished Lecture Se- Tractor Design No. 21. St Joseph, MI: American Society of
ries Tractor Design No. 4. St Joseph MI: American Society Agricultural Engineers.
of Agricultural Engineers. Killmann, G. et al. 1999. TOYOTA Prius - Development and
Dittrich, O. 1953. Theorie des Umschlingungsgetriebes mit market experiences. In: VDI-Berichte 1459: 109-121.
keilförmigen Reibscheibenflanken (Theory of the V-belt Düsseldorf: VDI-Verlag.
traction drive). Ph.D. thesis TH Karlsruhe. Kim, H.J., E. H. Kim and K. H. Ryn 2000. Development of a
Dziuba, F. and R. Honzek 1997. Neues stufenloses contiuously variable-speed transmission for agricultural
leistungsverzweigtes Traktorgetriebe. (A new power-split tractors. Proceedings of ICAME, Vol. I: 162-169. Seoul:
transmission). Agrartechnische Forschung Vol. 3 No.1: Korean Society for Agricultural Machinery.
19-27. Kirste, T. 1989. Entwicklung eines 30 kW-Forschungstrak-
Englisch, A. et al. 2002. 500 Nm CVT-Luk Components in tors als Studie für lärmarme Gesamtkonzepte (Develop-
Power Split. In: VDI-Berichte 1709: 147-163. Düsseldorf: ment of a 30 kW research tractor as a study for low noise
VDI-Verlag. concepts). Fortschritt-Berichte VDI Series 14, Nr. 43.
Faust, H., M. Homm and M. Reuschel 2002. Efficiency Opti- Düsseldorf: VDI-Verlag.
mised CVT Hydraulic and Clamping System. In: Koberger, M. 2000. Hydrostatische Ölversorgungssyteme für
VDI-Berichte 1709: 43-58. Düsseldorf: VDI-Verlag. stufenlose Kettenwandlergetriebe (Hydrostatic supply sys-
Fredriksen, N. 1999. Presentation HM-II - transmission on the tems for chain variator CVTs). Fortschritt-Berichte VDI
CVT Event. HM-II - CVT-Transmission VIP Presentation. Series 12, No. 413. Düsseldorf: VDI-Verlag.
Hannover: Nov. 9, 1999. Proceedings: 75-108. Paderborn: Krauss, A. and M. Ivantysynova 2004. Power Split Transmis-
Claas Industrietechnik (CIT). sions versus Hydrostatic Multiple Motor Concepts - a
Freimann, R. 2000. Investigation of Position Accuracy of an Comparative Analysis. SAE paper 04OH-8. Warrendale,
Autonomous, Off-road Vehicle Navigation. AgEng 2000 PA: Society of Automotive Engineers.
Warwick July 2-7, paper 00-IE-7ii. Abstracts Vol. 1: Kress, J.H 1962. Variable-Speed Transmission Combined
254-255. With Planetary Drive. US patent 3 251 243, filed
Fushimi, Y., T. Fujii and S. Kanehara 1996. A Numerical Ap- 21.05.1962, granted 17.05.1966.
proach to Analyze the Power Transmitting Mechanisms of Kress, J.H. 1968. Hydrostatic power splitting transmissions
a Metal Pushing V-Belt Type CVT: SAE paper 960720. for wheeled vehicles - classification and theory of opera-
Warrendale, PA: Society of Automotive Enginers. tion. SAE paper 680549.Warrendale, PA: Society of Auto-
Goering, C.E. et al. 2003. Off-Road Vehicle Engineering motive Engineers.
Principles. St. Joseph, MI: American Society of Agricul-
tural Engineers.

Lang, Th., A. Römer and J. Seeger 1998. Entwicklungen der N.N. 1954. Hydraulically propelled tractor. Farm Mechaniza-
Hydraulik in Traktoren und Landmaschinen (Develop- tion Vol. 6, No. 62: 224-225.
ments in hydraulics for tractors and agricultural machin- N.N. 2003. V-model - Development Standard for IT-Systems
ery) O+P Ölhydraulik und Pneumatik Vol. 42, No. 2: of the Federal Republic of Germany.
87-94. URL: http://www.V-modell.iabg.de/kurzb/vm/b-vm.doc.
Leitner, J., H. Aitzetmüller and R. Marzy 2000. Das Munich: IABG.
mechanisch-hydrostatisch leistungsverzweigte, stufenlose Nowatschin, K. et al. 2000. Multitronic - die neue
Fahrzeuggetriebe S-Matic (Mechannic-Hydrostatic Power Getriebegeneration von AUDI (multitronic - the new trans-
Split CVT S-Matic for Vehicles). Congress mission generation of AUDI). In: VDI-Berichte 1565:
Antriebstechnik Zahnradgetriebe Sept. 14-15 Dresden. 195-213. Düsseldorf: VDI-Verlag.
Proceedings: 80-93. TU Dresden. Lehrstuhl für Orshansky, E. 1972. Hydromechanical Transmissions. Engi-
Maschinenelemente und Maschinenkonstruktion. neering Report 800 (June 1972). New York: Orshansky
Linares, P. 2003. Transmissiones CVT con ramificatión Transmission Corporation. See also several US patents on
mechánica-hydrostática de la potencia (CVTs with the same subject.
mechancial-hydrostatical power split). Madrid: Escuela Pohlenz, J. and W.-D. Gruhle 2002. Stufenloses
Tecnica Superior de Ingenieros Agronomos, booklet hydrostatisch-mechanisch leistungsverzweigtes Getriebe
M-14928-2003. (Continuously variable hydro-mechanical power split
Machida, H. et al 1995. Design Principle of High Power Trac- transmision). O+P Ölhydraulik und Pneumatik Vol. 46,
tion Drive CVT. SAE paper No. 950675. Warrendale, PA: No. 3: 154-158.
Society of Automotive Engineers. Reiter, H. 1990. Verluste und Wirkungsgrade bei
Marschall, H. 1973. Antriebsvorrichtung, insbesondere für Traktorgetrieben (Energy losses and efficiencies of tractor
land- und bauwirtschaftlich genutzte Fahrzeuge (Drive transmissions). Fortschritt-Berichte VDI, Series 14, No.
concept, mainly for agricultural and construction vehicles). 46. Düsseldorf, Germany: VDI.
German patent 2 335 629, filed 13.7.1973, granted Rempfer, M. 2003. Grundlagen der automatischen
30.1.1975. Reifenluftdruckverstellung bei Traktoren (Design funda-
Martin, L.S. 1969. The development of a Digital Computer mentals for automatic tractor tyre inflation control).
Program for Analyzing the Performance of Hydrostatic Fortschritt-Berichte VDI Series 14, No. 111. Düsseldorf:
and Hydro-Mechanical Transmissions. SAE paper No. VDI-Verlag.
690566. Warrendale, PA: Society of Automotive Engi- Renault, L. 1907. Flüssigkeits-Wechselgetriebe (hydraulic
neers. gear box). German patent No. 222.301, filed 22.12.1907,
Martinus, M. and R. Freimann 2002. Process Safety Agricul- granted 24.05.1910.
tural Machinery Electronics - Implement Guided Tractor Renius, K.Th. 1968. Grundkonzeptionen der Stufengetriebe
Control. Agrartechnische Forschung Vol. 8, No. 3: moderner Ackerschlepper (Concepts of stepped transmis-
E61-E69. sions of modern tractors). Grundlagen der Landtechnik
URL: http://www.landtechnik-net.com. Vol. 18, No. 3: 97-106.
Meile 1961. Hydrostatic Drives for Tractors. ASAE paper Renius, K. Th. 1969. Stufenlose Drehzahl-Dreh-
No. 61-632. St Joseph, MI: American Society of Agricul- moment-Wandler in Ackerschleppergetrieben (Continu-
tural Engineers. ously variable units in tractor transmissions). Grundlagen
Mistry, S.I. and G.E. Sparks 2002. Infinitely Variable Trans- der Landtechnik Vol. 19, No. 4: 109-118.
mission (IVT) of John Deere 7000 TEN Series Tractors. Renius, K. Th. 1974a. Neuere Getriebekonzeptionen für
ASME International Mechanical Engineering Congress landwirtschaftliche Schlepper (Recent tractor transmission
Nov. 17-22 New Orleans. Paper IMECE2002-39347. New developments). Grundlagen der Landtechnik Vol. 24, No.
York, NY: American Society of Mechanical Engineers. 2: 41-46.
Mitsuya, H. et al 1994. Development of Hydromechanical Renius, K. Th. 1974b. Untersuchungen zur Reibung zwischen
Transmission (HMT) for Bulldozers. SAE paper No. Kolben und Zylinder bei Schrägschei-
941722. Warrendale, PA: Society of Automotive Engi- ben-Axialkolbenmaschinen (Investigation of the friction
neers. between piston and cylinder of swash plate axial piston
Mölle, R. 2003. Control and Operating Behavior of Continu- units). VDI-Forschungsheft 561: Düsseldorf, VDI-Verlag.
ously Variable Chain Transmissions. In: VDI-Beichte Renius, K. Th. 1976. Last- und Fahrgeschwindigkeits-
1798: 397-402. Düsseldorf: VDI-Verlag. kollektive als Dimensionierungsgrundlagen für die
Molly, H. 1966. Hydrostatische Fahrzeugantriebe - ihre Fahrgetriebe von Ackerschleppern (Load and vehicle
Schaltung und konstruktive Gestaltung. Teil I und II. (Hy- speed spectra fundamentals for dimensioning of tractor
drostatic vehicle drives - their control and engineering. Part drives). Fortschritt-Berichte VDI-Z, Series 1, No. 49.
I and II). ATZ Automobiltechnische Zeitschrift Vol. 68, Düsseldorf: VDI Verlag.
No. 4: 103-110 (I) and No. 10: 339-346 (II). Renius, K.Th. 1976a. European Tractor Transmission Design
Morris, W.H.M. 1967. The IHC Hydrostatic. Farm Concepts. ASAE paper No. 76-1526. St. Joseph, MI:
Mechnization and Buildings Vol. 19, No. 215:79. American Society of Agricultural Engineers.
Mortenson, P.C. 1960. Hydrostatic Transmissions. SAE Renius, K. Th. 1988. Einführung in das Forschungsprojekt
Transactions Vol. 68: 243-247. “Leiser Kleinschlepper”. (Introduction to the research pro-
Moyer, D.W. 1966. A Simple Transmission for a Deluxe Es- ject “low noise tractor”). Public presentation of the Munich
tate Tractor. SAE paper No. 660586. Warrendale, PA: So- Research Tractor 25.02.1988 TU Munich.
ciety of Automotive Engineers.

Renius, K.Th. 1993. Tractor Transmission Developments Schueller, J.K. and N. Khan 1995. Some conceptual designs
with particular Reference to Europe. Invited Special Lec- of heavy vehicle continuously-variable transmissions.
ture along with ICAMPE ´93 Seoul Oct. 19-22. Seoul: Ko- SAE paper No. 952134. Warrendale, PA: American Soci-
rean Farm Machinery & Tool Industry Cooperation ety of Automotive Engineers.
(KFMTIC). Schueller, J.K. 2002. Advanced Mechanical and Mechatronic
Renius, K.Th. 1994. Trends in Tractor Design with Particular Engineering Technologies and Their Potential Implemen-
Reference to Europe. J. of Agric. Engng. Res. Vol. 57, No. tation on Mobile Agricultural Equipment. ASAE paper No.
1: 3-22. 021064. St. Joseph, MI: American Society of Agricultural
Renius, K.Th. and B. Vahlensieck 1996. Wirkungsgrade Engineers.
stufenloser Traktor-Fahrantriebe (Efficiencies of continu- Schunder, H. 1995. Die Rexroth-Geschichte (The Rexroth
ously variable tractor drive transmissions). Landtechnik Story). Lohr a. Main/Germany: Mannesmann Rexroth
Vol. 51, No. 5: 248-249. GmbH.
Renius, K. Th. 1999. Tractors. Two Axle Tractors. In: CIGR Srnik, J. 1999. Dynamik von CVT-Keilkettengetrieben (Dy-
Handbook of Agricultural Engineering. Vol. III, Plant Pro- namics of CVT chain variators). Fortschritt-Berichte VDI
duction Engineering (edited by B.A. Stout and B. Cheze): Series 12, No. 372. Düsseldorf: VDI-Verlag.
115-184. Stöckl, B. 2002. Development of a Pressure Controlled
St Joseph MI: American Society of Agricultural Engineers. Clamping System for Continuously Varaible Belt and
Renius 2000. Trends in tractor transmission development. In- Chain transmissions. In: VDI-Berichte 1709: 571-582.
vited Special Lecture along with ICAME 2000, Seoul Nov. Düsseldorf: VDI-Verlag.
14. Seoul: The Korea Agricultural Machinery Industry Co- Sue, A., O. Dittrich and G. Poll 2002. CVT-Wirkungsgrad-
operative. berechnung aus der Verformung beim Umschlingungstrieb
Renius, K.Th. and M. Koberger 2001. Tractor Engines and (CVT efficiency modelling for pull type chain variator
Transmissions. In: Yearbook Agricultural Engineering based on deflections). In: VDI-Berichte 1709: 69-88.
Vol. 13: 43-47 and 263-264. Münster: Landwirt- Tarasinski, N. 2004. Elektrische Antriebe in Traktoren (Elec-
schaftsverlag. (E-mail: [email protected]). trical drives in tractors). In: VDI-Berichte 1855: 231-239.
Renius, K. Th. 2002. Global tractor development: Product Düsseldorf: VDI-Verlag.
families and technology levels. 30. Symposium “Actual Ulbrich, P.: 1994. Neues stufenloses hydrostatisch-mechan-
Tasks on Agricultural Engineering” Opatija, Croatia, isches Verzweigungsgetriebe für Traktoren (New hydro-
March 12-15. Proceedings: 87-95. static-mechanical power split transmission for tractors).
Renius, K.Th. 2003. Hydrostatische Fahrantriebe für mobile Meeting “Landtechnik” Stuttgart-Hohenheim Oct. 13-14.
Arbeitsmaschinen (Hydrostatic CVTs for mobile machin- Proceedings: 10.1-10.3. Düsseldorf: VDI-MEG.
ery drives). In: VDI-Berichte 1793: 65-78. Düsseldorf: Vahlensieck, B. 1999. Messung und Anwendung von
VDI-Verlag. Lastkollektiven für einen stufenlosen Kettenwand-
Resch, R. 2004. Leistungsverzweigte Mehrbereichs- ler-Traktorfahrantrieb (Measurement and application of
fahrantriebe mit Kettenwandlern. (CVTs with ranges and load spectra for a tractor chain variator CVT).
power split for chain variators). Ph.D. thesis TU Munich Fortschritt-Berichte VDI Series 12, No. 385. Düsseldorf:
2004 (in preparation for Fortschritt-Berichte VDI). VDI Verlag.
Ross, W.A. 1972. Designing a hydromechnical transmission Wadman, B. 1973. Responder automatic transmission ready
for heavy duty trucks. SAE paper No. 720725. for market. Diesel and Gas Turbine Progress Vol. 39, No.
Warrendale, PA: Society of Automotive Engineers. 6: 32-35.
Sauer, G. 1996 Grundlagen und Betriebsverhalten eines Welschof, G. 1974. Entwicklungslinien im Schlepperbau.
Zugketten-Umschlingungsgetriebes (Fundamentals and Kriterien für die heutige und zukünftige Entwicklung (De-
operating characteristics of a pull type chain variator). velopment trends in tractor engineering. Criteria for pres-
Fortschritt-Berichte VDI Series 12, No. 293. Düsseldorf: ent and future development). Grundlagen der Landtechnik.
VDI-Verlag. Vol. 24, No. 1: 6-13.
Scarlet, A.J. 1993. Integration of Tractor Engine, Transmis- Westenthanner, U. 2000. Hydrostatische Anpress- und
sion and Implement Depth Controls, Part II: Control Sys- Übersetzungsregelung für stufenlose Kettenwandler-
tems. J. of Agric. Engng. Res. Vol. 54, No.2: 89-112. getriebe (Hydrostatic clamping and ratio control for chain
Schönnenbeck, G. and P. Wagner 1992. Gesichtspunkte und variator CVTs). Fortschritt-Berichte VDI Series 12, No.
spezielle Prüfmethoden zur Schmierstoffauswahl für 442. Düsseldorf: VDI-Verlag.
stufenlose Kettenwandler (Considerations and test meth- Witt, R.H. 1965. A New Hydrostatic Drive for Small Trac-
ods for traction fluids of chain variator CVTs). Conference tors. SAE paper No. 650675. Warrendale, PA: American
“Tribologie 2000” TAE Esslingen, Germany Jan. 14-16. Society of Automotive Engineers.
Schönnenbeck, G. and P. Wagner 2002. The development of Worn, C.L.G. and A.C. Walker 1966. New Hydrostatic
Mechanical Infinitely Variable Chain/Belt Drives. In: Transmission Replaces Gearbox. SAE Journal Vol 74, No.
VDI-Berichte 1709: 1-22. Düsseldorf: VDI-Verlag. 4: 64-65. (See also SAE paper 650 689).
Schueller, J.K. and B.A. Stout 1995. Agricultural trends and Ziebart, E., H.-J. Bart and M. Meyerle 1983. Stufenlose
their effects on technological needs for farm equipment in hydrostatische Getriebesysteme für Straßenfahrzeuge (Hy-
the 21st century. Club of Bologna 6. Meeting Nov. 6-8, drostatic CVT systems for on-road vehicles). In:
1995. Proceedings: 65-76. Rome: UNACOMA. VDI-Berichte 466: 93-99. Düsseldorf: VDI-Verlag.
ASAE DISTINGUISHED LECTURE SERIES — TRACTOR DESIGN No. 29 37

03 - Continuously Variable Tractor Transmissions

Uploaded by

Copyright:

Available Formats

03 - Continuously Variable Tractor Transmissions

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

03 - Continuously Variable Tractor Transmissions

Uploaded by

Copyright:

Available Formats

This is not a peer-reviewed article.

Prof. Dr.-Ing. Dr. h.c.

For presentation at the 2005 Agricultural Equipment Technology Conference

INTRODUCTION AND HISTORY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

Karl Th. Renius

Prof. Dr.-Ing. Dr. h.c., Retired Head of Department of Agricultural Machinery

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 5

6 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 7

8 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

70 after a low number of units produced. Other compa-

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 9

10 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

Full load driveline efficiency

Tractor pulling with 90

full ballast on concrete %

Tractor pulling off-road 82

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 11

12 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

tor transmission input shafts. Probability

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 13

14 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

Traction forces within Radius of Important for

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 15

pros`s con`s comments

Steel belt drive CVTs

16 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 17

18 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 19

20 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

PME Synchronous line engine, electrical braking (charging the

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 21

22 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 23

Fig. 28 Continuously variable transmission with internal

24 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 25

8 Positive circu- Negative circu- 0

Power fixed path

Positive circu- 0.6

3 Power split lating power

-1 coupled CVT configuration

26 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

should consider efficiency benefits 0.80

5. Computer aided analysis is recom-

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 27

28 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

Full load driveline efficiency

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 29

example 95 % for 6-12 km/h), the result

30 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

Power split point

1.5) in its new series “Agrotron TTV” 0

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 31

32 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005

EXAMPLE OF A POWER SPLIT CVT CONCLUSIONS AND OUTLOOK

ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005 33

34 ASAE DISTINGUISHED LECTURE SERIES No. 29 – February 2005