International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 – 0882

357
Vol.4, Issue 4, April 2015

Car Dynamics using Quarter Model and Passive Suspension; Part V:

Frequency Response Considering Driver-seat

Galal Ali Hassaan

Emeritus Professor, Department of Mechanical Design & Production,
Faculty of Engineering, Cairo University, Egypt

that the sprung mass transfer function had a distinct

ABSTRACT sensitivity to a pavement roughness frequency of 3.5 Hz
The objective of the paper is to investigate the dynamics
[1]. Sun (2001) presented a theoretical model for
of a quarter-car model with passive suspension and
considering the driver-seat. The model is a 3 DOF one determining roughness index. He obtained the frequency
excited by the irregularities of the road pavement. The response of quarter-car model for frequencies up to 23
model dynamics are studied in terms of the frequency Hz [2]. Lauwerys, Swevers and Sas (2002) discussed the
response of the driver using its transmissibility. The development of a controller for a passenger car using
paper presents new technical terms used to assess the
active shock absorber through varying the damping
suspension parameters which are the accumulated and
mean transmissibility. The driver frequency response is characteristics with controlled valves. They tested the
studied for road pavement of irregularities having proposed controller using a quarter-car test rig. They
harmonic components of frequencies up to 10 Hz. The demonstrated the frequency response of the system for
natural frequencies of the system are calculated and the frequencies up to 30 Hz [3]. Zuo and Nayfeh (2003)
driver transmissibility is evaluated for suspension
parameters in the range 2.85 to 11.4 kN/m for stiffness considered road random excitations as input to an 8 DOF
and 0.5 to 8 kNs/m for damping coefficient. full-car model using passive suspensions. They
considered the optimization of passive suspensions as
Keywords - Car dynamics; quarter car model; passive structure-constrained LQG/H2 optimal control problem.
suspension; frequency response, driver transmissibility.
They demonstrated the vehicle frequency response for
I. INTRODUCTION exciting frequencies up to 100 Hz [4].

Lu (2004) studied using the frequency-adaptive

The study of car dynamics has special importance since
multi-objective suspension strategy for vehicle
it helps car designers to produce vehicles capable of
suspension control. He used a quarter-car model and
achieving ride comfort to passengers according to
presented the frequency response of the system for
international standards. On the other hand, such studies
exciting frequencies up to 20 Hz [5]. Melcer (2006) used
help authorities dealing with the requirement of speed
a quarter-car model in the solution of some problems of
control in residential districts or for security reasons to
vehicle-road interaction in the frequency domain. He
design speed humps properly without harming the
calculated the frequency response in terms of amplitude
vehicles or passengers. Frequency response is one of the
and phase with frequencies up to 100 rad/s [6]. Batterbee
dynamic means of investigating car dynamics.
and Sims (2007) used skyhook damping laws within
Papagiannakis and Gajarathi (1995) proposed a primary automotive suspension. They used a 2 DOF
pavement roughness statistics based on the vertical quarter-car model excited by realistic road profiles. They
sprung acceleration. They simulated the sprung mass estimated the frequency response of the system using
acceleration using a quarter-car model. They showed exciting frequencies up to 15 Hz [7]. Chi, He and

www.ijsret.org
 International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 – 0882
358
Vol.4, Issue 4, April 2015

Naterer (2008) applied three optimization algorithms for II. THE QUARTER-CAR MODEL
the design optimization of vehicle suspensions based on
a quarter-car model. The investigated the effect of car The quarter car model is used extensively in studying the
speed and road irregularity on the design variables to vehicle dynamics. Large number of researchers consider
improve ride quality. They covered frequencies up to 48 this model as a 2 DOF one by considering only the tire-
Hz in the frequency response of the vehicle [8]. Santos wheel assembly and the sprung mass. Here in this work I
et. al. (2010) modeled a MR damper as a component in a consider it as a 3 DOF by considering the driver and his
quarter-car model. The presented the frequency response seat. The line diagram of the 3 DOF quarter-car model is
of the sprung mass for frequencies up to 20 Hz [9]. shown in Fig.1.

Barbosa (2011) used a quarter-car model to identify

the vehicle dynamic behavior. He derived the vehicle
inertance function using the vehicle force transfer
function. He obtained the vehicle frequency response for
frequencies up to 100 Hz [10]. Slaski (2012) used signal
processing to estimate the power spectral density and the
magnitude characteristics of a car suspension. He used a
quarter-car model subjected to input excitation with
frequencies up to 25 Hz [11]. Patil, Pawar and Patil
(2013) reviewed the research work regarding vehicle's
passive and active suspensions. They presented the Figure 1 3 DOF quarter-car model.
vehicle frequency response for exciting frequencies up to
100 Hz showing resonance at frequencies less than or The model of Fig.1 consists of:
equal 10 Hz [12]. Razdan, Bhave and Awasare (2014) - One tire and wheel and its attachments having the
studied the active pneumatic suspension with control parameters: mass (m1), damping coefficient (c1)
strategy based on mass flow control for a small car.
and stiffness (k1).
They concluded that the active suspension using velocity
- Sprung mass which is one quarter of the car
feedback provided less transmissibility at resonance
chassis having the parameters: mass (m2),
compared to the passive suspension. They used
damping coefficient (c2) and stiffness (k2). c2 and
frequencies up to 25 Hz in their presentation of the car
k2 are the suspension parameters.
frequency response [13]. Saglam and Unlusoy (2014)
- The driver and its seat having the parameters:
developed an active hydro-pneumatic suspension
mass (m3), damping coefficient (c3) and stiffness
system. They used a quarter-car model and derived a
(k3). c3 and k3 represent the seat damping and
nonlinear model with state dependent matrices. They
elasticity.
presented the frequency response of the system for
passive and active suspensions for frequencies up to 20 III. MODEL PARAMETERS
Hz. Their system showed resonance near 1 and 9 Hz
[14]. Hassaan and Mohammed (2015) used a 10 DOF The 3 DOF quarter-car model parameters are borrowed
full-car model to investigate the frequency response of from more than one reference. The parameters of the tire
the vehicle against road disturbance. Their frequency and sprung mass are used by Florin, Ioan-Cosmin and
response covered frequencies up to 18 Hz [15]. Liliana [16], while the parameters of the driver and seat are

www.ijsret.org
 International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 – 0882
359
Vol.4, Issue 4, April 2015

used by Guclu [17]. The 3 DOF quarter-car model m1x1'' + (c1+c2)x1' – c2x2' + (k1+k2)x1 – k2x2 = c1y' +k1y
parameters used in this analysis work are given in Table 1. (1)

Table 1 3 DOF model parameters For the sprung mass, m2:

Parameters Description Value m2x2'' – c2x1' + (c2+c3)x2' – c3x3' – k2x1 + (k2+k3)x2 – k3x3 = 0
k1 (kN/m) Tire stiffness 135
c1 (kNs/m) Tire damping 1.4 (2)
coefficient
m1 (kg) Un-sprung mass 49.8 For the driver and seat mass, m3:
k2 (kN/m) Suspension stiffness 2.85, 5.7, 11.4
m2 (kg) Sprung mass 466.5 m3x3'' – c3x2' + c3x3' – k3x2 + k3x3 = 0 (3)
k3 (kN/m) Seat stiffness 15
c3 (Ns/m) Seat damping 150 Equations 1,2 and 3 can be written in the matrix form:
coefficient
m3 (kg) Driver-seat mass 90 Mx'' + Cx' + Kx = F (4)

The damping coefficient of the suspension, c2 is changed Where: M, C and K are the system mass, damping and
in the range: stiffness matrices of the 3 DOF system given respectively
by:
0.5 ≤ c2 ≤ 8 kNs/m m1 0 0
M = 0 m2 0
0 0 m3
The purpose of this change is investigate its effect of
the frequency response of the sprung mass due to the c1+c2 –c2 0
irregularities of the road. C= -c2 c2+c3 -c3 (5)
0 -c3 c3

IV. MATHEMATICAL MODEL k1+k2 –k2 0

K= -k2 k2+k3 -k3
The mathematical model of the 3 DOF quarter-car model 0 -k3 k3
is defined by three differential equations obtained by
The response vector x is:
applying Newton's third law to the free body diagram of x = [x1 x2 x3]T
the three lumped masses of Fig.1. The dynamic motions of The force vector F in Eq.4 is:
the system are: x1, x2 and x3. The irregularities of the road F = [F1 0 0]T (6)
is defined by the dynamic motion y which is assumed Where F1 is the force acting on the tire and wheel mass as
sinusoidal. Of course, the irregularities are random. The transmitted from the road through the viscoelastic
frequency contents of this irregularity can be defined using characteristics of the tire. That is:
FFT (fast Fourier transform). Thus, the sinusoidal input
motion y represents one component in this transform. All F1 = c1y' + k1y (7)
the parameters of the system are passive and assumed
For a harmonic exciting force component of angular
having the constant values presented in section 3. The
frequency ω, y takes the form:
differential equations are written in standard form as
follows: y = Y ejωt (8)

For the tire and wheel mass, m1: Combining Eqs.7 and 8 gives:

F1 = Y(k1 + jc1ω) ejωt (9)

www.ijsret.org
 International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 – 0882
360
Vol.4, Issue 4, April 2015

Where Y is the peak amplitude of the harmonic The MATLAB command 'eig' is used to extract the
component of the road irregularity. system natural frequencies using Eq.13 as an eigenvalue
problem [19]. The system natural frequencies depend on
The steady state solution of Eq.4 is:
the suspension stiffness as given in Table 2.
x = X ejωt (10) Table 2 Quarter-car model natural frequencies
Where X is the phasor of the amplitude vector of the three Suspension stiffness, 2.85 5.70 11.40
motions. k2 (kN/m)
First natural 0.3555 0.4964 0.6847
Now, combining Eqs.4 and 10 gives the amplitude phasor frequency (Hz)
X as: Second natural 2.2497 2.2553 2.2666
frequency (Hz)
X = (K – ω M + jωC) F
2 -1
(11) Third natural 8.3737 8.4604 8.6322
frequency (Hz)
Solving Eq.11 for any frequency range provides the
frequency response for the three masses of the system. VI. DRIVER TRANSMISSIBILITY
It is possible to study the frequency response of the
quarter-car model in terms of the transmissibility. This can The driver displacement transmissibility provides the
be done as follows: driver motion amplitude relative to the ground motion
amplitude at a specific frequency. The isolators have to
- The transmissibility here is a displacement
maintain this value less than unit value for proper
transmissibility defined as X/Y [18].
suspension design. This depends mainly on the parameters
- The phasor of the transmissibility TR is obtained
of the suspension elements and the exciting frequency.
by combining Eqs.6,9 and 11 as:
Figs.2, 3 and 4 show the frequency response of the driver
TR = (K – ω2M + jωC)-1 [k1 + jc1ω) 0 0]T (12)
as function of the three parameters, exciting frequency f,
- The transmissibility of any of the three masses damping coefficient c2 and stiffness k2. The three figures
will be the magnitude of its corresponding phasor. are generated by MATLAB using Eq.12 and the
- In the present study, the transmissibility of the parameters in Table 1.
driver is investigated.

V. QUARTER-CAR MODEL NATURAL

FREQUENCIES

The evaluation of the natural frequencies of any dynamic

system is of great importance since they assign the
location of resonance during harmonic excitation of the
system. The analyzed system has three degrees of
freedom. Then, it has three natural frequencies. They are
obtained using the equation:

(K – ωn2M)X = 0 (13) Figure 2 Driver transmissibility for k2 = 2.85 kN/m.

www.ijsret.org
 International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 – 0882
361
Vol.4, Issue 4, April 2015

- After the second intersection the damping effect is

reversed. That is decreasing the damping
coefficient will decrease the transmissibility
which this process is reversed between the first
and second intersections.

VII. DRIVER MAXIMUM

TRANSMISSIBILITY

- The driver transmissibility is maximum near

resonance.
- It may be maximum near the first or second
resonance as clear in Figs.2-4.
Figure 3 Driver transmissibility for k2 = 5.70 kN/m. - The location of the maximum transmissibility
depends on the suspension parameters c2 and k2.
- Fig.5 gives the maximum transmissibility of the
driver against c2 and k2 (transmission parameters).

Figure 4 Driver transmissibility for k2 = 11.40 kN/m.

- Figs.2,3 and 4 depict the driver transmissibility for

frequencies between 0 and 10 Hs, suspension Figure 5 Driver maximum transmissibility.
damping coefficient between 1 and 5 kNs/m and
- It decreases to minimum as the damping
suspension stiffness between 2.85 and 11.40
coefficient increase, then increases again.
kN/m.
- For the damping range 0.5 ≤ c2 ≤ 3.5 kNs/m,
- Two resonance values appear in the three
increasing the damping coefficient at a specific
transmissibility plots which are the first and
exciting frequency increases the maximum
second resonance.
transmissibility.
- The third resonance does not appear.
- For the damping coefficient range 3.5 ≤ c2 ≤ 8
- All the transmissibility curves intersect at 2 points
kNs/m, the maximum transmissibility increases
at zero frequency and a t a frequency between the
almost linearly with c2 independent of the
first and second resonance.

www.ijsret.org
 International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 – 0882
362
Vol.4, Issue 4, April 2015

suspension stiffness k2. value to a maximum value, then decreases (or

decreases, increases and decreases) again to
VIII. ACCUMULATED DRIVER almost zero value.
TRANSMISSIBILITY - Then, the mean value is used by the author as
another way for assessing the effect of the
- This is a new function introduced by the author to suspension parameters c2 and k2.
assess the effectiveness of the transmission - The mean is simply evaluated using the
elements for the whole range under study (0 ≤ f MATLAB command 'mean' [20].
≤.fmax). - Fig.7 shows the variation of the transmissibility
- The accumulated transmissibility, TRacc is defined mean of the driver with the suspension parameters
as the area under the transmissibility curve. That c2 and k2.
is:
fmax
TRacc = ∫ TR df (14)
0
- MATLAB is used to evaluate the accumulated
transmissibility of the driver through using its
command 'sum' [20].
- The variation of the accumulated transmissibility
with the suspension parameters is shown in Fig.6
as generated by MATLAB.

Figure 7 Driver mean transmissibility.

- The mean transmissibility of Fig.7 has almost the

same pattern as the accumulated transmissibility.
- Thus, within the studied range of suspension
parameters (c2 and k2), decreasing the suspension
damping coefficient and stiffness will decrease the
mean transmissibility during the frequency range
from 0 to 10 Hz.
- The mean transmissibility is < 1 for all the
Figure 6 Driver accumulated transmissibility.
combinations of the suspension parameters.
- The accumulated transmissibility decreases as the
damping coefficient and stiffness of the X. MINIMUM EXCITING FREQUENCY
suspension elements decreases. FOR TRANSMISSIBILITY < 1

IX. MEAN DRIVER TRANSMISSIBILITY - In suspension design, it is desired to produce

- The driver transmissibility increases from a unit driver transmissibility < 1 to attenuate the road

www.ijsret.org
 International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 – 0882
363
Vol.4, Issue 4, April 2015

disturbance. transmissibility and mean transmissibility. The main

- This can be achieved by controlling the road conclusions of this research work are as follows:
pavement such that its irregularities do not have (1) The dynamic model of the 3 DOF dynamic
frequency components less than specific values. system was derived assuming passive parameters for all
- This is the minimum disturbance frequency, fmin. dampers and springs in the system.
- Again fmin depends on the suspension parameters (2) The frequency response of the quarter-car model
c2 and k2 as shown in Fig.8. was presented in terms of the driver transmissibility.
(3) New terms were introduced to assess the driver
vibrations due to road disturbance (accumulated and
mean transmissibility).
(4) The three natural frequencies of the system were
evaluated for different values of the suspension damping
coefficient and stiffness.
(5) Using MATLAB, the frequency response of the
driver in terms of its transmissibility was presented for
harmonic road frequencies up to 10 Hz , suspension
parameters of damping coefficient in the range 0.5 to 8
kNs/m and stiffness in the range 2.85 to 11.4 kN/m.
(6) The third natural frequency did not show any
resonance effect in the frequency response of the driver.
Figure 8 Minimum frequency for transmissibility < 1. (7) The effect of the suspension parameters on the
- For high level of damping coefficient (c2 > 2 maximum driver transmissibility, accumulated
kNs/m) the suspension stiffness almost has no transmissibility, mean transmissibility and minimum
effect on the minimum frequency and fmin varies road frequency for transmissibility less than one was
almost linearly with c2. demonstrated.
- For c2 < 2 kNs/m, fmin decreases as k2 decreases at (8) Decreasing the values of the suspension
a specific frequency. parameters decreased both accumulated and mean
- It is possible to go down with fmin to < 0.5 Hz at transmissibility's.
small levels of the suspension parameters (c2 and (9) The maximum transmissibility was greater than
k2). one for all the combinations of the suspension
parameters.
XI. CONCLUSION (10) It was possible to go down by the minimum
road frequency to less than 0.5 Hz with suspension
Including the driver-seat in the quarter-car model has
stiffness of 2.85 kN/m and damping coefficient ≤ 1
increased the degree of freedom (DOF) to 3. This
kNs/m.
resulted in three natural frequencies, i.e. three resonance
locations with possible increase of the frequency REFERENCES
response near those values. Two new parameters were
introduced to the terminology of frequency response of [1] T. Papagiannakis and M. Gujarathi, A roughness
dynamic systems which were the accumulated model describing vehicle-pavement interaction.
Report No. WA-RD 372.1, Washington State

www.ijsret.org
 International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 – 0882
364
Vol.4, Issue 4, April 2015

University, Pullman, Washington, February, 1995. – A review, International Journal of Engineering

[2] L. Sun, Developing spectrum-based models for Research and Applications, 3 (1), 2013, 774-777.
international roughness index and present sensitivity [13] S. Razdan, S. Bhave and P. Awasare, Comparison of
index, Journal of Transportation Engineering, quarter-car model of active pneumatic suspension
November/December, 2001, 463-470. using mass flow control for a small car,
[3] C. Lauwerys, J. Swevers and P. Sas, Linear control International Journal of Current Engineering and
of car suspension using nonlinear actuator control, Technology, Special Issue-2, February, 2014, 597-
Proceedings of ISMA, I, 2002, 55-62. 601.
[4] L. Zuo and S. Nayfeh, Structured H2 optimization of [14] F. Saglam and Y. Unlusoy, Modeling and state
vehicle suspensions based on multi-wheel models. dependent Riccati equation control of an active
Vehicle System Dynamics, 40 (5), 2003, 351-371. hydro-pneumatic suspension system, Proceedings of
[5] J. Lu, A frequency-adaptive multi-objective the International Conference of Control, Dynamic
suspension control strategy, Journal of Dynamic Systems and Robotics, Ottawa, Ontario, Canada,
Systems, Measurement and Control, September, May 15-16, Paper 31, 2014, 1-8.
2004, 700-707. [15] G. A. Hassaan and N. A. Mohammed, Frequency
[6] J. Meker, Vehicle-road interaction analysis in a response of 10 degree of freedom full-car model for
frequency domain, Slovak Journal of Civil ride comfort, International Journal of Scientific
Engineering, 3, 2006, 48-52. Research Engineering, 4(1), 2015, 43-49.
[7] D. Batterbee and N. Sims, Hardware in the loop [16] A. Florin, M. Joan-Cosmin and P. Lilian, Passive
simulation of MR dampers for vehicle suspension suspension modeling using MATLAB, quarter-car
systems, Journal of Systems and Control model, input signal step type, New Technologies
Engineering, 221(2), 2007, 265-278. and Products in Machine Manufacturing, January,
[8] Z. Chi, Y. He and G. Naterer, Design optimization of 2013, 258-263.
vehicle suspension with a quarter-vehicle model, [17] R. Guclu, Active control of seat vibrations of a
Transactions of the CSME, 32 (2), 2008, 297-312. vehicle model using various suspension alternatives,
[9] J. Santos, O. Sename, L. Dugand, R. Menendez. and Turkish Journal of Engineering Environmental
R. Mendoza, A semi-active control-oriented damper Science, 27, 2003, 361-373.
model for an automotive suspension, 6th IFAC [18] S. Rao, Mechanical vibrations, Prentice Hall, 2011.
Symposium on Advances in Automatic Control, July, [19] R. Dukkinpati, Solving vibration analysis problems
Munich, Germany, Paper 16, 2010. using MATLAB, New Age International, 2007.
[10] R. Barbosa, Vehicle dynamic safety in measured [20] M. Cho and W. Matinez, Statistics in MATLAB: A
rough pavement, Journal of Transportation, May, prime, Chapman & Hall / CRC, 2014.
2011, 305-310.
[11] G. Slaski, Experimental determination of suspension
magnitude-frequency responses using electro-
hydraulic actuators-testing and data processing
method, The Archive of Automotive Engineering, 58
(2), 2012, 63-75.
[12] V. Patil, G. Pawar S. Patil, A comparative study
between the vehicle's passive and active suspensions

www.ijsret.org
 International Journal of Scientific Research Engineering & Technology (IJSRET), ISSN 2278 – 0882
365
Vol.4, Issue 4, April 2015

BIOGRAPHY

Galal Ali Hassaan

 Emeritus Professor of System Dynamics and

Mechanical Vibrations.
 Has got his B.Sc. and M.Sc. from Cairo
University in 1970 and 1974.
 Has got his Ph.D. in 1979 from Bradford
University, UK under the supervision of Late
Prof. John Parnaby.
 Now with the Faculty of Engineering, Cairo
University, EGYPT.
 Research on Automatic Control, Mechanical
Vibrations , Mechanism Synthesis and History
of Mechanical Engineering.
 Published 10’s of research papers in
international journals and conferences.
 Author of books on Experimental Systems
Control, Experimental Vibrations and Evolution
of Mechanical Engineering.
 Chief Justice of IJCT.
 Reviewer in a number of International Journals.

www.ijsret.org