Journal of http://jvc.sagepub.

com/
Vibration and Control

Design of a solenoid valve based active engine mount

Hossein Mansour, Siamak Arzanpour and Farid Golnaraghi
Journal of Vibration and Control 2012 18: 1221 originally published online 23 September 2011
DOI: 10.1177/1077546311417275

The online version of this article can be found at:

http://jvc.sagepub.com/content/18/8/1221

Published by:

http://www.sagepublications.com

Additional services and information for Journal of Vibration and Control can be found at:

Email Alerts: http://jvc.sagepub.com/cgi/alerts

Subscriptions: http://jvc.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://jvc.sagepub.com/content/18/8/1221.refs.html

>> Version of Record - Jun 1, 2012

OnlineFirst Version of Record - Sep 23, 2011

What is This?

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

Article
Journal of Vibration and Control
18(8) 1221–1232

Design of a solenoid valve based active ! The Author(s) 2011

Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
engine mount DOI: 10.1177/1077546311417275
jvc.sagepub.com

Hossein Mansour, Siamak Arzanpour and Farid Golnaraghi

Abstract
This paper describes the design of a versatile and fully controllable active engine mount. The proposed active mount is
capable of addressing vibration isolation requirements at various driving conditions. This design addresses a better ride
quality that has always been demanded by the automotive industry, as well as satisfying sophisticated vibration isolation
requirements for the unconventional engines, i.e. variable displacement, and hybrids. The proposed engine mount
replaces the decoupler of the original design with a solenoid actuator. The mathematical model of the active mount
is obtained. The dynamic characteristics of the mount are shown to be highly controllable over the operating frequency
range of excitation in engines. The effectiveness of the developed active engine mount for various working conditions of
engine is also evaluated. Several driving conditions are investigated and proper control strategies are utilized to dem-
onstrate the mount’s capability to fulfill the isolation requirements for each condition. The promising results, in addition
to compactness, low cost, fail safety, and durability are the main advantages of the proposed active engine mount, which
makes it viable for automotive applications.

Keywords
Active vibration control, driving condition, hydraulic engine mount, vibration isolation
Received: 13 July 2010; accepted: 1 June 2011

1. Introduction frequency/amplitude pattern of engine vibration in its

Vibration isolation in the engine compartment is a chal- course of action. The amplitude of engine vibration is
lenging design problem for all transportation means. usually more than 0.3 mm at low frequencies (1–50 Hz)
particularly for the automotive industry to attain a and less than 0.3 mm at high frequencies (50–200 Hz)
better ride quality, road handling and engine life. The (Golnaraghi and Jazar, 2001). Therefore the isolator is
sources of excitation on the engine can be divided into generally supposed to have high damping - high stiff-
two categories: 1) base excitation imposed by road ness at low frequency excitation and low damping -
pattern and acceleration of the vehicle; and 2) force low stiffness at high frequency excitation. Satisfying
excitation inside the engine. The excitation force of these requirements needs the characteristics of the
the engine is generated from the unbalance force of engine mount to be a function of excitation frequency
rotary parts (such as the crankshaft); the inertia of reci- as well as its amplitude (Brach and Haddow, 1993).
procating parts (such as the pistons); and the firing Hydraulic engine mounts are prevalent types of iso-
impulses inside the cylinders. A proper and effective iso- lators to address the isolation problem of engines
lation system is the one that can control 1) the force/
acceleration transmission to the frame; 2) induced accel-
eration on the engine; and 3) the relative engine-body Mechatronic Systems Engineering, Simon Fraser University, Surrey, BC,
Canada
displacement. As a result, design and control of a
proper engine mount is always a multi-objective problem. Corresponding author:
Siamak Arzanpour, Mechatronic Systems Engineering, Simon Fraser
Handling different types of input towards fulfilling University, Room 4376, 250–13450 102nd Avenue, Surrey, BC V3T
above-mentioned requirements is a conflicting task. 0A3, Canada
This problem is mainly due to the difference in the Email: [email protected]

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

1222 Journal of Vibration and Control 18(8)

(Flower, 1985; Foumani et al., 2003). The common fea- objective (Hagino et al., 1986; Yu et al., 2001), while
ture of all hydraulic mounts is that they exhibit high they ignored the two other important criteria (i.e. accel-
damping when they experience low frequency-high eration on the engine, and relative engine-body
amplitude excitations and low damping under high fre- displacement).Such a control strategy can cause cata-
quency-low amplitude situation. Hydraulic mounts ben- strophic acceleration on the engine or misalignment
efit from a dual dynamic performance mechanism that between engine and body in response to the actuating
makes their isolation characteristics adaptable with the force. As a result, an ideal control strategy is the one
amplitude and frequency of the excitation. However, this that considers all of the dynamic requirements simulta-
type of mount is still passive and is not truly flexible to be neously in defining the control function.
adapted with different loading and road conditions. In this paper, the design and control of a novel
In contrast with passive mounts, active (Arzanpour active mount is described. Generally, active systems
and Golnaraghi, 2008b; Little and Kashani, 1995; have several drawbacks such as size, not being failsafe,
Nakaji et al., 1999) and semi-active (Arzanpour and high energy consumption, high cost, and instability. To
Golnaraghi, 2008a; Hong et al., 2001; Mansour et al., minimize these, the proposed active mount is retrofitted
2010; Williams et al., 1993) mounts are proposed which inside a conventional hydraulic mount. Therefore, if the
are adaptable with various loading conditions. Semi- actuator fails, the passive parts of the mount can still
active mounts change one or more dynamic properties maintain the minimum vibration isolation require-
of the system, mainly damping, in order to dissipate ments. Also, the energy consumption and size of
vibration energy efficiently. In this sense, semi-active the actuator is reduced because the effect of the passive
isolators are adding controllability while maintaining elements is preserved and the actuator will just compen-
the simplicity and cost efficiency of passive mounts. sate for the lack of performance of the passive system.
Semi-active solutions can only generate resistive force The instability problem is mitigated by the rubber part
which is considered as their limitation in dealing with of the passive hydraulic system that restricts growing
complex excitations. The other drawback of semi-active the amplitude of the response of the mount. Another
systems is their low flexibility to provide the desired advantage of a retrofittable actuator is the cost mini-
isolation properties over a wide range of frequency. mization regarding the design of an isolator from
Active mounts benefit from actuators which makes scratch. More importantly, the solenoid based actuator
them capable of defining the dynamic properties at each which is developed for this application has not been
specific frequency. The active isolators are considered implemented for engine mount application. This actua-
as the next generation of engine mounts (Miller and tor is inexpensive compared to voice coil actuators that
Ahmadian, 1992; Swanson, 1993). In active systems, are mainly used for active engine mounts. A mathema-
the force generated with actuators cancels the transmis- tical model of the active engine mount is first derived
sion force to the system. Several types of active mounts which illustrates the ability of active mount to change
have been proposed. A pneumatic-based actuator is used the dynamic properties based on the utilized control
by Ozaki et al. (1999) to actuate an active engine mount. function. A practical control scenario is also provided
Active mounts with piezoelectric actuators are also to make the active mount adaptable with the road and
developed which have mostly utilized amplitude engine excitation. Finally several loading conditions
magnifying mechanisms to compensate for their very and their isolation requirements are described and the
small displacement (Ushijima and Kumakawa, 1993). ability of the developed engine mount to fulfil these
Electromagnetic actuators are other types of actuators requirements is demonstrated.
that are used to develop active mounts. Fursdon et al.
(2000) designed an active mount working with a voice
coil, which was further developed by Hillis et al. (2005)
2. Active hydraulic mount
using different adaptive control techniques. Recent To ensure the performance and ease of manufacturing,
studies on active mounts were mostly focused on control- our proposed active mount is designed to be retrofitted
ling the previously developed active mounts rather than into a passive hydraulic mount.
designing new mounts (Lee and Lee, 2009; Liang et al., A passive hydraulic engine mount consists of two
2010; Turnip et al., 2009). Active engine mounts have fluid-filled chambers connected through two fluid
opened their way into the very competitive automotive passages, i.e. decoupler and inertia track, as shown
industry since 2005. Honda (Hideki et al., 2004), Toyota schematically in Figure 1 (a). Typically, the fluid within
(Ozaki et al., 1999), and Nissan (Aoki et al., 1999) are the mount is a mixture of ethylene glycol and water.
examples of the companies that have started using active The upper or high-pressure chamber is bounded on top
mounts in their commercial products. by a main rubber structure and on the bottom by a
Most of the studies on active mounts are focused on steel plate that houses the inertia track and decoupler.
the ‘force transmission to the frame’ as the sole control The plate is fixed to the base of the mount through a rigid

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

Mansour et al. 1223

structure that surrounds the lower or compliance From the schematic view of the engine mount, the
chamber. The compliance chamber consists of a flexible transmitted force to the chassis (FT) is,
rubber bellow that expands and contracts as fluid passes
though the inertia track or the decoupler. FT ¼ Kr X þ Br X_ þ Ap P1 ð1Þ
The rubber element of the high-pressure chamber
has to carry the static load of the engine in addition According to equation (1) the generated force of a
to providing a portion of the stiffness and damping hydraulic mount comes from two different sources: 1)
required to suppress the engine vibration. The recipro- the bulk rubber; and 2) fluid pressure inside the pump-
cal movement of the engine forces the liquid inside the ing chamber.
pumping chamber through the decoupler and inertia The main rubber is designed to carry the static load
track into the compliance chamber. The ability of the of the engine and its stiffness and damping (i.e. Kr and
compliance chamber rubber to bulge relieves the pres- Br) are inherent properties of its design. As a result, the
sure in the pumping chamber. pressure inside the pumping (upper) chamber is the
The decoupler plate has a finite travel distance best parameter to control for adjusting the dynamic
within its cage; therefore, it limits the volume of the behavior of the engine mount. This identifies the best
fluid that can pass freely between the upper and lower location for the actuator that will affect the perfor-
chambers. Once the plate bottoms out on the cage, the mance of the isolator with minimum actuation effort.
fluid stops passing through the decoupler, and it now An electromechanical actuator is designed and
has to pass through the inertia track. The resistance and retrofitted inside the inertia track plate of a conven-
mass of fluid within the inertia track increase the over- tional hydraulic mount. Decoupler is eliminated from
all stiffness and damping of the mount under this con- the original design, although its performance is still
dition. On the other hand, during small amplitude available by letting the plunger move freely (it has a
excitations, the fluid passes freely through the decou- restricted motion) or by a control signal that derives
pler, giving the mount low damping and stiffness char- the actuator which acts similar to a reciprocating
acteristics. A more detailed discussion on passive pump. The plunger of the actuator is faced to the
hydraulic mounts can be found in Ahmed et al. (1999) pumping chamber from the top and to the atmosphere
Figure 1(b) shows the lumped model of a passive from the bottom. The plunger is sealed with a thin
hydraulic engine mount. C1 and C2 are the compliances rubber diaphragm and a compression spring keeps it
of the pumping and compliance chambers representing in its proper position. Figure 2 shows different parts
their ability to bulge, X is the relative displacement of of the electromechanical actuator designed for the
the engine and chassis or mount displacement, Kr and active mount.
Br are the stiffness and damping of the rubber, Q is the The actuator used for this active mount is composed
flow rate, I and R are the inertia and resistance of the of a plunger and a solenoid coil.Most of the previously
fluid-channels, P1 and P2 are the pressure in the pump- developed active engine mounts were using voice coil to
ing and compliance chambers and the subscripts i and d generate actuating force and the application of solenoid-
stand for the inertia track and decoupler (e.g. Qi is the based actuators is rather new in the area (Arzanpour and
flow within the inertia track). Golnaraghi, 2008b). A solenoid-based actuator is more

Figure 1. (a) Hydraulic engine mount schematic view and (b) its lumped model.

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

1224 Journal of Vibration and Control 18(8)

Figure 2. Design of the electromechanical actuator for the active mount.

passive mount except for the decoupler. A reciprocating

part is added to the model that faces to the pumping
chamber as the representation of the actuator.
The continuity equation for this lumped system can
be written for the pumping chamber as,

C1 P_ 1 þ Qi ¼ Ap X_ þ Aa Y_ ð2Þ

Where Aa is the effective pumping area, and Y is the

displacement of the plunger. Also the continuity equa-
tion for the compliance chamber can be described as,

C2 P_ 2 ¼ Qi ð3Þ

Finally, the fluid momentum equation is,

Figure 3. The lumped model of the active engine mount.
p1  p2 ¼ Ii Q_ i þ Ri Qi ð4Þ
compatible with the harsh environment of the active
mount. It is considerably cheaper, more durable, and it Solving equations (2) to (4) for P1 we will have
has less design/fabrication complexity compared to voice
 
coils. Ap XðsÞ þ Aa YðsÞ Ii S2 þ Ri S þ 1=C2
The actuator used in our design does not benefit from p1 ðsÞ ¼
C1 Ii S2 þ Ri S þ 1=C1 þ 1=C2
a permanent magnet; as a result, the magnetic coil can ð5Þ
only attract the plunger. This problem is addressed by
using a compression spring underneath the plunger and Although the above equation contains some proper-
giving a DC current to the actuator. The DC offset gen- ties of the hydraulic mount, the plunger displacement
erates a static displacement for the plunger against the Y(s) can be set to refine the dynamics of the isolator.
compression spring; therefore, if the imposed current Y(s) can be controlled as a function of the relative
goes below the level of the DC offset the plunger moves displacement of engine and frame X(s) according to
farther from the coil compared to its static equilibrium. equation (6)
This will enable us to produce a reciprocal movement on
the plunger.More details on the design of this active ZðsÞ ¼ YðsÞ=XðsÞ ð6Þ
mount can be found in (Mansour et al., 2010).
Figure 3 shows the lumped model of the proposed The transfer function Z(s) can be set in the control-
active mount. This model has all the components of a ler to act as the control law to drive the actuator based

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

Mansour et al. 1225

on the engine displacement. The last step to organize the force and the indices a, p, and o stand for active,
the engine mount fundamental equations is to convert passive, and overall respectively. The transfer function
them in terms of the dynamic stiffness equation. Z(s) is defined between the mount displacement and
Dynamic stiffness is a well-known quantity to iden- the plunger displacement; H(s) is between the plunger
tify an isolator dynamic performance over a frequency displacement and the active force; G(s) is between the
range. It relates the force transmitted to the chassis to mount displacement and the passive force; finally U(s)
the displacement of the engine at each excitation fre- is a double integral multiplied by the inverse apparent
quency. The relation can be obtained by dividing the mass which is between the overall force and the mount
resultant force to the relative displacement. Although displacement. H(s) and G(s) are derived from equa-
other properties of isolators, such as shock response tion (5) and shown in equations (8) and (9). It should be
or transient response, are other important factors, noted that most of the mount’s force is provided by the
dynamic stiffness is the parameter of choice by car man- passive part (i.e. G(s)) and the active part only tunes it.
ufacturers and researchers. Combining equations (5),
 
(1) and (6) leads us to Aa Ap Ii S2 þ Ri S þ 1=C2
HðsÞ ¼ ð8Þ
C1 Ii S2 þ Ri S þ 1=C1 þ 1=C2
A2p þ Aa Ap ZðsÞ
Kdyn ðsÞ ¼ Kr þ Br s þ  
C1 A2p Ii S2 þ Ri S þ 1=C2
  ð7Þ GðsÞ ¼ ð9Þ
Ii S2 þ Ri S þ 1=C2 C1 2
Ii S þ Ri S þ 1=C1 þ 1=C2
Ii S2 þ Ri S þ 1=C1 þ 1=C2 In real application, an appropriate control function
(i.e. Z(s)) will be picked from a predefined library of
Equation (7) implies that the control algorithm Z(s) functions based on the current operation condition of
can effectively define or modify the dynamic stiffness the vehicle (such as ‘high speed cruising on a smooth
response of the mount. The mount will behave similar road’). The vehicle operation condition can be obtained
to a passive isolator if the plunger of the actuator is from some states, such as engine RPM, throttle condi-
held stationary. Since various vibration patterns of tion, vehicle speed, and acceleration on the chassis.
engine require different dynamic performances by the These quantities are usually available in vehicles for
isolator, the function Z(s) is not unique. applications such as automatic transmission or EPS.
The abovementioned library of functions is designed
to define the right trend for isolation properties of the
3. Control scheme mount based on the requirements stated by the manu-
Closed loop control architecture used to control the facturer (Kinchen and Hogland, 2006).
designed active mount is shown in Figure 4. Although the desired value of the relative displace-
The control parameter is X(s), which is the relative ment is set to zero, the control functions are not
displacement of the engine and the frame. Y(s) is the designed to completely cancel the mount displacement.
displacement of the plunger (inside the actuator), F is The feedback from displacement is just to set the

Figure 4. The proposed control architecture to control the active mount.

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

1226 Journal of Vibration and Control 18(8)

amplitude and phase of the imposed force while the functionality of the fabricated mount. These cases
control functions are designed to satisfy all three accel- are including fixed, free and shaker displacement
eration and displacement requirements simultaneously. proportional movement for the plunger. In the first
It should be noted that the right trend is the only reli- case, the plunger is fixed by giving enough DC cur-
able known information about the desired isolation rent to the actuator. In this case, the active mount
characteristics, while defining a strict set point and for- acts like a passive mount and in mathematical model,
cing the system to stay around that point is just wasting Z(s) will be set equal to zero. In the second case, there
energy while there is still no guaranty that the vibration is no current applied to the actuator; as a result the
behavior is satisfactory. plunger is free to move based on the net force applied
on it. This condition can be modeled by a real and
negative value for Z(s). Finally, in the third case, pro-
4. Experimental analysis
portional control strategy is utilized, which means that
An experimental setup is built to investigate the Z(s) ¼ 1. The amplitude and phase of experimental
performance of the active mount. Figure 5 shows differ- results for these three cases are depicted in Figure
ent components of the testbed. The active engine mount 6(a). As it was expected in the case of fixed plunger
is supported by a frame on top and an LDS V722 shaker the active mount behaves similar to a passive hydraulic
excites it from the bottom. The setup is mounted on a mount while in the case of free plunger the pressure will
heavy concrete base to minimize the effect of the base be released in the pumping chamber and the mount
vibration on the measurements. Dytran accelerometer, behaves softer than the passive mount. In the latter
Model 3145AG LIVM, is also used as a feedback to the case, the active mount acts as a rubber passive mount
shaker controller. The displacement of the shaker with and its dynamic response is a straight line increasing
respect to the fixed frame (which means the elongation of with frequency. Finally, in a proportionally controlled
the engine mount) is measured by LVDT. A Sensotec case, the mount is a bit softer than the passive mount in
Model 41, 1000 lb precision pancake load cell is also lower frequencies while it acts harder in higher frequen-
used for sensing the transmitted force. In this study the cies. In the experimental results, the phase of the con-
frequency range of 5–60 Hz is studied which contains trolled case gets negative at very low frequencies, which
most of the vibration problems in engine isolation. should be due to measurement errors. Figure 6(b)
Excitations occurring beyond 60 Hz are not considered shows the simulation of the dynamic stiffness equation
in this paper since the plunger of the actuator can act for these three cases.
similar to a decoupler at higher frequencies. This paper
focuses more on the low frequency region where the pas- 5. Active mount performance
sive decouplers are mainly ineffective. The first natural investigation in real working
frequency of the frame is about 20 Hz that causes a small
peak around this frequency to appear in the results. This
conditions
fake peak can be ignored in all experiments. To evaluate the performance of the engine mount sev-
Three special cases are studied first to evaluate the eral engine conditions are simulated in the lab environ-
validity of the mathematical model and to approve the ment and the response of the mount for each condition

Figure 5. (a) Active mount’s experimental setup; and (b) its schematic view.

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

Mansour et al. 1227

Figure 6. Active mount (a) experimental and (b) numerical dynamic stiffness response when the actuator is fixed (dotted line) free
(solid line) and proportionally controlled (dashed line).

is recorded for both passive and active situations. These range; therefore, a high damping is required to suppress
conditions include no load, accessory load, cruising at the rotational mode of the engine.
high and moderate speed on a rough road, cruising The experimental results for this case are depicted in
at low to moderate speed on a smooth road, and accel- Figure 7. Compared with the hydraulic mount, the
erating the vehicle. Each condition happens at specific average stiffness of the active mount is reduced by
frequency/amplitude range and requires specific isola- half, while the damping is doubled. The vehicle is not
tion property. These conditions are generated for the moving, so the vehicle dynamics requirements need not
mount using an electromagnetic shaker in the fre- be considered.
quency/amplitude range of interest for each condition.
The engine isolation requirements for the conditions
are provided by General Motors for the mount that is
tested in this study. In evaluation of the mount for each
5.2. Accessory Load Condition
driving condition, lower and higher ranges of frequency The accessory load condition occurs for a stationary
are treated separately as their right trend is not neces- vehicle in drive, but with accessory load (A/C, headlights
sarily the same.The low frequency region is designed to and rear defog) at idle (no throttle input). Accessory load
decrease the negative effect of engine vibration on vehi- decreases the engine RPM to 600–750 RPM; therefore,
cle dynamics, while the higher range is designed to the engine will not work as smoothly as the no-load
address ride comfort concerns. case.Since the excitation force of the engine deviates
from a perfect sinusoidal wave, the first engine RPM
(i.e., 600–700 RPM) is no longer the dominant excitation
frequency. Harmonics (i.e., the whole multiples) of
5.1. No Load Condition engine RPM will emerge in the excitation force. The
For the stationary condition that applies in neutral or major excitation will be the second harmonic of the
park with no accessory load (such as A/C, headlights, engine RPM, which falls in the range of 20–40 Hz.
rear defog), the engine speed is between 650 and 800 A proper engine mount for this case should have low
RPM. The main concern in this condition is how to stiffness and low damping to enable the rigid body rota-
deal with the rigid body mode of the engine, and the tional mode of the engine and prevent the crankshaft’s
dominant excitation is the first engine RPM that is in torque from being transferred to the body. The experi-
the range of 5–20 HZ. For this case, the mount is required mental results of the active mount and the passive mount
to have low stiffness and high damping. Low stiffness are compared in Figure 8.
enables rigid body rotational mode of the engine and The result shows that the active mount is softer
reduces the amount of force transfer to the body due and has lower damping than the passive mount. The
to the crankshaft’s torque. However, random torque average level of stiffness is 170 N/mm compared with
input of the crankshaft falls in the same frequency 250 N/mm for the passive mount; for the damping, it is

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

1228 Journal of Vibration and Control 18(8)

Figure 7. Active mount dynamic stiffness response set for stationary in neutral or park with no unnecessary load (solid line)
compared to the passive mount response (dotted line).

reduced from 70 N/mm to 40 N/mm. As in the no-load stiffness and high damping in the frequency range of
case, vehicle dynamics requirements are not considered 6–20 Hz. The high stiffness requirement is intended to
because the vehicle is stationary. optimize the apparent mass relative to suspension
dynamics. In addition, high damping is required to
5.3. Cruising at High Speed or Moderate Speed optimize power-train response to suspension dynamics.
The experimental result for the frequency range of
on a Rough Road 6–20 Hz is compared with the result for the hydraulic
5.3.1. Noise and vibration requirements. For cruis- engine mount in Figure 10. The results show extreme
ing at steady speed at 45–80 mph, the engine normally increase in both stiffness and damping level. The aver-
works at 1000–3000 RPM. The excitation falls mainly age level of the stiffness in the frequency range of
in the frequency range of 35–400 Hz. According to 5–20 Hz is three times that of the passive one. In addi-
Kinchen and Hogland (2006), in this case, the engine tion, the magnitude of the damping is approximately
mount should have low stiffness and low damping to twice the damping of the passive mount. Both stiffness
stop the induced force of the engine from passing and damping follow the same trend as the passive ones
through the engine mounts. However, this range is do, with no dramatic change observed in the pattern.
mostly outside the scope of active vibration suppres-
sion, especially with a mechanical actuator. To evaluate 5.4. Cruising at Low to Moderate Speed
the performance, an experimental analysis is conducted
on a Smooth Road
on the lower frequency range (35–60 HZ). High fre-
quency requirements, above 60 HZ with very low For cruising at low speed on a smooth road, both
amplitude, can be considered in the design of the vibration and vehicle dynamics requirements have
main rubber. According to the experimental results, moderate stiffness and damping. Conventional passive
depicted in Figure 9, the active mount has a lower stiff- mounts are able to fulfill these requirements. The domi-
ness and damping than the passive mount. nant excitation frequency for this situation is 6–40 Hz
in which the decoupler of conventional mounts is
5.3.2. Vehicle dynamics requirements. For cruising closed. The same characteristics can be made in the
at high speed (above 45 mph) on a smooth road, and active mount by giving the actuator enough current
cruising at low to moderate speed (20–55 mph) on a to hold it against changing pressure in the pumping
rough road, it is required that the mount have high chamber (as depicted in Figure 6).

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

Mansour et al. 1229

Figure 8. Active mount dynamic stiffness response set for stationary vehicle with accessory load (solid line) compared to the passive
mount response (dotted line).

Figure 9. Active mount dynamic stiffness response set for cruising at steady speed in high gear (solid line) compared to the passive
mount response (dotted line).

60 Hz.An ideal mount should have moderate stiffness

5.5. Accelerating the Vehicle and low damping in this condition. The low damping is
This condition usually occurs when the driver acceler- needed to enable high frequency transmissibility of
ates the vehicle to roll into the highway speed. In this the mount, and the moderate stiffness is needed to
case, the engine vibration excitations cover a wide hold the engine against the reaction of the engine
range of 35–600 Hz. The upper bound of this excitation torque. As shown in Figure 11, this requirement can
frequency, 60–600 Hz, will be handled by the low-pass be achieved by our active mount. Although the damp-
characteristics of the main rubber, and the active actua- ing of the active mount is about 30% higher than the
tor should address only the frequency range of 35– damping of the passive mount in the frequencies below

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

1230 Journal of Vibration and Control 18(8)

Figure 10. Active mount dynamic stiffness response set for cruising at high speed or moderate speed on a rough road (solid line)
compared to the passive mount response (dotted line).

Figure 11. Active mount dynamic stiffness response set for accelerating the vehicle (solid line) compared to the passive mount
response (dotted line).

40 Hz, it decreases to one-third of the passive mount’s retrofitting an actuator inside a conventional passive
damping at 60 Hz. hydraulic mount. The actuator is aimed to control the
fluid pressure inside the pumping chamber with no
direct effect on the pressure of the compliance chamber.
6. Conclusion
Being retrtofittable to a passive mount, and utilizing a
Design and development of an active mount is special control strategy, provides the new mount with
described in details. The active mount is made by several advantages over a previously developed active

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

Mansour et al. 1231

mount. These advantages include: stability, cost effi- T transmitted

ciency, fail-safety and low power consumption. To a active
prove the ability of proposed mount for real applica- o overall
tions, several control strategies have been used to define p passive
the magnitude and phase of the dynamic stiffness. Each
of the control functions are designed to address the
isolation requirements of a specific driving condition. References
These conditions include: no load, accessory load, and
Ahmed A, Haque M and Rakheja S (1999) Nonlinear analy-
cruising in different speed ranges and road conditions.
sis of automotive hydraulic mounts for isolation of vibra-
The result of the experiments proved high degrees of tion and shock. International Journal of Vehicle Design 22:
flexibility and adaptability for the developed engine 116–128.
mount to satisfy the requirements both on stiffness Aoki K, Shikata T, Hyoudou T, Hirade T and Aihara T
and damping. (1999) Application of an active control mount (ACM)
for improved diesel engine vehicle quietness. SAE
Funding Transactions 108(3): 1153–1158.
This work was supported by the Canadian Auto 21 Center of Arzanpour S and Golnaraghi M (2008a) A novel semi-active
Excellence [grant number], Natural Sciences and Research magnetorheological bushing design for variable displace-
Council of Canada [grant number] and GM Canada [grant ment engines. Journal of Intelligent Material Systems and
number]. Structures 19: 989–1003.
Arzanpour S and Golnaraghi M (2008b) Development of a
bushing with an active compliance chamber for variable
Acknowledgements displacement engines. Vehicle System Dynamics 46:
The authors acknowledge the technical and fabrication sup- 867–887.
port provided by Mr Bernie Rice and Mr Joe Mihalic from Brach R and Haddow A (1993) On the dynamic response of
Cooper Standard Automotive, located in Mitchell, Ontario. hydraulic engine mounts. SAE Technical Paper Series
#931321.
Notation Flower W (1985) Understanding hydraulic mounts for
improved vehicle noise, vibration and ride qualities. SAE
Ap equivalent piston area of the rubber Technical Paper Series #850975.
Aa effective area of the actuator Foumani M, Khajepour A and Durali M (2003) Application
I fluid inertia of sensitivity analysis to the development of high perfor-
R fluid resistance mance adaptive hydraulic engine mounts. Vehicle System
C lumped compliance Dynamics 39: 257–278.
F resistance force Fursdon PMT, Harrison AJ and Stoten DP (2000) The design
K equivalent spring constant and development of a self-tuning active engine mount.
P pressure In: Proceedings of IMechE 2000 Conference Transaction.
Q flow pp. 21–32.
Golnaraghi M and Jazar N (2001) Development and analysis
s Laplace transformation variable
of a simplified nonlinear model of a hydraulic engine
t time
mount. Journal of Vibration and Control 7: 495–526.
x mount displacement Hagino Y, Furuishi Y, Makigawa Y and Kumagai N (1986)
Y actuator displacement Active control for body vibration of FWD car. SAE
Z control function of the actuator Technical Paper Series #860552.
H transfer function between actuator displace- Hideki M, Mikasa T and Nemoto H (2004) NV countermea-
ment and active force sure technology for a cylinder-on-demand engine: devel-
G transfer function between mount displacement opment of active control engine mount. SAE Transactions
and passive force 113(6): 18–186.
U transfer function between overall force and Hillis AJ, Harrison L and Stoten DP (2005) A comparison of
mount displacement two adaptive algorithms for the control of active engine
mounts. Journal of Sound and Vibration 268: 37–54.
Hong S, Choi S, Jung W, Ham I and Kim D (2001) Vibration
Subscripts control of an ER mount subjected to high static loads.
Journal of Sound and Vibration 242: 740–748.
1 upper chamber Kinchen D and Hogland D (2006) GM Future P/T Mount
2 lower chamber requirements. Cooper standard technical documents.
d decoupler Lee B and Lee C (2009) Model based feed-forward control of
i inertia track electromagnetic type active control engine-mount system.
r rubber Journal of Sound and Vibration 323: 574–593.

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013

1232 Journal of Vibration and Control 18(8)

Liang T, Xin-long T and Shi W (2010) Study on electromag- engine mount system. Vehicle Dynamic Systems 32:
netic actuator active engine mount with adaptive feed-for- 185–198.
ward control. In: Proceedings of Informatics in Control Ozaki H, Tsukamoto T, Ichikawa A, Yamazoe H, Shibata A,
Automation and Robotics (CAR) 2nd International Asia Maeno T, et al (1999) Development of active engine
Conference. vol. 1, pp. 112–115. mount. In: Proceedings of JSAE Annual Congress.
Little E and Kashani R (1995) Adaptive-passive and active pp. 141–144.
hydraulic engine mounts. Advanced Automotive Swanson D (1993) Active engine mounts for vehicles. SAE
Technologies ASME DSC-vol.56/DE-vol.86 135–140. Technical Paper Series #932432.
Mansour H (2010) Design and development of active and Turnip A, Hong K and Park S (2009) Modeling of a hydraulic
semi-active engine mounts. Master’s Thesis, Simon engine mount for active pneumatic engine vibration con-
Fraser University Burnaby, Canada. trol using the extended Kalman filter. Journal of
Mansour H, Arzanpour S and Golnaraghi M (2010a) Active Mechanical Science and Technology 23: 229–236.
decoupler hydraulic engine mount design with application Ushijima T and Kumakawa S (1993) Active engine mount
to variable displacement engine. Journal of Vibration and with piezo-actuator for vibration control. SAE Technical
Control (accessed 12 November 2010). Paper Series #930201.
Mansour H, Arzanpour S, Golnaraghi M and Parameswaran Williams E, Rigby S, Sproston J and Stanway R (1993)
A (2010b) Semi-active engine mount design using auxiliary Electrorheological fluids applied to an automotive engine
magneto-rheological fluid compliance chamber. Vehicle mount. Journal of Non-Newtonian Fluid Mechanics 47:
System Dynamics 49(3): 449–462. 221–238.
Miller L and Ahmadian M (1992) Active mounts-a discussion Yu Y, Naganathan N and Dukkipati R (2001) A literature
of future technological trends. Internoise 139: 421–421. review of automotive vehicle engine mounting systems.
Nakaji Y, Satoh S, Kimura T, Hamabe T, Akatsu Y and Mechanism and Machine Theory 36: 123–142.
Kawazoe H (1999) Development of an active control

Downloaded from jvc.sagepub.com at UNIV OF NEW HAMPSHIRE on March 3, 2013