Special Issue Article

Advances in Mechanical Engineering

2017, Vol. 9(8) 1–21
Ó The Author(s) 2017
Modeling and simulation study of a DOI: 10.1177/1687814017720404
journals.sagepub.com/home/ade
novel electromechanical clutch
actuation system

Fengyu Liu1,2, Li Chen1,2, Dongxu Li3, Jianlong Zhang2 and

Chengliang Yin2

Abstract
A novel motor-driven clutch actuator with force-aided lever has been recently developed. A unique spring is used to assist
apply of the clutch that can effectively reduce the motor power requirement, and the actuation system can save power
consumption during clutch operations with faster response and higher control precision. In this article, a mathematical
model of the conceptual clutch actuation system is developed to study the dynamic behavior of the actuator system via
computer simulation. The actuator model is further integrated into a dual-clutch transmission and driveline system to
demonstrate the effectiveness of the concept in shifting control. The simulation result demonstrates the smooth and fast
clutch-to-clutch gear shift and hence validates the feasibility of this new force-aided lever actuation system.

Keywords
Motor driven, force-aided, actuation system, dry clutch

Date received: 10 February 2017; accepted: 17 June 2017

Academic Editor: Aditya Sharma

Introduction well as compatibility with intelligent controls.4,12

However, the drawback of these systems lies in the
The automotive transmission is currently edging toward lower force/torque density, so the force/torque capacity
improving powertrain system efficiency and driving is difficult to meet the design requirement and the
comfort. Smooth gear shifting control is critical for actuator needs to grow in size, resulting in packaging
overall drive quality.1–3 Clutch actuation is a key com- infeasibility. To address this issue, various force-aided
ponent in the clutch control system,4 and depending on
the transmission type and clutch actuation needs, clutch
control can be executed with different actuation
mechanisms, particularly for automated manual trans- 1
State Key Laboratory of Ocean Engineering, Collaborative Innovation
mission (AMT),2,5 dual-clutch transmission (DCT),3,6 Center for Advanced Ship and Deep-Sea Exploration, Shanghai Jiao Tong
and various hybrid transmission.7,8 The commonly used 2
University, Shanghai, P.R. China
systems include electrohydraulic and electromechanical National Lab of Automotive Electronics and Control, Shanghai Jiao Tong
University, Shanghai, P.R. China
actuation systems,6,9 where the latter has a greater 3
Global Research and Development Center, General Motors Company,
potential for cost-saving and performance improve- Pontiac, MI, USA
ment.5,10,11 Due to its high mechanical efficiency and
control precision, electromechanical systems may Corresponding author:
Li Chen, State Key Laboratory of Ocean Engineering, Collaborative
become the mainstream in the future.10–12 Compared Innovation Center for Advanced Ship and Deep-Sea Exploration,
with electrohydraulic systems, electromechanical sys- Shanghai Jiao Tong University, Shanghai 200240, P.R. China.
tems have faster response, higher energy efficiency, as Email: [email protected]

Creative Commons CC-BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License
(http://www.creativecommons.org/licenses/by/4.0/) 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 (https://us.sagepub.com/en-us/nam/
open-access-at-sage).
2 Advances in Mechanical Engineering

mechanisms have been designed to increase the actua- section ‘‘System analysis and modeling,’’ the dynamic
tion force while keeping the package compact. model of the actuation system is established. A
Wheals et al.13,14 proposed a force-aided motor-dri- position-based closed-loop control algorithm is
ven actuator by storing and releasing energy using per- described in section ‘‘Controller design.’’ In section
manent magnet for AMTs and DCTs. Unfortunately, ‘‘Results and discussion,’’ the force-aided feature of
the mechanism also provides the same aiding force even new actuation system is studied and the function is vali-
when the clutch is fully open when zero normal force is dated through simulation. The simulation results
required. Therefore, the actuation motor has to gener- demonstrate the performance of the actuator system
ate a force to counter the undesired force from magnet during clutch engagement. Besides, the force-aided
at the open position, which will draw additional energy effect, system dynamic characteristics, and controllabil-
consumption. Yao et al.4,15–17 designed a wedge-based ity are discussed. With the advantages of the fast
clutch for automatic transmission, aiming to utilizing response, small actuation force, and accurate control,
the self-reinforcement property from the wedge feature. the new actuator system has great potential for practi-
Kim and Choi10,18 designed a wedge mechanism for an cal applications. Finally, the conclusion is drawn in sec-
AMT clutch. The intractable issue of the wedge tion ‘‘Summary.’’
mechanism lies in the self-weakened effect, which
occurs when the clutch slipping speed is reversed. This
effect can greatly restrict the application of the wedge Design of force-aided lever actuator
mechanism. Actuator principle
Preloaded spring, viewed as an energy storage com-
ponent, is adopted in several force-aided solutions. One In a traditional manual or DCT transmission, the
mechanism proposed by Jianwu Zhang releases the pre- clutch is connected to the flywheel and input shaft of
loaded spring during clutch engagement assisting an transmission. The clutches can be wet or dry.28 Figure
actuation motor.19,20 However, such design is not able 1 illustrates a normally open clutch. When no force is
to reduce the motor size as the aiding force decreases applied on the piston, the clutch remains disengaged.28
rapidly with the engagement travel, where bigger actua- In the assembly in Figure 1, a clutch pack includes
tion force is required. Another design involves a pre- pressure plate, friction plate, diaphragm, piston, and
loaded spring acting on a lever with a moving actuation lever. The hydraulic system generates an
pivot,21–24 which has been used on a six-speed dry actuation force Fm to push the lever down and causes
DCT. The experimental results show that the actuation the lever to rotate clockwise about the pivot. As a con-
system has a fast response.25 In general, the controll- sequence, the piston moves down and engages the
ability of the system is essentially rooted in the mono- clutch plate through diaphragm. During the engage-
tonicity of the clutch diaphragm, and the actuation ment, the lever is pushed by an actuation system, and
system becomes uncontrollable when the diaphragm then, it moves the piston to compress the diaphragm,
has a ‘‘downward force-travel characteristic.’’26 Lee clutch friction plate, and pressure plate. Control preci-
et al.27 integrated the spring, lever, wedge, and many sion and response time are critical for a clutch actua-
other components into an actuation mechanism, and tion system, which determines how smooth and how
for such complicated system, the reliability, maintain- fast a gear shift can achieve.3,6,29 In addition to the
ability, and cost limit its practicality. hydraulic-based system shown here, the electromecha-
To date, electromechanical dry-clutch actuation sys- nical actuation system has become popular and shown
tem with large aiding force and good controllability is promises in meeting the control requirements with a
still of research interest, which includes monotonic or low cost potential.14,21
non-monotonic diaphragms. This article is to propose Figure 2 shows an example of the electromechani-
a new force-aided mechanism by adding a rotating pre- cal actuation system design featured by a moving
loaded spring to a lever. The spring can rotate along pivot, which adjusts the force ratio between the two
with the lever. With bigger rotational angle, the spring ends of the lever. Besides, one end of the lever is con-
can store and release more energy to assist the clutch nected to a preload spring, and the lever motion is dri-
apply. In this case, the aiding force becomes larger as ven by the preload spring. The spring is compressed
the clutch piston travels during the engagement. The and installed along the vertical direction at the end of
aiding force is nonlinear with respect to the rotating a lever point B. The other end of the lever point C
angle, and comprehensive modeling and analysis are directly acts against the diaphragm of clutch. In this
necessary to study the function and characteristics of arrangement, the diaphragm is compressed by the
the new mechanism. spring force through the lever. A soft leaf spring is
This article is organized as follows. In section attached to the lever on one side and a stiff support D
‘‘Design of force-aided lever actuator,’’ the principle of on the other. As soon as the electric motor drives the
force-aided lever and the design are presented. In fulcrum roller O through a power screw, the fulcrum
Liu et al. 3

Figure 1. The schematic representation of the dry-clutch assembly with a traditional actuator.

Figure 2. The schematic representation of the dry-clutch assembly with an ‘‘existing production actuator.’’

roller in turn pushes the lever along its profile and the compressed spring is along the direction of OB.
causes a change in the force arm OB and OC. As a With the force acting on the pivot O in this position,
consequence, the piston force Fc is controlled by the the effective moment generated by the spring force is
position of fulcrum roller O. zero. As soon as the motor force Fm pushes the lever
In this concept, the driving motor is used to move downward and causes it to rotate clockwise around the
the pivot, and the associated power consumption would pivot O, point B of the lever will move up. Accordingly,
be less than the case when the motor is directly acting the spring will rotate counterclockwise about the pivot
on the lever. However, at the same time, it has limited D. When the spring moves away from its original posi-
the controllability of the lever rotating dynamics, espe- tion, the spring starts to generate positive (clockwise)
cially under the negative stiffness of the diaphragm. In moment on the lever in the same direction as the clutch
what follows, a new electromechanical clutch actuation applies for assistance. The effective moment from the
system with a force-aided feature is proposed. spring depends on the angle of the rotation of the lever.
The principle of the force-aided lever for single A larger spring stiffness and bigger rotation angle result
clutch is described in Figure 3. When the clutch is open, in a higher aiding force.
4 Advances in Mechanical Engineering

Figure 3. The schematic representation of the dry-clutch assembly with force-aided lever actuation system.

Table 1. Comparison between ‘‘existing production actuator’’ and ‘‘force-aided actuator.’’

‘‘Existing production actuator’’ ‘‘Force-aided actuator’’

Principle Move pivot rollers Force-aided lever

Actuator travel (mm) 35 34
Application for ‘‘downward force-travel characteristic’’ diaphragm 3 O
Application for ‘‘upward force-travel characteristic’’ diaphragm O O
Motor rated power (W) 170 50
Minimum actuation time (ms) 110 180
Lever length (mm) 80 180

Although both of the electromechanical systems illu- it eliminates the need for hydraulic circuits, oil pump,
strated in Figures 2 and 3 consist of a motor, lever, and solenoid valves, oil pressure sensors, and other hydrau-
spring, their principles differ fundamentally. In the con- lic accessories. In addition, the force-aided spring
cept illustrated in Figure 2, the force to engage clutch assists the motor during clutch engagement, so that a
solely comes from the spring, and the motor moves the smaller motor can be used with less power consump-
pivot of lever to change the actual applying force. In tion. Therefore, a 12-V system is sufficient, which is
force-aided lever actuator in Figure 3, the pivot is fixed, desired from the economic consideration. Besides, the
and both motor and spring contribute to the clutch equivalent aiding force increases with the lever rota-
normal force during the clutch apply. The motor pro- tion, which means that the spring can provide larger
vides the torque on the lever directly, so it has a better aiding force when needed. This characteristic will be
controllability of the lever rotating dynamics. studied in the following section. Finally, the electrome-
Consequently, the lever system may have a better stabi- chanical system improves control robustness under var-
lity under the feedback control through the motor. ious transmission operation conditions.
Table 1 lists the main design properties of the two
systems discussed above for comparison. In the existing
actuator concept, the pivot of lever is essentially con- Physical structure
trolled, so that a diaphragm with upward force-travel Figure 4 and Table 2 illustrate the layout of the novel
characteristics can be used, and plus, it needs a high clutch actuation system and the name of each part for
precision in the lever curve manufacturing. However, a dry-DCT transmission. For each clutch, the actuation
for the force-aided actuator, the controllability of the system includes a preload spring (16), lever (4), linear
system is not sensitive to the lever geometry, but the unit with motor (3), and connectors (2, 6, 12, 14, 15,
trade-off here is that it needs a longer lever. 17, and 18). The lever can rotate about the pivot (13).
In summary, the force-aided lever mechanism pro- Driven by motor (3) through a slider (1), the lever
posed here has the following advantages. First of all, rotates and pushes the clutch piston (9) to apply the
this electromechanical actuation system appears to be clutch. A preload spring (16) is installed on the other
more cost-effective compared to a hydraulic system as end of the lever, which can rotate about the pivot (19).
Liu et al. 5

move the lever for clutch apply and release. It can be a

linear motor or a rotary motor with a power screw. We
used the latter option in our prototype for cost consid-
erations. The parameters for the actuation system are
listed in Table 3.

System analysis and modeling

A numerical model of the force-aid actuation system in
a vehicle is developed to study its dynamic characteris-
tics to further validate the performance of the electro-
mechanical actuator. The model is developed in
MATLAB/Simulink. It consists of two layers, as shown
Figure 4. 3D model of the clutch actuation system. in Figure 5. The first layer is the actuator model,
including the DC motor, power screw, lever mechan-
ism, and accessories; the second layer is the drivetrain
Table 2. Description of actuation system. model with a DCT transmission including the engine,
dual-mass flywheel (DMF), gearbox with two clutches,
1 Slider 2 Motor shaft and vehicle to simulate a vehicle gear shift.
3 Linear unit with motor 4 Lever I Each actuation system model is composed of four
5 Transmission case 6 Roller of slider main components: motor module, power screw, lever
7 Lever II 8 Clutch piston II
9 Clutch piston I 10 Bearing mechanism, and diaphragm, as shown in Figure 6.
11 Input shaft 12 Support of pivot O Most components of this system are treated as rigid
13 Pivot O 14 Roller of sleeve body, including the lever, motor, and some connectors.
15 Sleeve 16 Preload spring The elastic components such as force-aided spring and
17 Spring piston 18 Support of pivot D
19 Pivot D
diaphragm are treated as pure stiffnesses with the possi-
ble inertias are neglected. The dotted line illustrates the
position of the lever when the clutch is fully open. Fm is
the actuation force from DC motor, Fc is the piston
Both of the pivots (13 and 19) are fixed on the trans- force from diaphragm, Fq is the force between slider
mission case. And a sleeve (15) with a roller is installed and lever, Fs is the spring force, u is the rotational angle
between the lever (4) and the force-aid spring (16) to of the lever, a is the rotational angle of the spring, and
transfer the moment generated by the spring to the l, l1 , l2 , and l0 are the lengths of the force arm OA, OC,
lever and then to the clutch piston. A motor is used to OD, and OB, respectively.

Table 3. Parameter of Actuator.

Symbol Value Unit Description

Pmr 50 W Motor rated power

Km 0.131 N m/A Motor moment coefficient
R 0.567 O Motor electric residence
L 4.13 3 1024 H Motor inductance
Kf 9.632 3 1024 Nms Motor damping coefficient
Im 1.81 3 1025 kg m2 Motor inertial
l 0.18 m Length of OA
l1 0.060 m Length of OC
l2 0.070 m Length of OD
l0 0.030 m Length of OB
Ks 7 3 107 N/m Preload spring stiffness
Fp 3500 N Spring preload
I 0.002 kg m2 Lever inertia about pivot O
c 0.1 N m s/rad Lever rotary damping
ab 0.158 rad Bevel angle of the screw
rm 4 3 1023 m Acting radius on the power screw of the motor
mBk 0.09 – Slipping friction coefficient
msk 0.10 – Static friction coefficient
6 Advances in Mechanical Engineering

Figure 5. Diagram of dynamic model of DCT with new actuation systems.

Figure 6. Dynamic analysis of the actuation system.

Motor and the number of turns of wire around the conductor

The dynamics of the motor consists of electromagnetic coil.
and mechanical parts. The resistance of the circuit R
and the self-inductance of the armature L are assumed
to be constant. The following differential equations Power screw
describe the behavior of this electromechanical system The ball screw unit consists of a screw, a screw nut, and
bearing balls. For convenience of force analysis, the
dim ball screw unit is supposed to be expanded as shown in
Vapp  Kb u_ m = L + R im ð1Þ
dt Figure 7.
Im €
um =  Kf u_ m + Km im  Tmo ð2Þ The force Fa is generated by the motor and calcu-
lated by
in which um is the motor rotating angle and Kb is the
electromotive force (EMF) constant, which also Tmo
Fa = ð3Þ
depends on certain physical properties of the motor. Im rm
is the motor inertia, Kf is a linear approximation for
in which, rm is the acting radius of the motor torque.
viscous friction, Vapp is the applied voltage, Tmo is the
The force balances of the screw yield
load torque or motor output torque, im is motor cur-
rent and Km is the armature constant related to physical
Fa = Fbs sin ab + FfB cos ab ð4Þ
properties of the motor, such as magnetic field strength
Liu et al. 7

Figure 8. Coulomb friction model.

Figure 7. Force analysis of the ball screw unit. in which, I is the inertia of lever rotating with pivot O
and c is the rotating damping of lever. a is the spring
rotating angle. l, l1 , l2 , and l0 are the geometry lengths
in which, Fbs is the normal force acting on the screw by of the lever. The force Fq is the slider force, which can
the balls, FfB is the friction force on the screw, and ab is be calculated as
the bevel angle of the screw.
The force balance of the slider yields Fm
Fq = ð10Þ
cos u
0 0
Fm = Fbs cos ab  FfB sin ab ð5Þ
The spring force Fs can be calculated as
0
in which, Fbs is the normal force acting on the screw  
0
nut by the balls and FfB is the friction force of the screw Fp
Fs = Ks l2  l0 +  ls ð11Þ
nut. Ks
The equilibrium of the action and reaction forces
where ls , Ks , and Fp denote the spring length DB, stiff-
yields
ness, and preloaded force of spring, respectively.
0
Fbs = Fbs ð6Þ According to the motion constraint of lever, the follow-
ing equations are obtained
0
FfB = FfB ð7Þ
sin a sin u
= ð12Þ
The friction force FfB is modeled using the coulomb l0 ls
friction model as shown in Figure 8. The mathematical
expression is given as follows ls2 = l22 + l02  2l2 l0 cos u ð13Þ
(    The motion constraint of lever and the ball screw
mBk Fbs sign u_ m u_ m .e0 yields the calculation of u as follows
FfB =   ð8Þ
½msk Fbs , msk Fbs u_ m \e0
rm tan ab um
tan u = ð14Þ
in which, mBk and msk denote the slipping friction and l
static friction coefficient, respectively, and e0 is a small
positive constant, which is used to differentiate the stick
friction from the slip friction. Diaphragm
In this article, the DCT clutches are normally open.
Figure 3 shows the diaphragm installation in one clutch
Lever pack. During the clutch engagement, the spring exhibits
Figure 6 shows the schematic diagram for the system the diaphragm characteristics first in clearance phase
force analysis. The dynamic equation can be written as before the pressure plate comes in contact with the fric-
tion plate, and then, it will behave like a lever spring
Fq l once they touch.12 The piston force Fc and the clutch
I€
u= + Fs l2 sin a  Fc l1  cu_ ð9Þ
cos u normal force Fn are functions of piston travel l, which
8 Advances in Mechanical Engineering

Figure 9. Two typical characteristics of diaphragm for normally open clutch: (a) ‘‘Upward force-travel characteristic’’ diaphragm and
(b) ‘‘Downward force-travel characteristic’’ diaphragm.

Figure 10. Simplified two-speed DCT dynamic model.

are obtained by experiment using curve fitting with a from 1.6 to 5.5 mm, which may cause a stability issue
fifth-degree polynomial and exponential function30 for the actuator design.26

Fc = p1 l5 + p2 l4 + p3 l3 + p4 l2 + p5 l ð15Þ
  The powertrain model
1  ea3 ðla2 Þ
Fn = a1 x  a2  The transmission is based on a dry DCT that has six
a3 ð16Þ
forward speeds and one reverse, with the odd gears on
ð0:5tan hð100ðl  a2 ÞÞ + 0:5Þ the hollow shaft connected to clutch 1 and even gears
on the solid shaft connected to clutch 2.29 To demon-
in which, p1 , p2 , p3 , p4 , and p5 and a1 , a2 , and a3 are fit-
strate the functionality of the actuator, a typical upshift
ting parameters for the piston force and the clutch nor-
is investigated without loss of generality. In this case,
mal force, respectively. The piston travel l can be
only a simplified transmission model is used with only
calculated as
necessary components relevant to the shift considered,
l = l1 tan u ð17Þ as shown in Figure 10. Here, the model has been made
generic to be able to represent all possible upshift with
Two typical diaphragm characteristics, ‘‘upward proper parameters (inertias, ratios, initial conditions,
force-travel characteristic’’ diaphragm and ‘‘downward etc.), while the gear selector and its pre-selection proce-
force-travel characteristic’’ diaphragm, are used for the dure are neglected. The drivetrain model further
validation of the new clutch actuator, as shown in includes engine, DMF, two clutches, gearbox with two-
Figure 9. The piston force Fc increases with piston speed gear sets and final drive, half shafts, and a front-
travel l, which is called ‘‘upward force-travel character- wheel-drive (FWD) vehicle body module. The gear
istic’’ as shown in Figure 9(a). Figure 9(b) shows the shift involves the engagement of the oncoming clutch
‘‘downward force-travel characteristic’’ diaphragm. and the release of the offgoing clutch. Two single-disk
Piston force Fc decreases with piston travel l changing dry clutches are housed in a compact assembly which
Liu et al. 9

also contains the clutch actuators whose structure has The clutch friction torques Tc1 and Tc2 are calculated
been described. using the same friction model. Let Tc denote the general
A 4-degree-of-freedom (DOF) model is built as friction torque, and the calculation is given as
shown in Figure 10. The first DOF derives from the
motion of the lumped inertia Id1 , which represents the Tc = mNRcp Fn ð26Þ
engine and one part of the DMF. The second DOF
derives from the motion of the inertia Id2 which repre- in which, N is the number of friction interfaces of the
sents the other part of the DMF. The third DOF clutch and Rcp is the equivalent radius of the clutch fric-
derives from the motion of the transmission, in which, tion interface. Fn is the clutch normal force, which is
the two clutches and the final drive are represented by obtained from the actuation system model and treated
the inertias Ic1 , Ic2 , and Ifd , respectively. The fourth as an input of the powertrain model. Instead of using
DOF derives from the motion of the inertia Iv , which the discontinuous Coulomb model,32 the friction coeffi-
represents the vehicle. The angular velocities of the 4- cient m is smoothened for computation convenience as
DOF motions are denoted by u_ e , u_ d2 , u_ fd , and u_ v , follows
respectively. The elasticity of the DMF is modeled by 
ms  
 1:0 ezjDuj tan h j  Du_  Rcp
_
the stiffness kd and the damping cd . The elasticity of m = mk 1:0 +
the drive shafts is modeled by the stiffness kv and the mk
damping cv . The rotational damping from friction and ð27Þ
lubrication of the components before the clutches is
modeled by be , and the rotational damping of those in which, mk and ms denote the slipping friction and sta-
after the clutches is modeled by bfd . i1 , i2 , and ifd are tic friction coefficients, respectively; § and z are two
the first gear ratio, second gear ratio, and final drive tuning parameters; and Du_ is the slipping speed of the
ratio, respectively. clutch.
The governing equations of 4-DOF model for the
two-speed DCT system are as follows
Controller design
  
ue = Te  be u_ e  kd ðue  ud2 Þ + cd u_ e  u_ d2 ð18Þ
Id1 € Actuator controller
  For a clutch actuator controller, the input is clutch nor-
Id2 €
ud2 = kd ðue  ud2 Þ + cd u_ e  u_ d2  Tc1  Tc2 ð19Þ
mal force command, and the output is the motor vol-
i21 i2fd Ic1 + i22 i2fd Ic2 + Ifd €ufd tage, which will drive the motor to apply the clutch
ð20Þ normal force through the lever mechanism. The dia-
= i1 ifd Tc1 + i2 ifd Tc2  Tg out  bfd u_ fd gram is shown in Figure 11.
The controller is composed of a force lookup table, a
Iv €uv = Tg out  Tload ð21Þ position loop, and a current loop. The force lookup
table is based on the diaphragm characteristic
in which, Tg out stands for the transmission output shaft (Figure 9), and the motor position command is calcu-
torque lated from the target clutch normal force. Theoretically,
    if the lookup table is known precisely, clutch normal
Tg out = kv ufd  uv + cv u_ fd  u_ v ð22Þ force can be controlled well if the piston traveling is
controlled precisely. Note that the uncertainties of the
The engine output torque Te is dependent on the diaphragm or flat spring usually exist. It needs advance
throttle position (TP) and the angular velocity u_ e of the control method to deal with these uncertainties, such as
engine, which are expressed as clutch torque estimation or robust control.
  Then, the motor position command um cmd is com-
Te = f TP, u_ e ð23Þ pared with the actual motor position and fed into a
proportional–integral–derivative (PID) position loop to
Considering the frontal drag and rolling resistance,
output the motor current command, with a saturation
the vehicle load torque Tload is calculated as31
of 8 A. Finally, the current loop will output the motor
  voltage Vapp with a saturation of 12 V. The PID con-
Tload = c1 + c2  V + c3  V 2 Rtr ð24Þ
troller is defined with coefficients (Kpp , Tpi , and tp ) as
in which, Rtr is the tire radius and c1 , c2 , and c3 are the follows
three coefficients. The vehicle velocity V is calculated as  ð 
1 dep
im cmd = Kpp 1+ ep dt + tp ð28Þ
V = u_ v Rtr ð25Þ Tpi dt
10 Advances in Mechanical Engineering

Figure 11. Actuator controller diagram.

 ð 
1 studied. Second, based on the dynamic model, the
Vapp = Kip 1 + ei dt ð29Þ
Tii actuation performance will be benchmarked with the
same actuator without the force-aided spring in terms
in which of clutch engagement and system response. Besides, the
stability of the new actuation system will be examined.
ep = u m cmd  um ð30Þ
Finally, the simulation of the actuation system in a
e i = im cmd  im ð31Þ complete drivetrain environment including a DCT
transmission is used to further demonstrate the actua-
tor capability during a clutch-to-clutch shift. The para-
Clutch-to-clutch upshift control strategy meters of actuation system used in simulation are listed
Unlike the conventional automatic transmissions, the in Table 3.
DCT needs more precise clutch control, since no torque
converter is available to dampen possible dynamic
impacts during a shift.31 The clutch torque needs to be
Effect of the influencing factors on aiding force
controlled precisely for a good vehicle launch and Suppose the clutch actuator is at equilibrium and the
smooth gear shifting. Both the torque and timing are components are static, that is
critical. The objective of the shift control is to maintain
the possible torque disturbances minimum with the u_ = 0, €u = 0 ð32Þ
proper clutch control and coordination to achieve good
drive quality.33 Substitution of equation (32) into the equations (9) and
The new force-aided lever actuator system needs to (10) yields the expression as follows
execute the typical clutch-to-clutch control strategy
Fm l l2
with four phases, that is, fill phase, torque phase, iner- Fc = + Fs sin a ð33Þ
cos2 u l1 l1
tia phase, and end phase. The desired Fn is controlled
by the new actuator motor position. In the fill phase, Equation (33) shows that at equilibrium, the force that
the oncoming clutch needs to overcome the clearance compresses the diaphragm is equal to the sum of the
quickly with minimum impact; during the torque phase, actuation force Fm and spring force Fs . The following
the drivetrain torque needs to be transferred from the equation gives the equivalent aiding force Fseq acting
offgoing clutch to the oncoming clutch; in the inertia on the clutch piston that is generated by the force-aided
phase, the actuator needs to adjust the clutch normal spring
force in order to synchronize the friction plate and steel
plate quickly and smoothly. As the shift enters the iner- Fseq = Fs ll21 sin a ð34Þ
tia phase, the torque carried by the offgoing clutch is
zero. The speed ratio starts to change and the torque By substituting equations (11)–(13) into the right side
carried by the oncoming clutch needs to be controlled of equation (34), the following equation is obtained
accurately to facilitate the ratio change. In the end 2 0 ffi13
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2  
phase, the oncoming clutch increases to the maximum l0 l0 l0
value in order to lock the clutch. The shift strategies are Fseq = 4Fp  Ks l2 @  1 + 1+ 2 cos uA5
l2 l2 l2
shown in Figure 12.
ðl0 =l2 Þsin u l2
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 
2 l
1 + ðl0 =l2 Þ  2ðl0 =l2 Þcos u 1
Results and discussion
ð35Þ
The performance validation through simulations for
this new system includes three parts. First, the influence The equivalent aiding force Fseq can show the force-
of different parameters on force-aided feature will be aided feature of this actuation system. Higher Fseq
Liu et al. 11

Figure 12. Clutch-to-clutch control strategy during upshift.

means more aiding force from the spring. Equation Generally speaking, a large spring preload and long
(35) shows that Fseq is related to the rotational angle u, l0 =l2 should be considered if a high aiding force is
spring preload Fp , stiffness Ks , and geometry length desired. However, the geometry dimensions for an
l0 =l2 and l1 =l2 . Here, the maximum aiding force Fseq is actuator are usually constrained in practice. Besides, a
taken as the objective. This single-objective problem is large spring preload may possibly result in an unstable
generally simpler than those multi-objective prob- system, and it is unfavorable from manufacturing point
lems.34,35 There are various methods to solve the single of view.
or multi-optimization problems.36 In this article, two
variable trade-off surfaces37 are used for analyzing the
influence of the aiding force. System response for step input
The lever rotational angle u, spring preload Fp , stiff- This section will study ultimate responsiveness of the
ness Ks , and geometry length l1 =l2 and l0 =l2 are treated actuation system by following the position step input.
as the five independent influencing factors. Calculated Using the control scheme designed in section
from equation (35), the effect of the spring preload Fp ‘‘Controller design,’’ the actuation system is validated
and the lever rotational angle u is shown in Figure by the dynamic models with two diaphragms of differ-
13(a). The aiding force Fseq increases when the angle u ent characteristics discussed in section ‘‘System analysis
is small and then tends to decrease when the angle u and modeling.’’ Figure 14 shows the simulation results
becomes larger. It can be seen that the aiding force Fseq for a diaphragm with the ‘‘upward force-travel charac-
can be approximately linear with respect to the angle u teristics’’ (Figure 9(a)). The actuation system generates
within the range u 2 ½0, 0:26rad or u 2 ½0, 15. In addi- a normal force on the clutch plate by the lever to a cer-
tion, larger spring preload Fp generates larger aiding tain position controlled by the motor. The motor angu-
force Fseq . Figure 13(b) shows the effect of geometry lar position reference is given as a step function with an
length l1 =l2 and l0 =l2 . Increasing of l1 =l2 will reduce the amplitude of 36 rad, which corresponds to a 3000 N
aiding force Fseq . When the range of l1 =l2 is relatively normal force, as shown in Figure 14(a). The motor cur-
small, the effect on aiding force Fseq is relatively obvi- rent, motor speed, piston force, and clutch normal
ous. Figure 13(c) shows that the aiding force is not sen- force are illustrated in Figure 14(b)–(e). The motor
sitive to the spring stiffness. speed increases rapidly at first because the motor
12 Advances in Mechanical Engineering

Figure 13. Equivalent aiding force analysis under different parameters: (a) aiding force Fseq versus rotational angle u and spring
preload Fp , (b) aiding force Fseq versus length l1 =l2 and l0 =l2 , and (c) aiding force Fseq versus stiffness Ks .

current starts at a high value to accelerate the motor. spring, the motor torque is adjusted, so that the normal
After that, the motor current is kept constant at 8 A, force is retained following the same profile. Therefore,
which is the maximum current allowed by the motor, the force-aided feature can be compared. The spring can
while the motor speed decreases as the piston force Fc reduce about 30% motor torque in both clutch engage-
increases. The clutch normal force Fn keeps zero until ment and disengagement. And the force-aided feature is
1.05 s since the clutch is through the clearances (previ- more obvious with the increase in normal force.
ously defined as the fill phase). In the end, the motor cur- The simulation results in Figures 14–18 show the con-
rent decreases rapidly in order to avoid any overshoot. trollability for both monotonicity and non-monotonicity
Similar simulation results for a diaphragm with the of the clutch diaphragm since the motor can control the
‘‘downward force-travel characteristics’’ (Figure 9(b)) are actuation force Fm directly by motor torque Tmo .
shown in Figure 15. The motor speed is controlled to drop
initially followed by an increase from 1.02 to 1.18 s, since Stability validation
the piston force initially increases and then decreases.
Because of the force-aided feature of the actuation This section validates the stability and shows the advan-
system, the clutch engagement can be much faster, as tage of the system by comparing to a production actua-
shown in Figure 16. The red line is the command of tion system. Both the force-aided lever actuation system
step input. The response time can be reduced by 24% and the production actuation system are controlled by
with the help of the spring (from 205 to 160 ms) for the same PID controller with respect to the motor posi-
‘‘upward force-travel characteristic’’ diaphragm and by tion. Both the ‘‘upward force-travel characteristic’’ and
19% (from 220 to 180 ms) for ‘‘downward force-travel a ‘‘downward force-travel characteristic’’ diaphragms
characteristic’’ diaphragm. (Figure 9) are used for the two actuation systems for
comparison. In addition, the simulation results of the
‘‘existing actuator’’ actuator are provided for compari-
Torque-saving characteristic verification son. The dynamic model of the ‘‘existing actuator’’ is
Due to the spring installed, motor torque is reduced dur- derived in Appendix.
ing clutch engagement and disengagement. Figures 17 Figure 19 shows the simulation results with ‘‘upward
and 18 show the process of clutch engagement and dis- force-travel characteristics’’ diaphragm (Figure 9(a)). The
engagement. The clutch normal force is changed from 0 motor position is controlled by the PID controller, and
to 1000 N and to 3000 N finally. With or without the the equilibrium state of the actuation system is reached at
Liu et al. 13

Figure 14. Step response of the force-aided lever actuation system with ‘‘upward force-travel characteristic’’ diaphragm: (a) motor
rotational angle um , (b) motor current im , (c) motor speed u_ m , (d) piston force Fc , and (e) clutch normal force Fn .

l = 1.6 mm. To study the system natural robustness of Figure 20 shows the simulation results under the
the two actuation systems, a disturbance force is added same conditions with ‘‘downward force-travel charac-
on the piston (point C in Figure 3)) at 6 s, as shown in teristics’’ diaphragm (Figure 9(b)). The equilibrium
Figure 19(a). Both of the two actuation systems with the state of ‘‘existing production actuator’’ seems to be rea-
PID controller can hold the motor position by adjusting lized and a steady-state normal force at l = 1.6 mm is
motor current, as shown in Figure 19(b) and (c). The pis- obtained (Figure 20(e) and (f)). However, when a small
ton position of ‘‘existing production actuator’’ actuation disturbance Fd = 15 N is applied on piston (Figure
system is shifted from 1.6 to 1.7 mm in the presence of 20(a)), ‘‘existing production actuator’’ leaves its equili-
the disturbance, as shown in Figure 19(e). The simulation brium state easily, and the piston travel l changes from
results show that both the new actuator and the existing 1.6 to 8.0 mm (Figure 20(e)), which cause the clutch
production actuator actuation system are stable under a normal force to change from 0 to 3180 N (Figure 20(f)).
diaphragm with ‘‘upward force-travel characteristic.’’ The The operation point passes through the whole ‘‘down-
new actuation system proposed in this article demon- ward’’ area (1.6 mm \ l \ 5.5 mm in Figure 9(b)) and
strates higher robustness disturbance than the existing stops at new ‘‘upward’’ area (l . 5.5 mm in
production system.
14 Advances in Mechanical Engineering

Figure 15. Step response of the force-aided lever actuation system with ‘‘downward force-travel characteristic’’ diaphragm:
(a) motor rotation angle um , (b) motor current im , (c) motor speed u_ m , (d) piston force Fc , and (e) clutch normal force Fn .

Figure 16. Fast response comparison between actuators with and without spring: (a) fast response with ‘‘upward force-travel
characteristic’’ diaphragm and (b) fast response with ‘‘downward force-travel characteristic’’ diaphragm.
Liu et al. 15

Figure 17. Force-aided feature during clutch engagement and disengagement for ‘‘upward force-travel characteristic’’ diaphragm:
(a) clutch normal force during engagement processing, (b) clutch normal force during disengagement processing, (c) motor torque
during engagement processing, and (d) motor torque during disengagement processing.

Figure 18. Force-aided feature during clutch engagement and disengagement for ‘‘downward force-travel characteristic’’ diaphragm:
(a) clutch normal force during engagement processing, (b) clutch normal force during disengagement processing, (c) motor torque
during engagement processing, and (d) motor torque during disengagement processing.
16 Advances in Mechanical Engineering

Figure 19. Stability validation with ‘‘upward force-travel characteristic’’ diaphragm (‘‘force-aided + PID’’ stands for proposed
actuation system with a well-tuned PID controller; ‘‘existing actuator + PID’’ stands for ‘‘existing production actuator’’ system with
a well-tuned PID controller): (a) disturbance on piston at point C Fd , (b) motor current im , (c) motor rotation angle um , (d) motor
speed u_ m , (e) piston travel l, (f) clutch normal force Fn , and (g) piston force Fc .
Liu et al. 17

Figure 20. Stability validation with ‘‘downward force-travel characteristic’’ diaphragm (‘‘force-aided + PID’’ stands for proposed
actuation system with a well-tuned PID controller; ‘‘existing actuator + PID’’ stands for existing production actuation system with a
well-tuned PID controller): (a) disturbance on piston at point C Fd , (b) motor current im , (c) motor rotation angle um , (d) motor
speed u_ m , (e) piston travel l, (f) clutch normal force Fn , and (g) piston force Fc .

Figure 9(b)). By comparison, the new actuation system The simulation results in Figures 19 and 20 show that
is stable and keeps the piston travel l at 1.6 mm. ‘‘force-aided + PID’’ actuation system has strong
18 Advances in Mechanical Engineering

Table 4. Parameters of powertrain.

Symbol Value Unit Description

Id1 0.058 kg m2 Lumped inertia

Id2 0.1 kg m2 Lumped inertia
Ic1 0.02 kg m2 First gear shaft inertia
Ic2 0.02 kg m2 Second gear shaft inertia
Ifd 0.04 kg m2 Final drive inertia
cd 3 Nms Dual-mass flywheel damping
kd 600 N m/rad Dual-mass flywheel stiffness
cv 100 Nms Output shaft damping
kv 33,000 N m/rad Output shaft stiffness
i1 3.917 – First gear ratio
i2 2.249 – Second gear ratio
ifd 3.7 – Final drive ratio
ms 0.37 – Static friction coefficient
mk 0.35 – Slipping friction coefficient
Rcp 0.1 m Clutch plate radius
N 2 – Number of friction plate surface
§ 4 – Coefficient for friction calculation
j 100 – Coefficient for friction calculation
Rtr 0.314 m Wheel radius
M 1400 kg Vehicle mass M
c1 189 N Load torque coefficient
c2 1.36 N/(m/s) Load torque coefficient
c3 0.45 N/(m/s)2 Load torque coefficient

stability both with an ‘‘upward force-travel characteris- speeding of the engine and the associated inertias dur-
tic’’ diaphragm and a ‘‘downward force-travel character- ing the ratio change can produce a positive driveline
istic’’ diaphragm. torque. The output torque fluctuation can be clearly
observed in the inertia phase. As there is no spike in
the output torque profile, it indicates no sudden jerk
Application to clutch-to-clutch shift during the shift, and the gear shift process may be
The two actuator models are integrated in the vehicle deemed acceptable in terms of running smoothness.
model. The parameters for the powertrain system used The motor position follows the command closely, and
in the simulation are listed in Table 4. the clutch normal force profile also shows that it can
The vehicle is simulated with the acceleration under be controlled well. The results show that this new
the 25% throttle position. Figure 21 shows the simula- force-aided lever actuator system is sufficient for
tion results from the first to second gear. Clutch 1 is clutch-to-clutch gear shift control.
the offgoing clutch and clutch 2 is the oncoming clutch. With the force-aided spring, the motor power
The total shift time here is 0.9 s. The fill phase is fin- requirement can be reduced for gear shifting. Figure 22
ished in 0.1 s. The torque phase lasts 0.2 s, and the iner- shows the motor power for the oncoming clutch during
tia phase takes 0.6 s. the first to second shift with or without the force-aided
In the fill phase, the motor position command (red spring. In both cases, the normal force is controlled to
dashed line in Figure 21(a)) is set to the position where follow the same profile and the vehicle is controlled
the clutch plates touch each other, and the motor cur- with the same acceleration profile. From the simulation
rent and speed are controlled to the maximum value results, it can be concluded that the spring can reduce
allowed in order for the clutch to go through the clear- about 30% motor power.
ance with a minimum time. At the end of the fill phase,
the motor current is changed to a negative value and its
Summary
speed is reduced to avoid the impact at clutch kiss
point. In the torque phase, the normal force on the The clutch actuation system is a key enabler for fast
oncoming clutch increases gradually to take over the and smooth gear shift process for a dry DCT. In this
normal force on the offgoing clutch. During this time, article, a force-aided lever actuation system with a
the engine torque is transmitted from the offgoing clutch spring driven by an electric motor has been proposed.
to the oncoming clutch. After that, the oncoming clutch The design concept and physical structure have been
normal force is adjusted in the inertial phase, and the presented. The dynamic model of a DCT equipped
speed ratio is changed to the second gear. The down with two force-aided lever actuators has been
Liu et al. 19

Figure 21. Simulation results for upshift process: (a) motor position, (b) motor current, (c) motor speed u_ m , (d) clutch normal
force Fn , (e) engine speed u_ e , (f) engine torque Te , and (g) transmission output torque Tg out .

developed. The dynamic simulation results demonstrate transmission controls. This is about 30% reduction in
the feasibility of the force-aided lever actuator to fur- motor power required. In addition to the fast response,
ther validate the actuator design. With the aid of the small actuation force, and accurate control, the new
additional spring, it has been demonstrated that a actuator cost is less and has shown great potential for
motor of 50 W is sufficient to actuate the clutch for practical applications.
20 Advances in Mechanical Engineering

10. Kim J and Choi SB. Design and modeling of a clutch

actuator system with self-energizing mechanism. IEEE-
ASME T Mech 2011; 16: 953–966.
11. Glielmo L, Iannelli L, Vacca V, et al. Gearshift control
for automated manual transmissions. IEEE-ASME T
Mech 2006; 11: 17–26.
12. Liu YG, Qin DT, Jiang H, et al. Clutch torque formula-
tion and calibration for dry dual clutch transmissions.
Mech Mach Theory 2011; 46: 218–227.
13. Wheals J, McMicking J, Shepherd S, et al. Proven high
efficiency actuation and clutch technologies for eAMTä
and eDCTä. SAE technical paper, 2009.
14. Wheals J, Turner A, Ramsay K, et al. Double clutch
transmission (DCT) using multiplexed linear actuation
Figure 22. Motor power comparison between with and technology and dry clutches for high efficiency and low
without spring. cost. SAE technical paper, 2007.
15. Yao J, Chen L, Yin C, et al. Modeling of a wedge
Declaration of conflicting interests clutch in an automatic transmission. SAE technical
paper, 2010.
The author(s) declared no potential conflicts of interest with
16. Yao J, Chen L and Yin C. Experimental validation of a
respect to the research, authorship, and/or publication of this
wedge clutch in automatic transmissions. In: Proceedings
article.
of the international conference on advanced vehicle technol-
ogies and integration, China, Changchun, 24–28 October
Funding 2012.
The author(s) disclosed receipt of the following financial sup- 17. Yao J, Chen L and Yin C. Modeling and stability analy-
port for the research, authorship, and/or publication of this sis of wedge clutch system. Math Probl Eng 2014; 2014:
article: This project was supported by the National Natural 712472 (12 pp.).
Science Foundation of China (grant no. 51475284). 18. Kim J and Choi SB. Self-energizing clutch actuator sys-
tem: basic concept and design. In: Proceedings of
FISITA2010 world automotive congress, Budapest, 30
References May–4 June 2010. Budapest, Hungary: Springer.
1. Oh JJ, Kim J and Choi SB. Design of self-energizing 19. Taguchi Y, Soga Y, Mineno A, et al. Development of an
clutch actuator for dual-clutch transmission. IEEE- automated manual transmission system based on robust
ASME T Mech 2016; 21: 795–805. design. SAE technical paper, 2003.
2. Gao B, Lu X, Chen H, et al. Dynamics and control of 20. Zhang J. Servomechanism for automatic clutch. Google
gear upshift in automated manual transmissions. Int J patents, 2004.
Vehicle Des 2013; 63: 61–83. 21. Wagner U, Berger R, Ehrlich M, et al. Electromotoric
3. Van Berkel K, Hofman T, Serrarens A, et al. Fast and actuators for double clutch transmissions. In: LuK
smooth clutch engagement control for dual-clutch trans- symposium, Schweinfurt, Germany, 3–4 April 2006,
missions. Control Eng Pract 2014; 22: 57–68. pp.137–153. Schaeffler Technologies AG & Co. KG.
4. Yao J, Chen L, Liu F, et al. Experimental study on 22. Faust IH, Bünder IC and DeVincent E. Dual clutch
improvement in the shift quality for an automatic trans- transmission with dry clutch and electro-mechanical
mission using a motor-driven wedge clutch. Proc actuation. ATZautotechnology 2010; 10: 32–37.
IMechE, Part D: J Automobile Engineering 2014; 228: 23. Burkhart D and Dreher A. Clutch system. Google patents,
663–673. 2009.
5. Oh J, Kim J and Choi S. Robust feedback tracking con- 24. Deur J, Milutinovic M, Ivanovic V, et al. Modeling of a
troller design for self-energizing clutch actuator of auto- dry dual clutch utilizing a lever-based electromechanical
mated manual transmission. SAE technical paper, 2013. actuator. J Dyn Syst: T ASME 2016; 138: 091012 (11 pp.).
6. Manyala J. Gearshift actuator dynamics predictions in a 25. Ivanović V, Hoić M, Deur J, et al. Design of test rigs for
dual clutch transmission. SAE technical paper, 2013. a dry dual clutch and its electromechanical actuator. SAE
7. Chen L, Xi G and Sun J. Torque coordination control technical paper, 2012.
during mode transition for a series–parallel hybrid elec- 26. Kimmig K, Bührle P, Henneberger K, et al. Dry double
tric vehicle. IEEE T Veh Technol 2012; 61: 2936–2949. clutch-success with efficiency and comfort. In: 9th Schaef-
8. Van Berkel K, Veldpaus F, Hofman T, et al. Fast and fler symposium, Schweinfurt, Germany, 26–28 August
smooth clutch engagement control for a mechanical hybrid 2010, pp.154–163. Schaeffler Technologies AG & Co. KG.
powertrain. IEEE T Contr Syst T 2014; 22: 1241–1254. 27. Lee JW, Jeong EH and Kam JH. Clutch actuator for a
9. Balau A-E, Caruntu C-F and Lazar C. Simulation and vehicle. Google patents, 2013.
control of an electro-hydraulic actuated clutch. Mech 28. Kimmig K and Agner I. Double clutch wet or dry, that is
Syst Signal Pr 2011; 25: 1911–1922. the question. In: 8th LuK symposium, Schweinfurt,
Liu et al. 21

Germany, 3–4 April 2006, pp.120–135. Schaeffler Tech- can be neglected since it is constrained by the leaf
nologies AG & Co. KG. spring. Assume the barycenter of lever is in the middle.
29. Liu YG, Qin DT, Jiang H, et al. A systematic model for The dynamic equations of lever are as follows
dynamics and control of dual clutch transmissions. J
Mech Design 2009; 131: 061012. Fs + Fc  FBl cos a = m€y ð36Þ
30. Tarasow A, Wachsmuth G, Lemieux J, et al. Online esti-  
mation of time-varying torque characteristics of automo- l l l2
€ = ðFc  Fs Þ + FBl
Ia   ca ð37Þ
tive clutches using a control oriented model. In: IEEE 2 2 cos a cos a
American control conference (ACC), Washington, DC,
16 August 2013, pp.6752–6757. New York: IEEE. in which
31. Kulkarni M, Shim T and Zhang Y. Shift dynamics and  
control of dual-clutch transmissions. Mech Mach Theory l
y = l0 +  l2 tan a ð38Þ
2007; 42: 168–182. 2
32. Duan C and Singh R. Influence of harmonically varying
normal load on steady-state behavior of a 2DOF torsional
Fs = ks ðls0  l0 + l2 tan aÞ ð39Þ
system with dry friction. J Sound Vib 2006; 294: 503–528. where Fc is the piston force from the diaphragm; FBl is
33. Gao B, Chen H and Sanada K. Two-degree-of-freedom the fulcrum roller force; Fs is the spring force; ls0 and ks
controller design for clutch slip control of automatic
are the spring initial length and stiffness, respectively; m
transmission. SAE technical paper, 2008.
and I are the mass and inertia of lever, respectively; and
34. Obayashi S, Jeong S and Chiba K. Multi-objective design
exploration for aerodynamic configurations. In: 35th l and l2 are the length of BC and BO at horizontal direc-
AIAA fluid dynamics conference and exhibit, Toronto, tion, respectively.
ON, Canada, 6–9 June 2006. Reston, VA: AIAA. The power screw is used to transfer the moment
35. Huang W, Wang Z, Ingham DB, et al. Design explora- motion to the linear motion. The dynamic equation of
tion for a single expansion ramp nozzle (SERN) using the power screw is given as follows
data mining. Acta Astronaut 2013; 83: 10–17.
36. Silvas E, Hofman T and Steinbuch M. Review of optimal Ips €um = Tmo  Rp FBl sin a ð40Þ
design strategies for hybrid electric vehicles. IFAC P
2012; 45: 57–64. um Rp = l2 ð41Þ
37. Skinner BA, Parks GT and Palmer PR. Comparison of
Rp refers to the distance that the power screw travels
submarine drive topologies using multiobjective genetic
algorithms. IEEE T Veh Technol 2009; 58: 57–68.
when the motor rotates a certain angle and Ips is the
inertia of power screw.
The dynamics of the motor are same as equations (1)
Appendix 1 and (2). The piston force Fc and clutch normal force Fn
Figure 23 shows the force analysis of the existing pro- are obtained from equations (15) and (16). The piston
travel l is calculated as
duction actuation system. When the roller O is being
moved by the motor, the lever is displaced simultane-
l = yc0  ðl  l2 Þtan a ð42Þ
ously with a displacement x of barycenter at horizontal
direction, and a rotate angle a around the fulcrum in which, yco stands for the initial vertical position of
roller with respect to power screw center line. The lever piston.
displacement y at vertical direction is very small and

Figure 23. Dynamic analysis of ‘‘existing production actuator’’ system.