TEHNOMUS - New Technologies and Products in Machine Manufacturing Technologies

PASIVE SUSPENSION MODELING USING MATLAB, QUARTER CAR

MODEL, IMPUT SIGNAL STEP TYPE

Andronic Florin 1, Manolache-Rusu Ioan-Cozmin 2, Pătuleanu Liliana 3

1, 2, 3
„Ștefan cel Mare” University of Suceava, 13 Universității, 720229, Suceava, Romania,
[email protected]

Abstract: The purpose of a vehicle suspension system is to improve ride comfort and road handling. In
current article is simulated and analyzed the handling and ride performance of a vehicle with passive
suspension system, quarter car model with two degree of freedom. Since, the equations of the system cannot be
solved mathematically has developed a scheme in Matlab Simulink that allows analyzing the behavior of the
suspension. The schema that was created in Matlab Simulink, were compared with the State space model and the
Transfer function. After completing the simulation scheme can be introduced excitation signals, this case a step
signal.

Keywords: Automotive Suspension, Matlab, Transfer function, State space model

1. Introduction frequency of the actuator[11]. Passive system are

the most common.
The vehicle suspension system differ So far, several models have been developed [3, 7-
depending on the manufacturer which ensures a 10], such as quarter car, half car or full car
wide range of models. Whichever solution is suspension. The following references will be made
adopted to design, a suspension system has the to the model suspension quarter car, passive
primary role to ensuring the safety function. It is suspension system.
known that road unevenness produce oscillations
of the vehicle wheels which will transmitted to 2. The mathematical model of quarter car
their axles. It becomes clear that the role of the suspension
suspension system witch connect the axles to the The system shown in Fig.1 is an quarter car
car body is to reduce as much vibrations and system were m1 - is the sprung mass, m2 - is the
shocks occurring in the operation. This causes, the unsprung mass, k1 - is the stiffness coefficient of
necessity to using a suspension of a better quality. the suspension, k2 - is the vertical stiffness of the
A quality suspension must achieve a good tire, b1 - is the damping coefficient of the
behavior of the vehicle and a degree of comfort suspension, b2 - is the damping coefficient of the
depending on the interaction with uneven road tire, x1 - the vertical displacement of sprung mass,
surface [1, 6]. When the vehicle is requested by x2 - is the vertical displacement of unsprung mass,
uneven road profile, it should not be too large w - is the road excitation. We will consider only
oscillations, and if this occurs, they must be mass movements on the vertical axis ignoring the
removed as quickly. The design of a vehicle rotational movement of the vehicle.
suspension is an issue that requires a series of Since the distance x1–w is hard to measure
calculations based on the purpose. and the deformation of the tire x2–w is negligible,
Suspension systems are classified in the result that we can use as an input size, the
well-known terms of passive, semi-active, active displacement x1–x2 against which we will analyze
and various in between systems. Typical features the behavior of the suspension system.
are the required energy and the characteristic

258
 TEHNOMUS - New Technologies and Products in Machine Manufacturing Technologies

performed using Simulink library blocks

computing.
In order to analyze the behavior of the
quarter car suspension system were used as input
parameters the next values:
m1 = 466.5 kg;
m2 = 49.8 kg;
k1 = 5700;
k2 = 135000;
b1 = 290;
b2 = 1400;
ki = 5.52;
kd = 10.0;
kp = 0.552
corresponding to the above equations.
Figure 1 The passive suspension system

The equations of motion can be obtained using

the Newton's second law for each of the two
masses are in motion and Newton's third law of
their interaction . These will be:

x1  b1  x1  x2   k1  x1  x2   0
m1  (1
) Figure 2 Matlab representation of the equation
(3)
x2  b1  x1  x2   k1  x2  x1   b2 x2  k2 x2  b2 w(2 k2 w
m2 
)
Separate terms x1 and x2 :

x1  b1  x1  x2   k1  x1  x2 

m1  (3
)

x2  b2 w  k2 w  b1  x1  x2   k1  x2  x1   b2 x2 (4

m2   k2 x2
)
Equations (3) and (4) is a second-order
differential equations of a passive suspension
system. Solving this systems of equations is
Figure 3 Matlab representation of the equation
difficult so we can use Matlab Simulink software.
(3) and (4)
Solving the system and its verification will be done
by three methods:
Given the input parameters that we run the
- write the equations in Matlab using
program and the graphical representation of the
Simulink library blocks; system of equations, we can calculate the mass
- using the "transfer function"; movements of the vehicle and its suspension. In
- using the "state-space" model. Fig. 3 is presented car body displacement, it is
noted that the excitation signal on mass m1
3. Graphic representation using Matlab displacement typically occurs over a period of 15s
library blocks after the oscillation amplitude is zero. In terms of
In Fig. 2 is represented the equation (3) and comfort, passenger severe amplitudes persist
in Fig.3 to equation (3) is added the equation (4), during the 10s minimal except that the first
now the system is completely. Solving was oscillations are strongest.

259
 TEHNOMUS - New Technologies and Products in Machine Manufacturing Technologies

m1 s 2 x1  s  

 x1  s 
m s 1
2

 b1 s  k1 m2 s 2  b2 s  k 2   (2)
b1 s  k1
 w  s  b2 s  k 2 

The transfer functions for the displacement of

m1 and m2 are:
x1  s 
H1  s   (3)
Figure 3 Car body displacement (unsprung ws
mass) x1
x2  s 
H2 s  (4)
ws
Using equations (8), (9) and (10) can be
determined the transfer function H1(s):
 b2 s  k2  b1s  k1 
H1  s  
m1 s 2 B1  B2 B3
B1   b1 s  k1  (5)
Figure 4 Suspension displacement (sprung mass)
x2 
B2  m1 s  b1 s  k1
2

B3  m s 2
2
b sk 
2 2
4. Graphic representation using Transfer
If the variable s is separated, the transfer
function
function becomes:

To verify the results from Fig.3 and Fig. 4 b1b2 s 2   k1b2  b1k 2  s  k1k 2
we move on to the second method of calculation, H1  s  
m1m2 s 4  A1 s 3  A2 s 2  A3 s  k1k2
using the "transfer function". Solving the system of
equations (1) and (2) we will do, by using the A1   m1b1  m1b2  m2 b1  (12)
Laplace transform, switching the original into a A2   m1k1  m1k2  m2 k1  b1b2 
image function of a complex argument s.
A3   b1k 2  k1b2 
The system of equations is:

m1 x1 s 2  b1  x1  x2  s  k1  x1  x2   0 (5) Equation (12) we write in Matlab using

block "transfer function", its representation is in
Figure 5.
m2 x2 s 2  b1  x2  x1  s  k1  x2  x1   (6)
b2  x2  w  s  k2  x2  w   0

From the first equation (5) of the system is

obtained:
m1 s 2  b1 s  k1
x2 ( s )  x1( s ) (1)
b1 s  k1
If you add the equations (5) and (6) of the
system and x2(s) will be replaced, will obtains:
Figure 5 Transfer function

260
 TEHNOMUS - New Technologies and Products in Machine Manufacturing Technologies

X  A  t  X  t   B  t U  t 
(13)
Y  t   C  t  X  t   D  t U  t 
where, X(t) is the " state vector", Y(t) - is the
"output vector", U(t) - is the "input (or control)
vector", A(t) - is the "state (or system) matrix", B(t) -
is the "input matrix", C(t) - is the "output matrix",
D(t) - is the "direct transmission matrix".
The matrix X(t) contains the following
variables:
 x1 
 
x
X t    2  (14)
 x1 
 
 x2 
Figure 6 Transfer function parameters
from equations (5), (6), (13) and (14) "state
After entering the parameters in this space" matrixes of the system will be:
function and initial parameters of the
 b1 b1 k1 k1 
suspension system described in Fig. 1 and run  m m1

m1 m1 
 1 
the program we get x1 moving vehicle mass as  b1 b  b2 k1 k k 
X  t     1
 1 2  X t  
it shown in Fig. 7. We note that the  m 2 m 2 m2 m2 
displacement calculated by the "transfer  1 0 0 0 
 
function" is identical to that in Fig. 3, is the  0 1 0 0 
initial displacement calculated by solving the  0 0 
 
system of equations with Matlab block  2b k 2 
 w 
diagram.   m2 m2   
  w
Using the equation (10) we can calculated  0 0  
 0 0 
the suspension mass displacement in the same
way as we calculated the car mass  x1 
 
displacement.  x2  
 x1  x2 
0 0 1 0  0 0 
     w 
 0 0 0 1  X  t   0 0   
 0 0   
w
 0 0 1 1
(15)

Matlab representation of equations (15) are

made using a predefined block where are
introduced the data from matrices.

Figure 7 Car body displacement (unsprung

mass) x1 calculated with transfer function

5. Graphic representation using State-space

model

Another method of modeling the suspension

system whit Matlab software is using the general
Figure 8 State-space model
form of the State-space model:

261
 TEHNOMUS - New Technologies and Products in Machine Manufacturing Technologies

W is the excitation signal representing the

road surface, in our case is a step type signal,
its parameters are represented in Fig. 12.

Figure 9 State-space model parameters

Figure 12 Step signal parameters

6. Conclusions

In this paper we simulated a passive

suspension system, a system that is most
Figure 10 Car body displacement (unsprung commonly used.
mass) x1 calculated with State-space model To verify the accuracy of modeling were
used State Space and Transfer Function.
In Fig. 10 the yellow line from the graph Results obtained, using the three methods
above, represent the vehicle mass with the same parameters of the suspension
displacement x1, the displacement is identical system, are identical.
to those shown in Fig. 3 and Fig. 7. The purple Was used a step type signal for a broad
line is the suspension mass displacement x2, application of the suspension system. This signal
this is similar to the one shown in Fig. 4. The can be modified, for example, a sinusoidal or a
blue line represent the difference x1-x2. required signal.
The parameters of a passive suspension
system are generally fixed, being chosen to achieve
a certain level of compromise between road
holding, load carrying and comfort.

ACKNOWLEDGMENT
This paper was supported by the project
"Improvement of the doctoral studies quality in
engineering science for development of the
knowledge based society-QDOC” contract no.
POSDRU/107/1.5/S/78534, project co-funded by
the European Social Fund through the Sectorial
Operational Program Human Resources 2007-
2013.
Figure 11 Matlab representation of the entire
system

262
 TEHNOMUS - New Technologies and Products in Machine Manufacturing Technologies

References:
[6]. Augustaitis, V. K., Gičan V., Šešok N., Iljin I.
[1]. Kamalakannan, K., ElayaPerumalb A.,
Computer-aided generation of equations and
Mangalaramananc, S., Arunachalamd K.
structural diagrams for simulation of linear
Performance Analysis and Behaviour
stationary mechanical dynamic systems.
Characteristics of CVD (Semi Active) in
ISSN 1392-1207.-Mechanika 1.-Kaunas:
Quarter Car Model. ISSN 1995-6665.-Jordan
Technologija, 2011. p.255-263.
Journal of Mechanical and Industrial
[7]. Poussot-Vassal, C., Sename, O., Dugard, L.,
Engineering, Volume 5, Number 3.- Zarqa,
Gáspár, P., Szabó, Z., Bokor J. A LPV based
Jordan.-The Jordanian Ministry of Higher
semi-active suspension control strategy. 3rd
Education and Scientific Research: Hashemite
IFAC Symposium on System, Structure and
University, 2011. p.261-265.
Control, SSSC 07.- Iguaçu Brésil, 2007.
[2]. Senthil kumar, M., Vijayarangan, S.
p.106-111.
Analytical and experimental studies on active
[8]. Ram Mohan, T., Venkata G., Sreenivasa, K.,
suspension system of light passenger vehicle
Purushottam A. Analysis of passive and semi
to improve ride comfort. ISSN 1392-1207.-
active controlled suspension systems for ride
Mechanika 1.-Kaunas: Technologija, 2007.
comfort in an omnibus passing over a speed
p.34-41.
bump. ISSN: 2076-734X, EISSN: 2076-
[3]. Guglielmino, E., Sireteanu T., Stammers W.
7366.-International Journal of Research and
C., Ghita, G., Giuclea M. Semi-active
Reviews in Applied Sciences Volume 5, Issue
Suspension Control - Improved Vehicle Ride
1. Islamabad: Academic Research Publishing
and Road Friendliness.- London: Springer.
Agency, 2010. p.7-17.
2008.-294p.
[9]. Goncalves, D.F. Dynamic analysis of semi-
[4]. De Man, P., Lemerle, P., Mistrot, P.,
active control techniques for vehicle
Verschueren, J-Ph., Preumont, A. An
applications.-Virginia: Thesis submitted to
investigation of a semi-active suspension for a
the Faculty of the Virginia Polytechnic
fork lift truck. ISSN 0042-3114.-Vehicle
Institute and State University. 2001.-180p.
System Dynamics, Volume 43, Number 2.-
[10].Segla, S., Reich S. Optimization and
London: Taylor and Francis Ltd, 2005. p.107-
comparison of passive, active, and semi-active
119.
vehicle suspension systems.-12th IFToMM
[5]. Giua, A., Melas, M., Seatzu, C., Usai, G.
World Congress.-Besançon (France), 2007.-
Design of a predictive semiactive suspension
6p.
system. ISSN 0042-3114.-Vehicle System
[11].Streiter, R. Entwicklung und realisierung
Dynamics, Volume 41, Number 4.-London:
eines analytischen regelkonzepts fur eine
Taylor and Francis Ltd, 2004. p.277–300.
aktive Federung. Dissertation, TU, Berlin,
1996.

263