See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/281692226

A method of applying two-pump system in automatic transmissions for

energy conservation

Article  in  Advances in Mechanical Engineering · July 2015

DOI: 10.1177/1687814015590306

CITATIONS READS

7 3,913

3 authors, including:

Y.F. Liu
Beihang University (BUAA)
39 PUBLICATIONS   261 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Dedicate Hybrid System Configuration Synthesis View project

Automotive Transmission View project

All content following this page was uploaded by Y.F. Liu on 03 May 2016.

The user has requested enhancement of the downloaded file.

Research Article

Advances in Mechanical Engineering

2015, Vol. 7(7) 1–11
Ó The Author(s) 2015
A method of applying two-pump DOI: 10.1177/1687814015590306
aime.sagepub.com
system in automatic transmissions for
energy conservation

Peng Dong1, Yanfang Liu2, Yang Liu2 and Xiangyang Xu2

Abstract
In order to improve the hydraulic efficiency, modern automatic transmissions tend to apply electric oil pump in their
hydraulic system. The electric oil pump can support the mechanical oil pump for cooling, lubrication, and maintaining the
line pressure at low engine speeds. In addition, the start–stop function can be realized by means of the electric oil pump;
thus, the fuel consumption can be further reduced. This article proposes a method of applying two-pump system (one
electric oil pump and one mechanical oil pump) in automatic transmissions based on the forward driving simulation. A
mathematical model for calculating the transmission power loss is developed. The power loss transfers to heat which
requires oil flow for cooling and lubrication. A leakage model is developed to calculate the leakage of the hydraulic sys-
tem. In order to satisfy the flow requirement, a flow-based control strategy for the electric oil pump is developed.
Simulation results of different driving cycles show that there is a best combination of the size of electric oil pump and
the size of mechanical oil pump with respect to the optimal energy conservation. Besides, the two-pump system can also
satisfy the requirement of the start–stop function. This research is extremely valuable for the forward design of a two-
pump system in automatic transmissions with respect to energy conservation and start–stop function.

Keywords
Automatic transmission, electric oil pump, mechanical oil pump, two-pump system, forward driving simulation, energy
conservation

Date received: 20 November 2014; accepted: 24 April 2015

Academic Editor: Anand Thite

Introduction MOP and use an electric oil pump (EOP). The oil flow
of EOP can be controlled to improve the pump effi-
Conventional automatic transmissions have bigger ciency. In addition, EOP can help to realize the start–
power loss than manual transmissions. Manufactures stop function which can further improve the vehicle
of automatic transmissions have been trying to solve fuel economy.
this problem using high-efficiency components or new
structure design with less drag torque.1–6 One big part 1
Department of Mechanical Engineering, Ruhr University Bochum,
of power loss comes from the mechanical oil pump Bochum, Germany
(MOP). In order to ensure that the mechanical system 2
School of Transportation Science and Engineering, Beihang University,
has sufficient lubrication and cooling at low engine Beijing, China
speeds, the displacement volume of MOP is usually
Corresponding author:
designed big. However, it causes excessive oil flow and Yanfang Liu, School of Transportation Science and Engineering, Beihang
big power loss at high engine speeds. Therefore, mod- University, Beijing 100191, China.
ern automatic transmissions tend to reduce the size of Email: liuyf@buaa.edu.cn

Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 3.0 License
(http://www.creativecommons.org/licenses/by/3.0/) which permits any use, reproduction and distribution of the work without
further permission provided the original work is attributed as specified on the SAGE and Open Access pages (http://www.uk.sagepub.com/aboutus/
openaccess.htm).
Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015
2 Advances in Mechanical Engineering

In recent years, many new automatic transmissions
start applying EOP in their hydraulic system.7–10 The
EOP is designed for the start–stop function.
Meanwhile, it is used to support the line pressure and
supplement the oil flow for lubrication and cooling at
low engine speeds. A two-pump system with two EOPs
is designed to improve the efficiency and performance
of a six-speed automatic transmission.11 Some control
algorithms are developed for the EOP control of a two-
pump system.12–14 They try to reduce the energy con-
sumption of the two-pump system. High-efficiency
EOP and MOP have also been developed and applied
in automatic transmissions.15,16 Ahlawat et al.17 ana-
lyzed the effect of different pumps on fuel economy of
a dual-clutch transmission vehicle. Through modeling
and simulation, it is suggested to apply a two-pump
system (one EOP and one MOP) in this dual-clutch
transmission.
Some of the above studies integrate the EOP into the
existing transmission structure. Some try to reduce the
energy consumption by optimizing the control algo-
rithms or the design of the pumps. However, these stud-
ies do not involve the way to determine the size of EOP
and MOP for optimal energy conservation. The best
combination of the size of EOP and the size of MOP
with respect to energy conservation is usually difficult
for engineers to figure out.
This article develops a method to deal with this issue
using an eight-speed automatic transmission in mass
production as studying object.18,19 Figure 1 shows its
two-dimensional structure. A detailed mathematical
model is developed to calculate the power loss of the
transmission. A leakage model is developed to calculate
the leakage of the hydraulic system. The flow require-
ment can be calculated by these two models for the
two-pump system in the forward driving simulation. A
flow-based control strategy for the EOP is developed.
Different combinations of MOP and EOP result in dif-
ferent energy consumptions. They are investigated
through the simulation of various driving cycles.
Results show that there is a best combination for the
two-pump system with respect to energy conservation.
The start–stop function also has a requirement for the
two-pump system. Finally, the prototype is tested on
the vehicle. It is proved that the developed two-pump
system and the flow-based control strategy can satisfy
the flow requirement of the eight-speed automatic
transmission.
Power loss model
The power loss of an automatic transmission includes
mechanical loss (gear, bearing, seal, shifting loss, and
drag torque loss of the open shifting elements), hydrau-
lic loss (torque converter and MOP), and electric loss
(electric power demand for the solenoids, sensors,
transmission control unit, and EOP). The power loss
transfers to heat which must be dissipated by the trans-
mission housing and the oil. The necessary oil flow for
cooling is at the expense of the hydraulic loss (from
MOP and EOP) and the electric loss (from EOP). In
this section, the power loss of the eight-speed automatic
transmission is calculated. Then, the necessary oil flow
for cooling can be known based on the calculated
power loss.

TG610 TG47 PG1 TG58

C3 C1
B1
C4

3 6 4 5
1
in

10
PG3 7
8
2
final
drive 9

11

PG2 C2

out diff. out

PG: Planetary gear

TG: Transfer gear

Figure 1. Two-dimensional structure of an eight-speed automatic transmission.

Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015

Dong et al.
Gear meshing loss
As shown in Figure 1, this eight-speed automatic trans-
mission has three planetary gear sets and four transfer
gear sets. The mechanical power transferred by a plane-
tary gear set can be divided into two parts.20 One part
is known as ‘‘coupling power’’ which has no losses
because there is no conversion of torque and speed.
The other part is called ‘‘rolling power’’ which is a
result of relative speed between the central gears (sun
gear and ring gear) and the carrier. This power gener-
ates losses in the rolling teeth and the planetary bear-
ing. The calculation of rolling power for each part of a
planetary gear set is expressed by equations (1)–(3)
PRS = TS  ðvS  vC Þ
PRR = TR  ðvR  vC Þ
PRC = TC  ðvC  vC Þ = 0
The sum of these three powers is just the loss of a
planetary gear set (4)
Ploss gear = PRS + PRR + PRC
where TS, TR, and TC are the torques of the sun gear,
ring gear, and carrier, respectively; vS, vR, and vC are
the angular speeds of the sun gear, ring gear, and car-
rier, respectively; and PRS, PRR, and PRC are the rolling
powers of the sun gear, ring gear, and carrier, respec-
tively. Equations (1)–(3) also apply to a transfer gear
set. TS and TR can be replaced by the torques of the
pinion and the wheel. The carrier can be assumed to be
the housing with its angular speed vC being 0. The loss
of the planetary gear set is mainly a torque loss. It can
be calculated based on the kinetics of the planetary gear
set (5)
TR + TS  i0  hw0 = 0
where i0 is the standard gear ratio, h0 is the efficiency
factor of a planetary gear set with motionless carrier,
and w defines the direction of the power flow. The effi-
ciency factor h0 includes two parts according to equa-
tion (6)
Y Y
h0 = hV  hL
where hV is the gear meshing efficiency and hL is the
planetary bearing efficiency. For a transfer gear set,
there is no need to take planetary bearing efficiency into
consideration. According to VDI Richtlinie 2157,20 the
gear meshing efficiency is assumed to be 99.0% for an
external gear pair and 99.5% for an internal gear pair.
The planetary bearing efficiency is assumed to be
99.5%.
Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015
3
Bearing loss
There are three types of bearings in this eight-speed
automatic transmission: ball bearings, tapered roller
bearings, and needle roller bearings. The friction torque
of a bearing includes two parts.21 One part is a function
of speed as shown in equation (7). The other part is a
function of load as shown in equation (8)

T0 = f0  ðn  nÞ2=3 dM
3
 107 if ðn  nÞ  2000 ð7Þ
3 7
f0  160  dM  10 if ðn  nÞ\2000
where T0 is the friction torque as a function of speed, f0
is the bearing factor for friction torque as a function of
speed, n is the kinematic viscosity of the oil at a specific
ð1Þ operating temperature, n is the operating speed, and dM
is the mean bearing diameter
ð2Þ

f1  P1  dM if ball tapered roller bearings
ð3Þ T1 = ð8Þ
f1  F  dM if needle roller bearings
where T1 is the friction torque as a function of load, f1
is the bearing factor for friction torque as a function of
ð4Þ load, P1 is the decisive load for friction torque, and F is
the load for bearings. The total power loss of a bearing
is calculated according to equation (9)
Ploss bearing = ðT0 + T1 Þ  2  p  n ð9Þ
Seal loss
The seals of this eight-speed automatic transmission
include the radial shaft seal rings and the o-rings in the
shifting elements. For the o-rings, there is an axial fric-
tion force when the clutch piston moves in the axial
direction besides the friction force in the circumferen-
tial direction. However, this axial friction force only
exists in the filling phase of the shifting process. It is
ð5Þ
ignored in the power loss calculation because of the
short action time of the piston. The calculation of seal
loss is expressed by equation (10)22
Ploss seal = 7:69  106  dsh
2
n ð10Þ
where dsh is the shaft diameter.
ð6Þ
Drag torque loss of the open shifting elements
Table 1 shows the shifting logic of the eight-speed
automatic transmission. This eight-speed automatic
transmission has five shifting elements which are all
multi-plate clutches or brakes. In each gear, three shift-
ing elements are locked. Only two open shifting ele-
ments create drag torque losses due to the relative
motion between the inner and outer plates.
4 Advances in Mechanical Engineering

Table 1. Shifting logic of the eight-speed automatic plate, ri is the inner radius of the friction plate, m is the
transmission. oil dynamic viscosity, nu is the oil angular shear velo-
city, v is the relative angular speed difference, h is the
Gear Engaged shifting elements
oil film thickness, and z is the unit interval of the plate
B1 C1 C2 C3 C4 gap.
Owing to the increasing relative speed, the centrifu-
Reverse   
gal force overcomes the shear force and begins to drain
1   
2    the oil from the gap in region 2. The laminar flow
3    breaks up, and the air is mixed in the gap. Therefore,
4    the drag torque begins to descend. In region 3, the oil is
5    completely drained by the increasing centrifugal force,
6   
and the drag torque drops to a negligible level. When
7   
8    the relative speed continues increasing and enters into
region 4, the wobbling of the plates will cause the
mechanical contact; thus, the drag torque loss increases
again. However, region 4 is avoided in the design of the
eight-speed automatic transmission. The drag torque is
Rigion 1 Rigion 2 Rigion 3 Rigion 4
always calculated based on equation (11) even in region
2 and region 3. This can provide a safety factor for the
calculation of the oil flow requirement. The power loss
of the open shifting element is, thus, calculated accord-
ing to equation (12)
Drag torque

Ploss drag = Tdrag  v ð12Þ

Torque converter loss

In the control strategy of the torque converter, the
Oil-air
Full oil mixture Full air Wobbling
lock-up clutch is always open in the first gear and the
reverse gear. In other gears, the pump and the turbine
Relative speed are mechanically connected. A big hydraulic loss
appears when the torque converter works under the
Figure 2. Drag torque profile of the open multi-plate shifting hydraulic condition. This power loss is also considered
element. in the power loss model and is calculated based on
equation (13)
As shown in Figure 2, the drag torque loss depends
on the relative speed between the two halves of the Ploss TC = TP  vP  TT  vT = TP  ðvP  mTC  vT Þ
shifting element. = l  rATF  v2P  D5TC  ðvP  mTC  vT Þ
In region 1, the drag torque increases linearly with ð13Þ
the relative speed. It is caused by the laminar shear of
the oil in the gap between the plates. Assuming that the where TP is the pump torque, TT is the turbine torque,
oil is Newtonian and the laminar velocity is in the cir- vP is the pump angular speed, vT is the turbine angular
cumferential direction, the drag torque in this region speed, mTC is the torque ratio coefficient, l is the per-
then can be expressed by equation (11)23 formance coefficient, rATF is the oil density, and DTC is
the profile diameter of torque converter.
R R Rro 
2p 
Tdrag = N  t  r  dA = N  m  dydzu  r  drdu
A 0 ri Pump loss
Rro vr 2
’ 2  N  p  m  h  r  dr The pump provides the circulated oil flow for the auto-
 4 ri 4  matic transmission. It also consumes the engine power
= N pvm
2h  ro  ri and generates the heat loss. This loss is calculated by
ð11Þ equation (14)
 
where N is the number of friction facings, t is the vis- 1  hpump
cous torque per unit area, A is the friction area, r is the Ploss pump = pline  qpump  ð14Þ
friction radius, ro is the outer radius of the friction hpump

Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015

Dong et al.
Figure 3. Power loss of the eight-speed automatic transmission in different gears at 90 °C.
where pline is the hydraulic line pressure, qpump is the
oil flow provided by the pump, and hpump is the pump
efficiency. For the MOP, this efficiency includes the
volumetric efficiency and the mechanical efficiency.
For the EOP, this efficiency also includes the electric
efficiency.
All the above losses are torque or speed dependent.
The sum of them of different gears at 90 °C is shown in
Figure 3. It can be seen that there exists big power loss
at the first gear and the reverse gear because the torque
converter works under the hydraulic condition. The
normal operating temperature of the automatic trans-
mission fluid (ATF) is from 80 °C to 120 °C. When the
temperature drops, the loss will increase because the
viscosity of the ATF becomes bigger. However, there is
still one big loss not shown in Figure 3. This loss occurs
not in a specific gear but in the shifting process. It is
the shifting loss which is caused by the friction slip of
the shifting elements.
Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015
5
Shifting loss
All shifts of this eight-speed automatic transmission are
clutch-to-clutch shifts. For the engaging and disenga-
ging shifting elements, there is a pressure-controlled
slipping process during the gear shifting. There are
four basic shifting types in automatic transmissions,
namely power-on upshift, power-on downshift, power-
off upshift, and power-off downshift. Their differences
and control logic are explained in detail in Dong et
al.24 In the shifting process of power-on upshift and
power-off downshift, most of the power loss occurs at
the on-coming shifting element. On the other hand, in
the shifting process of power-on downshift and power-
off upshift, most of the power loss occurs at the off-
going shifting element. The power loss of the actuated
shifting element is calculated by equation (15)
Ploss shift = md  ðpSE  pkiss Þ  A  r  N  signðDvÞ  Dv
ð15Þ
6 Advances in Mechanical Engineering

150
Vehicle speed Oil flow control valve
Driving pedal position
Vehicle speed [km/h]

Pedal position
100 0.4

50 0.2

0 0
110 130 150 170 190
Cooling
Time [s]

Figure 5. Hydraulic schematic diagram of the eight-speed

20 8 automatic transmission.
Power loss
Gear 7
15 6
(proportional solenoid valves and slide valves) have
Power [kW]

5
leakages. The leakage oil flow must be compensated by
Gear

10 4 the two-pump system. There are two kinds of leakage

5 2
0 0
110 130 150 170
Time [s]
Figure 4. Simulation results of the NEDC.
where md is the dynamic friction coefficient, pSE is the
pressure of the shifting element, pkiss is the kiss-point
pressure which compensates for the return spring and
seal friction force, and Dv is the angular speed differ-
ence. The shifting loss can be taken into account for
each shift in the forward driving simulation. It is usually
greater than other power losses. The heat occurs in such
a short time and cannot be dissipated by the oil within
a short period of time. The friction materials should be
able to bear such a high temperature. There are slots on
the plate surface to let the oil go through when the shift-
ing element is locked. Figure 4 shows one part of the
New European Driving Cycle (NEDC) and its calcu-
lated power loss in the forward driving simulation. It
can be seen that the peak points of the power loss curve
are just caused by the gear shifting at a certain driving
pedal position.
Leakage model
Figure 5 shows the hydraulic schematic diagram of the
eight-speed automatic transmission. The detailed
description of its functionality is introduced in Wang
et al.25 In this hydraulic system, all the valves
3 in the hydraulic system. One considers the leakage of
the eccentric ring gap, as shown in Figure 6(a). The
leakage between the spool and the bore in the slide
1
valve is just like this form. The other is called the orifice
190 leakage, as shown in Figure 6(b). The leakage of the
proportional solenoid valve is considered in this form.
The calculation of the leakage in an eccentric ring
gap is expressed by equation (16)26
ð d1 + d2 Þ  3 !
2  p  d1 d
2
2
3 e2
QL1 =  Dp  1+   
12  n  l  rATF 2 d1 d2 2
2
ð16Þ
where d1 is the inner diameter of the bore, d2 is the outer
diameter of the spool, l is the overlap length between
bore and spool, Dp is the pressure difference, and e is
the eccentric ring gap. The orifice leakage in the propor-
tional solenoid valve is calculated by equation (17)26
sffiffiffiffiffiffiffiffiffiffiffiffi
2  Dp
QL2 = aD  A0  ð17Þ
rATF
where aD is the flow coefficient and A0 is the cross-
sectional area of the orifice.
It is known from equation (16) that the leakage of
an eccentric ring gap is much influenced by the tem-
perature. With the temperature rising, the kinematic
viscosity of the oil n will decrease; thus, the leakage will
increase. However, the temperature has a small impact
on the orifice leakage. In different transmissions, the
leakage is different owing to different tolerances. The
calculation here assumes that the transmission has the
biggest tolerance; thus, the eccentric e has the biggest
value equivalent to (d12 d2)/2. The leakage also varies

Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015

Dong et al.
(a)
Figure 6. Two types of leakage in the hydraulic system (a) Leakage of the eccentric ring gap and (b) leakage of the orifice.
with the line pressure and the control pressure of the
hydraulic system. When the pressure difference Dp
becomes greater, the ring gap leakage and the orifice
leakage will increase. The line pressure is controlled
according to the engine torque in the actual driving.
But in the leakage calculation, it is always assumed to
be the maximum value to ensure that the calculated oil
flow requirement is sufficient.
In summary, the total flow requirement for the leak-
age is calculated by equation (18)
X X
QLeakage = QL1 + QL2
Flow-based control strategy for the EOP
In order to replace the big MOP with a two-pump sys-
tem (one EOP and one MOP), a flow-based control
strategy is developed. This flow-based control strategy
just makes use of the information of the calculated flow
requirement. No more additional energy is required for
the EOP on the condition that the flow from MOP is
sufficient for the transmission. Hence, this control
strategy can reduce the load of the EOP to a maximum
extent. First, based on the power loss model, the
required oil flow for the cooling and lubrication is cal-
culated by equation (19)

L Ptotal loss  Phou sin g
Qcooling = min 25 , ð19Þ
min cATF  rATF  DT
where Ptotal_loss is the total heat loss generated, Phousing
is the heat dissipated through the housing, cATF is the
heat capacity of the ATF, and DT is the permissible
temperature increase in the ATF. In equation (19), the
total heat loss generated is defined as
X X
Ptotal loss = Ploss gear + Ploss bearing
X X
+ Ploss seal + Ploss drag
X
+ Ploss shift + Ploss TC + Ploss
The heat dissipated through the housing is calculated
by equation (21)
Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015
7
QL
l
d1
d2
(b)
Phou sin g = k  Aca  ðToil  T‘ Þ ð21Þ
where k is the heat coefficient which includes the inter-
nal heat transfer between oil and housing, the heat con-
duction through the housing wall, and the external heat
that dissipates to the environment. The calculation of
this coefficient is described detailedly in ISO/TR
14179-2:2001.22Aca is the area of the housing, Toil is the
temperature of the ATF, and TN is the ambient
temperature.
The maximum oil flow for the lubrication and cool-
ing system is limited to 25 L/min by an oil flow control
ð18Þ
valve (OF_CV) in Figure 5. The reason is that the oil
flow increases along with the engine speed. The pres-
sure and the oil flow will become very large at high
engine speed without this limitation. Therefore, the
flow requirement for the necessary cooling oil is limited
to maximum 25 L/min by the OF_CV valve.
The total flow requirement should also consider the
leakage mentioned in section ‘‘Leakage model.’’ Hence,
the total flow calculation is expressed by equation (22)
 
Qtotal = max QTC , Qline , Qcooling + QLeakage ð22Þ
where QTC is the minimum flow requirement of the
hydraulic torque converter and Qline is the minimum
flow demand for maintaining the hydraulic line pres-
sure. Since the MOP is connected to the pump of the
torque converter, the oil flow provided by the MOP
depends on the engine speed. The necessary flow
required from EOP is obtained by subtracting this oil
flow from the total flow requirement. Then, the speed
of the EOP can be controlled according to equation
(23)
QEOP Qtotal  QMOP
nEOP = =
VEOP  hEOP vol VEOP  hEOP vol
ð23Þ
Qtotal  npump  VMOP  hMOP vol
=
ð20Þ VEOP  hEOP vol
pump where QEOP and QMOP are the oil flows from the EOP
and MOP, respectively; VEOP and VMOP are the displa-
cement volumes of the EOP and MOP, respectively;
hEOP_vol and hMOP_vol are the volumetric efficiencies of
8 Advances in Mechanical Engineering
30
Oil flow [L/min]
20
10
0 − 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
0 20 40
Time [s]
Total flow requirement
Flow of MOP
Necessary flow of EOP
Oil flow of EOP with smoothing
Figure 7. Simulation results of the oil flow in a simple driving
cycle.
the EOP and MOP, respectively; and hpump is the rota-
tion speed of the MOP.
This flow-based control strategy is verified through
the forward driving simulation. For a given displace-
ment of the MOP, Figure 7 shows the simulation results
of the total flow requirement, the oil flow provided by
the MOP, and the necessary flow from the EOP in a
driving simulation. It can be seen that the EOP provides
necessary oil flow when the MOP cannot satisfy the
total flow requirement. When the MOP can satisfy the
system flow requirement alone, the EOP is switched off.
However, the fluctuation of the EOP flow require-
ment will make the target control speed of the EOP
oscillate when applying this control strategy in the
actual vehicle driving. Especially when shift occurs, the
oil flow requirement will increase dramatically. The
rotation speed of EOP cannot react quickly enough to
follow the system flow requirement. In reality, the heat
generated in such a short time is dissipated gradually.
Hence a first-order time delay function is applied here
to make the EOP speed change steadily. As shown in
Figure 7, this function helps to smooth the peak points
of the oil flow in actual application.
This flow-based control strategy can help us to select
the most economical combination of MOP and EOP
through the forward driving simulation. For a two-
pump system, when a small size of MOP is selected, the
energy loss of the MOP especially at high engine speed
will be reduced. However, for the compensation for the
oil flow at low engine speed, a big EOP must be
selected. Besides, the EOP needs to work for a long time
Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015
US- city cycle Only current
US- highway cycle MOP
1000 US-race cycle
Total energy consumption [kJ]
NEDC cycle
Only one Best solution
EOP for most cycles
500
60
Displacement of MOP [mL]
Figure 8. Energy consumption with different combinations of
MOP and EOP in different driving cycles.
which increases the electric load of the vehicle. The total
energy consumption thus will not be improved. On the
contrary, a big MOP can reduce the electric load of the
EOP, but it consumes a lot of energy with the engine
speed increasing. Therefore, a best combination of EOP
and MOP needs to be investigated with respect to the
reduction in the total energy consumption.
The total energy consumption of EOP and MOP is
expressed by equation (24)
Rt2
Etotal = ðPEOP + PMOP Þ  t  dt
t1
ð24Þ
Rt2 pline ðtÞQEOP ðtÞ 
= hEOP + pline ðthÞQMOP ðtÞ  t  dt
MOP
t1
where PEOP and PMOP are the power demands of EOP
and MOP, respectively, and hEOP and hMOP are the
total efficiency of EOP and MOP, respectively.
First, different sizes of MOP are selected in the simu-
lation. According to the flow-based control strategy,
the insufficient oil flow will be supplied by the EOP;
thus, its power depends on the displacement of the
selected MOP. Figure 8 shows the simulation results of
energy consumption with different combinations of
EOP and MOP in different driving cycles. Because each
driving cycle has a different driving distance and speed-
time profile, the total energy consumption for a specific
displacement of MOP varies in different driving cycles.
However, there is a minimum energy consumption
point in each curve. It can be seen in the US race cycle
that the minimum point appears at the 4-mL displace-
ment of MOP. In other cycles, the minimum energy
consumption point locates at the 7-mL displacement of
MOP. The reason is that compared with other cycles,
the US race cycle has a higher acceleration and driving
speed. The engine normally runs in the high speed
Dong et al. 9

100%

80%

60%

40%

20%

0%
US-city cycle US-highway cycle US-race cycle NEDC cycle
Only one MOP Only one EOP A 7ml displacement of MOP together with an EOP

Figure 9. Comparison of energy consumption in different driving cycles.

400

330
300
Power [W]

200

100

0
0 250 500 750 1000 1250

Time [s]

Figure 10. Power demand of the EOP in the NEDC.

range. A big size of MOP will consume much energy.

Therefore, the US race cycle tends to have a smaller Table 2. Maximum power demand of EOP.
MOP with respect to energy conservation.
In all driving cycles, only one big MOP in current trans- Driving cycle Maximum power
demand of EOP (W)
mission has the biggest energy loss. Although applying
only one EOP can reduce this energy loss, it is not the best US city cycle 402
choice as shown in Figure 8. In most of the driving cycles, US highway cycle 306
especially when driving in cities, the simulation results US race cycle 457
show that a 7-mL displacement of MOP contributes most NEDC 330
to the energy conservation. Even in the US highway cycle, EOP: electric oil pump; NEDC: New European Driving Cycle.
a 7-mL displacement of MOP is also the best choice for
the eight-speed automatic transmission. Therefore, a 7-mL
displacement of MOP is selected in the new designed two- all the driving conditions, a peak power of 600 W EOP
pump system. Compared with only one MOP in the cur- is selected in the design. However, the EOP will not run
rent design, Figure 9 shows the percentage of the energy at the operating points of peak power in most
consumption in each driving cycle when applying only one situations.
EOP or a two-pump system with a 7-mL displacement of
MOP.
The power demand of the EOP PEOP corresponding Test verification
to a certain displacement of MOP is also calculated in Driving simulation results give us an optimal combina-
the forward driving simulation. It can give us a refer- tion of EOP and MOP with respect to energy conserva-
ence for the selection of the EOP power rating. For tion. Based on it, the new two-pump system is designed
example, Figure 10 shows the power demand of the and applied in the eight-speed automatic transmission.
EOP in the NEDC corresponding to the selected 7-mL Figure 11 shows the prototype with the novel two-
displacement of MOP. In this driving cycle, the maxi- pump system. The EOP is mounted outside the trans-
mum power demand of EOP is 330 W. mission housing. The downsized MOP is located in the
Table 2 depicts the maximum power demand of original position behind the torque converter.
EOP in some driving cycles for the selected 7-mL dis- The developed flow-based control strategy for EOP
placement of MOP. In order to satisfy the demand of and the novel two-pump system are tested on the

Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015

10 Advances in Mechanical Engineering

150 100

EOP 90

Oil temperature [ ̊С ]
Vehicle speed [km/h]
100 80

70

50 60

50

0 40
0 400 800 1200
Time [s]
Equivalent vehicle speed
Oil temperature

Figure 13. Oil temperature in the test of the NEDC.

Figure 11. Prototype of the eight-speed automatic

transmission with the novel two-pump system.

Figure 14. Test rig.

Figure 12. Test vehicle.
vehicle. Figure 12 shows the test vehicle equipped with
the prototype. The test data are timely read by the
computer through the calibration software CANape.
Figure 13 depicts the measured oil temperature in
the NEDC. The ATF temperature rises because of high
accelerations and frequent shifts. However, the two-
pump system supplies sufficient oil flow for cooling and
lubrication under the flow-based control strategy. The
oil temperature can be maintained at a normal level
between 80 °C and 90 °C during driving. In addition,
the durability test on the test rig (Figure 14) has also
proved that no parts fail due to the heat stress. This
two-pump system and the flow-based control strategy
can satisfy the oil flow requirement.
Conclusion
A power loss model of an eight-speed automatic trans-
mission is developed to calculate all power losses
depending on the engine torque and speed. Based on
the power loss model, the required oil flow for cooling
and lubrication is timely calculated in the driving simu-
lation. Together with the compensation of the oil
flow for leakage, the total flow requirement for the
two-pump system is determined. A flow-based control
strategy is applied for the EOP control. The driving
simulation results show that a two-pump system is bet-
ter than only one EOP or only one MOP with respect
to energy conservation. A best combination for the size
of MOP and EOP can be found through the driving
simulation. This best combination can not only reduce
the energy loss but also satisfy the power demand of
the start–stop function. Test results verified that the

Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015

 Dong et al.
two-pump system can supply enough oil flow for the
eight-speed automatic transmission. This method can
be applied for the forward design of a two-pump sys-
tem in automatic transmissions. It proves that a good
combination of MOP and EOP can also improve the
energy conservation of automatic transmissions, and
this combination can be found through the developed
method.
Declaration of conflicting interests
The authors declare that there is no conflict of interests
regarding the publication of this article.
Funding
This work was supported by the National Science Foundation
of China (grant no. 51105017) and the National Science and
Technology Support Program of China (grant no.
2011BAG11B00).
References
1. Katou N, Taniguchi T, Tsukamoto K, et al. AISIN AW
new six-speed automatic transmission for FWD vehicles.
SAE technical paper 2004-01-0651, 2004.
2. Uozumi S, Taniguchi T, Tsukamoto K, et al. AISIN AW
new six-speed automatic transmission for RWD vehicles.
SAE technical paper 2004-01-0652, 2004.
3. Kasuya S, Taniguchi T, Tsukamoto K, et al. AISIN AW
new high torque capacity six-speed automatic transmis-
sion for FWD vehicles. SAE technical paper
2005-01-1020, 2005.
4. Ota H, Nozaki K, Honda A, et al. Toyota’s world first
8-speed automatic transmission for passenger cars. SAE
technical paper 2007-01-1101, 2007.
5. Scherer H, Bek M and Kilian S. ZF new 8-speed auto-
matic transmission 8HP70-basic design and hybridiza-
tion. SAE technical paper 2009-01-0510, 2009.
6. Aoki T, Kato H, Kato N, et al. The world’s first trans-
verse 8-speed automatic transmission. SAE technical
paper 2013-01-1274, 2013.
7. Kobayashi T. New generation of 6 and 8 speed trans-
verse automatic transmission. In: Proceedings of the 12th
international CTI symposium, automotive transmissions,
HEV and EV Drives, Berlin, 2–5 December 2013, vol. 1,
pp.141–147.
8. Dörr C, Homm M and Indlekofer G. The new automatic
transmission 9G-Tronic from Mercedes-Benz. In: Pro-
ceedings of the 12th international CTI symposium automo-
tive transmissions, HEV and EV drives, Berlin, 2–5
December 2013, vol. 1, pp.153–160.
9. Tomomatsu H, Tanaka Y, Nakatani K, et al. Automatic
transmission control system developed for Toyota mild
hybrid system (THS-M). SAE technical paper
2002-01-1253, 2002.
10. Shunsuke Y, Toshitaka I, Koichi F, et al. Development
of the V6 Accord Hybrid five-speed automatic transmis-
sion system. Honda R&D Tech Rev 2005; 17: 14–19.
Downloaded from ade.sagepub.com at BEIHANG UNIV LIB on September 10, 2015
View publication stats
11
11. Hwang J, Jo S, Wi T, et al. Development of two oil
pumping system for automatic transmission. SAE techni-
cal paper 2014-01-1766, 2014.
12. Kim Y, Lee J, Jo C, et al. Development and control of
an electric oil pump for automatic transmission-based
hybrid electric vehicle. IEEE Trans Veh Technol 2011; 60:
1981–1989.
13. Kim Y, Song M, Kim J, et al. Power-based control of an
electric oil pump for an automatic-transmission-based
hybrid electric vehicle. Proc IME D: J Automobile Eng
2012; 226: 1088–1099.
14. Song M, Oh J, Kim J, et al. Development of an electric
oil pump control algorithm for an automatic-transmis-
sion-based hybrid electric vehicle considering the gear
shift characteristics. Proc IME D: J Automobile Eng 2014;
228: 21–36.
15. Miyachi E, Ishiguro M and Mizumoto K. Development
of electric oil pump. SAE technical paper 2006-01-1595,
2006.
16. Toyoda F, Kobayashi Y, Miura Y, et al. Development of
variable discharge oil pump. SAE technical paper
2008-01-0087, 2008.
17. Ahlawat R, Fathy HK, Lee B, et al. Modelling and simu-
lation of a dual-clutch transmission vehicle to analyse the
effect of pump selection on fuel economy. Vehicle Syst
Dyn 2010; 48: 851–868.
18. Tenberge P and Müller J. 8-Gang-Automatikgetriebe,
AT-Getriebestrukturen mit Planetenradstufen und Stirn-
radstufen. In: Proceedings of the 6th international CTI
symposium: innovative automotive transmissions, Berlin,
3–7 December 2007. Duesseldorf: Car Training Institute
(CTI).
19. Tenberge P, Gläser K, Xiang-wu Liu, et al. 8-speed auto-
matic transmission. In: Proceedings of 4th international
CTI symposium: innovative automotive transmissions,
Detroit, MI, 8–9 June 2010. Duesseldorf: Car Training
Institute (CTI).
20. VDI Richtlinie 2157. Planetengetriebe—Begriffe, Sym-
bole, Berechnungsgrundlagen. Berlin: Beuth Verlag, 2010.
21. Schaeffler. Technical pocket guide. 1st ed. Herzogenaur-
ach: Schaeffler Technologies GmbH & Co.KG, 2014,
pp. 506–507.
22. ISO/TR 14179-2:2001. Gears-thermal capacity—part 2:
thermal load-carrying capacity (1st ed).
23. Design practices: passenger car automatic transmissions.
4th ed. Warrendale, PA: SAE International.
24. Dong P, Tenberge P, Qu W, et al. Optimized shift-control
in automatic transmissions with respect to efficiency, shift
loads and comfort. In: VDI-Berichte 2158, drivetrain for
vehicles, Friedrichshafen, 19–20 June 2012, pp.167–183.
Duesseldorf: VDI Wissensforum GmbH.
25. Wang S-h, Xu X-y, Liu Y-f, et al. Design and dynamic
simulation of hydraulic system of a new automatic trans-
mission. J Cent South Univ Technol 2009; 16: 697–701.
26. Will D and Gebhardt N. Hydraulik: Grundlagen, Kom-
ponnenten, Schaltungen. 4th ed. Berlin, Heidelberg:
Springer-Verlag, 2008.