Downloaded from SAE International by North Carolina State Univ, Monday, September 17, 2018

Downloaded from SAE International by North Carolina State Univ, Monday, September 17, 2018

680380

Hydrostatic Power-Splitting
Transmissions for
Wheeled Tractors —
Application and
Observed Performance

Gene Hunck
John Deere Waterloo Tractor Works, Deere & Co.

THE USE OF infinitely variable power transmission in a will describe two widely different hydrostatic power-splitting
wheeled tractor is just as desirable as for other vehicles such transmissions which were a part of that investigation, their
as the automobile. However, for tractors the ratio of power installation in test vehicles, and the resulting performance
used to that available is high and any significant loss in op- of e a c h . The two transmissions are:
erating performance is quickly recognized. 1 . Planetary power-splitting hydrostatic of the "input-
T h e tremendous a c c e p t a n c e of the hydrostatic transmis- coupled" arrangement.
sion in lawn and garden tractors the past few years indicates 2 . Hydraulic differential (rotating case) of the "output-
the value placed on "convenience of control" for small v e - coupled" arrangement.
h i c l e s . Undoubtedly, this same " control convenience" value Both of these transmissions are of the "three-element"
applies to larger utility tractors and to some lesser degree or "three-shaft" class.
to all agricultural and industrial wheeled tractors.
Hydrostatic power-splitting transmissions appear to offer PLANETARY POWER-SPLITTING
performance benefits for the larger tractors. At least t h e o - HYDROSTATIC TRANSMISSION
retical calculations indicate that by selecting the type of
power-splitting arrangement, certain portions of the total Construction of the "input-coupled" planetary transmis-
transmission operating range can be improved. How real sion is shown in Fig. 1 . The hydraulic pump and motor are
these benefits are and just what problems these hybrid units in a b a c k - t o - b a c k arrangement. The input shaft drives the
create can be fully determined only by building and testing pump barrel with a splined section and continues on through
a vehicle with such a transmission. For this purpose, our the manifold assembly and hollow motor shaft to the sun
company undertook an investigation several years ago on pinion of a planetary gearset. The ring gear of the planetary
hydrostatic power-splitting transmissions. In this paper, I attaches to the hydraulic motor shaft and the planetary c a r -

• ABSTRACT-

This paper describes two particular hydrostatic power- tractors, and resulting performance of e a c h . Peak perfor-
splitting transmission packages (an "input-coupled" plane- m a n c e of both was very close, but the "output-coupled" h y -
tary and an "output-coupled" hydraulic differential). It draulic differential transmission provided higher operating
discusses their construction, installation in experimental efficiency over a broader speed range.
 Downloaded from SAE International by North Carolina State Univ, Monday, September 17, 2018

rier is the output member. This arrangement is recognized transmitted mechanically to the drive wheels. (See Fig. 2,
as being " input coupled" because it is the input shaft that position C . ) To reduce output speed from this 4 mph lockup
couples directly to one of the hydraulic units. A variable point, the pump is taken overcenter so that it now starts
pump-variable motor is used in this installation. With the accepting oil from the motor which is being driven by the
power-splitting planetary and variable motor combination, planetary ring g e a r . The hydraulic unit's function is r e -
the total operating speed range of the vehicle appeared pos- versed in the 0 - 4 mph range in that the motor is acting as
sible without resorting to any stepping or range shifting. a pump and the pump is acting as a motor, feeding torque
The usual supercharge pump, filters, oil cooler, relief back into the input shaft. When the pump displacement
valves, e t c . , were required with this transmission, as is n e c - reaches m a x i m u m , the motor and ring gear are rotating at
essary with most hydrostatic transmissions. A mechanical a speed just sufficient to counteract the forward input speed
power take-off countershaft with its associated gear train of the sun pinion. The resulting carrier speed is zero at this
and clutch pack were also a part of the experimental p a c k - point. (See position B, Fig. 2 . ) For these swashplate p o -
a g e . These items were all included in the experimental sitions the vehicle is hydraulically locked at the zero output
unit as it was desired to be able to use the vehicle on a speed point. If parked on a steep incline, the full engine
variety of agricultural and industrial evaluation tests. horsepower would be available for braking (or at least to
The predominate usage range for most wheeled tractors traction l i m i t ) .
is 3 - 8 mph so this is the zone where the transmission was T o obtain reverse travel the motor is destroked, allowing
designed to provide its best performance. A utility type it and the ring gear to spin faster in a direction opposite to
tractor chassis was used since that particular chassis con- the input sun pinion. In the 0 - 8 mph reverse zone, the pump
figuration is most readily adapted to an in-line hydrostatic is again pumping and all power must be transmitted h y -
package. A 50 hp engine was used, but a fixed displace- draulically. T h e m e c h a n i c a l power portion is flowing from
ment implement pump in an open-center circuit removed the planetary sun into the input shaft. This is a continuation
2 hp so that the available transmission input was 48 h p . of the regenerative c y c l e as occurred in the 0 - 4 mph zone
With a 2500 rpm input, tire size and final reduction ratio but direction of power flow is reversed. (Position A of Fig.
selected, a "lockup" point or ring gear stall point is o b - 2 shows maximum reverse speed.) For speeds above 4 mph,
tained at a travel speed of 4 mph. At this point the h y - the pump is stroked to drive the motor and the ring g e a r .
draulic circuit flow is theoretically zero and all power is The ring speed added to the sun pinion input gradually re-
duces the planetary's effective reduction and the carrier
output speed increases. On the experimental unit, the pump
Input reaches its full stroke at 8 mph (position D, Fig. 2 ) . To
2#
my/////. continue into the transport range, the motor is again d e -
MV stroked until its displacement is equal to that of the pump.
PV
At this point, the planetary set rolls as a ball with its input
speed equal to t h e output (position E, Fig. 2 ) . The total
theoretical speed range of the vehicle was 8 mph reverse
to 16 mph forward. Slippage or leakage within the hydraulic

^7777777777777777 components reduces this total range slightly.

Overall operating efficiency (engine to axles) of the e x -

perimental vehicle was measured to include all of the para-
P.T.O. sitic loads that would normally occur in a wheeled tractor
transmission and final drive. Included for this experimental
Fig. 1 - "Input-coupled" planetary power-splitting hydro- installation were: a 12 gpm supercharging and lube pump
static transmission construction as used for performance tests operating at 1 8 0 - 2 0 0 psi, disc type parking or steering assist

V T
A . 8 M P H Reverse B. OMPH C. 4 M P H Forward

Fig. 2 - Swashplate positions for different travel

speeds of planetary power-splitting hydrostatic
D. 8 M P H Forward E. 1 6 M P H F o r w a r d transmission equipped experimental tractor
 Downloaded from SAE International by North Carolina State Univ, Monday, September 17, 2018

brakes in their disengaged position, a power takeoff train that draulic components at a slightly more efficient operating
was constant running with a clutch pack at the rear of the pressure. Also churning losses for all input connected m e m -
vehicle, two spur gear meshes, and one spiral bevel set. bers has been reduced. A noticeable improvement is seen
All of these parasites detracted from the tractor's perfor- in the vehicle's performance for this overload condition
m a n c e , of course, but would be quite typical for a wheeled over the reduced speed range.
tractor installation. Not shown here was the vehicle's performance in the
Fig. 3 shows the measured e n g i n e - t o - a x l e performance reverse range. As mentioned before, to improve the o p -
over the forward speed range. Peak operating efficiency erating efficiency in the forward range with the planetary
had been anticipated at the 4 mph lockup point. In the arrangement, one must, in general, detract from the reverse
vehicle assembly, the peak performance actually occurred range. In reverse, this experimental vehicle was limited
after the lockup point, at approximately 6 mph. At this and some operator complaints were received during field
point, the pump is stroking and hydraulic power flowing. evaluation tests. Suprisingly enough, it was during agri-
Performance in the 8 - 1 6 mph range falls fairly sharply. cultural evaluations and not during industrial evaluation
The hydraulic motor efficiency drops as its swashplate angle tests that these complaints were voiced. T h e major o b -
is reduced. The motor barrel windage as well as all output jections occurred while using the vehicle for towing heavily
member's windage loss also increase at the higher speeds. loaded wagons and then attempting to back the loaded units
Since peak efficiency did not occur at the 4 mph lockup up steep inclines. This type operation could be encountered
point, performance during heavy draft operation in the 2 - 4 at ramps to unloading elevators or dumps.
mph range was not as high as theoretical calculations had On industrial applications with a front end loader, drive
indicated. Refinements in the charging and control circuit wheel traction was usually the limiting factor during reverse
and in the power shaft drive would have undoubtedly i m -
proved the system's overall performance a few per cent.
A break in the system's operating efficiency can be seen
at the 4 mph point. Considerable speculation had been made
prior to building the vehicle as to how large this break would
be or if in fact it could even be detected during test. Only
with careful load and ratio adjustments could the break
shown here be measured. During regular operation of the
vehicle the operator could not detect the motor stall point.
The break in the efficiency curve was credited to the motor's
added resistance to rotation at or near its stall point, aiding f fjf I I I I I I I I
w 2 4- 6 8 10 12 14 16
the system below 4 mph where the hydraulic circuit is brak-
Speed, mph
ing, as compared to above 4 mph where the hydraulic motor
must be driven to rotate the ring in a forward direction. Fig. 4 - Tractor performance with planetary power-splitting
Fig. 4 shows the same overall ( e n g i n e - t o - a x l e ) perfor- hydrostatic transmission (engine pulled underspeed to 2000
m a n c e with the unit overloaded to pull the engine under- rpm, 4 2 . 5 hp input above 2.5 mph). Note: Performance
speed. reflects that of transmission and final drive including losses
For this condition, the engine torque (and transmission of supercharge and lube pump, idling PTO train, and service
system pressure) has increased, placing the transmission h y - brakes in disengaged position

Speed, m p h

Fig. 3 - Tractor performance with planetary power-splitting

hydrostatic transmission, 2500 rpm, 4 8 hp input above 3 mph.
Note: Performance reflects that of transmission and final
drive including loses of supercharge and lube pump, i d l i n g , Fig. 5 - Experimental tractor with planetary power-splitting
PTO train, and service brakes in disengaged position hydrostatic transmission as assembled in 1 9 6 1
 Downloaded from SAE International by North Carolina State Univ, Monday, September 17, 2018

operation. Although an excessive amount of hydraulic horse- Again either " input" or " output" coupled transmission could
power was flowing within the transmission in the reverse be used. In these units also, the shaft most directly asso-
range, the output torque was normally sufficient for loader ciated to the grounded hydraulic element defines its c o u -
work. This was particularly true during digging where r e - pling. Fig. 6 shows two methods of construction that have
verse operation required backing up an incline with a fully been used to provide "output" coupled hydraulic differential
loaded bucket. This small utility type vehicle would nor- transmissions.
mally do only stock pile or light loader work, but it was The hydraulic differential transmission as shown in Fig. 6
interesting to evaluate it also in heavy digging situations. utilizes a shell or case which carries the trunions for both
As mentioned earlier, the reason for considering a power- the pump and motor, as well as the manifolding between
splitting hydrostatic transmission in a wheeled tractor was to the two units. This type of unit has been produced in Europe
place its speed range in the most desired area and to e n - for in-plant industrial machine drives and for vehicle trans-
hance the vehicles' operating efficiency in the area of great- missions. An investigative program was undertaken by the
est usage. The experimental vehicle with the planetary Engineering Department of Deere's operation at Mannheim,
power-splitting transmission did accomplish those items, West Germany. The writer was fortunate enough to also be
even though the shape of the observed performance curve associated with that program for a short t i m e during some of
was somewhat different than theoretical prediction had in- its tests and evaluations. On this program, an experimental
dicated. Fig. 5 shows the experimental vehicle with its vehicle using a hydraulic differential of the construction
planetary power-splitting hydrostatic transmission installa- shown in Fig. 6B was assembled. Again a production t r a c -
tion. tor chassis was used with modifications made as required
to permit the installation of the experimental transmission.
HYDRAULIC DIFFERENTIAL TRANSMISSION
An engine which developed 47 hp at 2400 rpm was used in
In this type transmission the differential effect is o b - this installation. The transmission's operation can best be
tained directly with the hydrostatic transmission elements. described at fixed points in its speed range. Fig. 7 shows the
pump and motor swashplate position at the different output
speeds of the experimental v e h i c l e . These travel speeds
result from an input speed of 2400 rpm and the particular
final reduction and tire size used. It can be seen in Fig. 7
that the motor swashplate remains at maximum angle for
positions A, B and C . Output speed and direction is a func-
tion of the pump swashplate position for this range. Hy-
draulic system flow, however, is a function of both the pump
angle and output (case) speed as it is the relative velocity

V 1 between the hydraulic elements barrel and swashplate as

i
well as the angle that produces flow. As the output speed

X is increased from 0 to 3.9 mph (position C of Fig. 7 ) , the

relative speed between the pump swashplate and barrel a c -
tually decreases by 690 rpm, resulting in an effective speed
of 1710 rpm. Whereas in the reverse range, the opposite
would be true and relative velocity would increase to 3090
rpm for the same output speed change. The output range
Fig. 6 - Two construction arrangements of output-coupled between positions C and D of Fig. 7 are accomplished by
hydraulic differential transmissions reducing the motor swashplate angle. With the motor's

C. 3.9 M P H Forward D. 1 3 . 5 M P H F o r w a r d E. 1 6 . 2 M P H F o r w a r d

Fig. 7 - Swashplate positions for different travel speeds of hydraulic differential transmission equipped tractor
 Downloaded from SAE International by North Carolina State Univ, Monday, September 17, 2018

barrel anchored to ground, its effective velocity is a direct installation. The operating performance curve rises fairly
function of case or output speed. As the motor swashplate rapidly, then remains practically constant from the 4 mph
angle approaches zero, the pump's output has also been a p - to 10 mph travel speed. Performance decreases slightly to
proaching zero because the relative speed of the rotating the 1.0 speed ratio which would correspond to the 13.5 mph
case is nearing that of the input shaft or pump barrel. The point. In the overdrive range, performance decreases quite
entire unit now starts rolling as a ball at the lockup point rapidly. At maximum transport speed, corresponding to a
with the only hydraulic flow being that required to replenish 1 . 2 : 1 ratio, performance is down some 12-13<7o from the
leakage or slippage o i l . peak of the normal working range. While the transmission
For output speeds above the 1:1 ratio an overdrive can can function at higher output speeds, the 1 . 2 overdrive ratio
be accomplished in the hydraulic differential transmission appears to be the practical limit for this particular assembly.
by taking the motor swashplate overcenter. Hydraulic oil At higher speed ratios the large hydraulic power flow coupled
flow is re-established as the motor is stroking at a low angle with the churning and frictional losses of the large rotating
and high velocity, whereas the pump is at full stroke with case made performance unattractive.
low relative v e l o c i t y . An overdrive ratio of 1 . 2 : 1 was used F i g . 9 shows the tractor operating efficiency again with
on this particular experimental vehicle to achieve the trans- the hydraulic differential transmission under an engine pull-
port speed of slightly over 16 mph as shown in position E of down or overload situation. The engine torque increases
Fig. 7. in accordance with the engine's torque characteristic but
Controlling the position of the pump and motor swash- total input horsepower decreases because of the lower speed.
plates becomes more difficult on the hydraulic differential Measured input horsepower for this condition is 4 4 hp, down
transmission because they are now rotating at output speed. from the original 47 h p . The total operating range is now
If actuating servo cylinders are installed within the rotating compressed because of the lower input speed but the m a x i -
c a s e , space and size b e c o m e quite c r i t i c a l . If the actuating mum performance increases for the overload operation.
servo is separated from the rotating case, some type of r o - Again, this increase in performance is credited to the fact
tary joint capable of carrying sizeable swashplate reaction that churning losses and frictional losses are reduced at the
loads is required. With the hydraulic differential transmis- lower input speed while the hydraulic elements themselves
sion, it is also necessary to give some attention to balance are slightly more efficient at the increased pressures.
of the large rotating mass at its different swashplate posi- Reverse operation was not as critical with the hydraulic
tions. differential transmission as had been experienced with the
Overall operating efficiency (engine to a x l e ) of the e x - "input-coupled" planetary power-splitting transmission.
perimental vehicle was again measured to include all of Performance in reverse would not be as high as that shown
the parasite loads normally occurring on a similar production for the forward range, however.
tractor as well as those different items associated with the Fig. 10 shows the experimental tractor with the hydraulic
experimental transmission. Included in this tractor installa- differential transmission installation. Performance in the
tion were: a 6 gpm supercharge pump operating at 60 psi, common working speed range for this output-coupled hy-
an idling power take-off train, service brakes in their en- draulic differential transmission was higher than that r e -
gaged position, one spur gear mesh, and the spiral level r e - corded for the input-coupled planetary gear differentail unit.
duction. A direct comparison is slightly misleading, however, b e -
Fig. 8 shows measured operating efficiency of the experi-
mental tractor with the hydraulic differential transmission

Speed, m p h
Speed, mph
Fig. 9 - Tractor performance with hydraulic differential
F i g . 8 - Tractor performance with hydraulic differential transmission (engine pulled underspeed to 2100 rpm, 4 4 hp
transmission, 2400 rpm input speed, 47 hp input above 3 input above 2 . 5 mph). Note: Performance reflects that of
mph. Note: Performance reflects that of transmission and transmission and final drive including losses of supercharge
final drive including losses of supercharge and lub pump, and lube pump, idling PTO train, and service brakes in dis-
idling PTO train, and service brakes in disengaged position engaged position
 Downloaded from SAE International by North Carolina State Univ, Monday, September 17, 2018

2 . Supercharge pressure of the hydraulic differential trans-

mission required an automatic boost from 60 psi to approxi-
mately 600 psi during speed ratio changes. Measured per-
formance reported here, however, was for the steady state
or 6 0 psi level and would not reflect the higher pressure
required during ratio changes.
After taking these items into consideration to make the
comparison of the two power-splitting arrangements c o m -
patible, the peak performance points were very close. How-
ever, the "output-coupled" hydraulic differential transmis-
sion still provided a higher operating efficiency over a
broader speed range than did the "input-coupled" planetary
power-splitting transmission.

CONCLUSION

Fig. 10 - Experimental tractor with hydraulic differential Construction and control of a rotating case hydraulic dif-
transmission installation ferential transmission is more complex than that of a plane-
tary power-splitting transmission. In turn, the planetary
cause: power-splitting transmission is more complex than a straight
1 . The supercharge pump and, hence, the cooling loop or full hydrostatic transmission. In both cases, however, the
capacity of the experimental tractor with the planetary additional complexities made possible continuous operating
power-splitting transmission was much greater than that of performance areas not attainable with a simpler hydrostatic
the hydraulic differential transmission. system.

This paper is subject to revision. Statements and opinions been edited by SAE for uniform styling and format. Discussion will be printed
advanced in papers or discussion are the author's and are with the paper if it is published in SAE Transactions. For permission to publish
his responsibility, not the Society's; however, the paper has this paper in full or in part, contact the SAE Publications Division and the
authors.
Society of Automotive Engineers, Inc. 8 page booklet. Printed in U.S.A.