Proceedings of the 2008 International Conference on Electrical Machines

Paper ID 1285

Analysis of an Electromagnetic Automobile Suspension System

Miguel Chaves1,2,4, Jos Maia3,4, Jorge Esteves1,2,4
mchaves@deea.isel.ipl.pt jmaia@est.ips.pt jesteves@deea.isel.ipl.pt
(1) Instituto Superior de Engenharia de Lisboa, DEEA, R. Conselheiro Emdio Navarro 1, 1950-062 Lisboa, Portugal (2) Centro de Electrotecnia e Electrnica Industrial, R. Conselheiro Emdio Navarro 1, 1950-062 Lisboa, Portugal (3) Escola Superior de Tecnologia de Setbal, DEE, Campus do IPS, 2910-761 Setbal , Portugal (4) CIEEE, Av. Rovisco Pais 1049-001 Lisboa, Portugal

Abstract-The automobile suspension systems currently in use can be classified as passive, semi-active and active. The passive suspension systems are the most commonly used due their low price and high reliability. However, this system cannot assure the desired performance from a modern suspension system. An important improvement of suspension performance is achieved by the active systems. Nevertheless, they are only used in a very reduced number of automobile models because they are expensive and complex. Another disadvantage of active systems is relatively high energy consumption. The use of electromagnetic linear actuators is an alternative for the implementation of active automobile suspensions systems. Moreover, this solution has the advantage of the suspension energy recovery. The authors workgroup have proposed an electromagnetic actuator that is already implemented in a workbench. The prototype can be improved in order to obtain higher power with less volume, with the development of new materials. In this paper it is analysed a pseudo-levitating suspension system. Simulation and experimental results are also presented. Keywords-Automobile suspension, electromagnetic linear actuator, pseudo-levitating suspension.

ms Ks mu Kt
a)

Zs

ms Ks Bs mu Kt
b)

Zs

ms Ks Bs mu Kt
c)

Zs

Bs
Zu

Fa
Zu

Zu

Zr

Zr

Zr

Fig. 1. Commonly used suspension systems: a) Passive; b) Semi-active; c) Active. One possible solution that has been proposed by [1] to solve these difficulties is to use electromagnetic actuators in order to improve the performance of suspension systems without increasing the energy consumption and costs. All referred types of suspension systems could be built using electromagnetic actuators. Other advantage of electromagnetic suspensions is the possibility to work in the generator mode operation, this characteristic allows recovering energy from the vehicle body oscillations. In order to experimental test the mathematical models a workbench was implemented in the laboratory. A sprung mass, an electromagnetic actuator, a spring, a wheel, and a road perturbation generator compose the suspension system prototype. The analysed suspension system was introduced as a pseudo-levitating suspension system [1], and it uses an electromagnetic actuator and a spring between the sprung and the unsprung masses. The spring supports the body weight and the electromagnetic actuator is controlled in order to compensate the spring deformation force. With this topology smaller power electromagnetic actuators can be used. II. PSEUDO-LEVITATING SUSPENSION SYSTEM A. Pseudo-Levitating Suspension Model The suspension system analysis is made considering the representation of a quarter of the sprung mass ( of vehicle model) and considering vehicles with independent suspensions. The model of of vehicle is presented in figure 2.

I.

INTRODUCTION

The main function of vehicle suspension system is to isolate the vehicle body and passengers from the oscillations created by the road irregularities and to produce a continuous road-wheel contact. Three types of vehicle suspensions are commonly used: passive, semi-active and active. All the systems known as implemented in automobiles are based in hydraulic or pneumatic operation. However, it is verified that these solutions cannot solve satisfactory the vehicles oscillations problem or they are very expensive and contribute to the increasing of the energy vehicle consumption. The figure 1 presents the three types of suspension systems that are generally used, where ms represents the sprung mass, mu the unsprung mass, Ks the spring stiffness, BS the damping coefficient, Fa the actuator force, Kt the tyre stiffness, Zr the road position reference, Zu the unsprung mass position reference and Zs the sprung mass position reference.

978-1-4244-1736-0/08/$25.00 2008 IEEE

Proceedings of the 2008 International Conference on Electrical Machines

ms Ks mu Kt Fa
Zu

Zs

The suspension system can be written as the state space equation(5) and the output equation given by (6).
. 0 z s K .. s z s = ms . 0 zu Ks .. z u mu 1 0 Ks ms 0 0 (K + K t ) 0 s mu 0 0 z 0 s 1 0 . zs m + s 1 zu 0 . 1 0 z u mu 0 0 F a 0 zr Kt mu

(5)

Zr

a)

b)
K s .. ms z s ( zu zr ) = 0 ( z s zu ) 1 0 0 0 Ks 1 z 0 . s ms m zs s 1 0 + 0 z 1 0 . u 0 zu 0 Fa 1 z 0 r

Fig. 2. Pseudo-Levitating suspension system: a) physical model; b) Experimental test bench. From the model in figure 2a it is possible to write the system equations (1).
.. ' ' ms z s = K s z s z u + Fa ms g (1) .. ' ' mu z u = K s z 's z u Fa K t z u z 'r mu g

(6)

Considering the following variable change (2a) and (2b).

From the system (5) its possible to obtain the open loop transfer functions and compute the system poles that are the system critical frequencies. In this case the frequencies obtained are f1=1,4Hz and f2=10Hz, which are the typical critical frequencies of an automobile suspension system. The low frequency can be associated to the interaction between the sprung mass and the spring; the second one is the critical frequency referred to the tyre.

z s' = z s0 + z s ' 0 z u = z u + z u (2a) ' 0 zr = zr + zr

0 K s z s0 z u = ms g (2b) 0 K t z u z r0 = (ms + mu )g

B. Pseudo-Levitating Suspension Control Law Forcing to zero the sprung mass acceleration in the first equation it is obtained the ideal actuator force given by (8), which totally compensates the spring deformation.

( (

) )

z s = 0 (7)
Fa = K s (Z s Z u ) (8)
However with a force given by (8) the unsprung mass subsystem will be unstable because it will not have attenuation. In order to damp the unsprung mass oscillations near to its resonance frequency a gain k was introduced, the force equation becomes (9).

..

Rewriting the system (1) it is obtain the suspension system equation in (3).

.. m s z s = K s (z s z u ) + Fa (3) .. mu z u = K s ( z s z u ) Fa K t ( z u z r )
Choosing the state space, the output vector and input vector as (4a), (4b) and (4c) respectively.

Fa = K s (Z s Z u ) + k Z u (9)
This control law provides a minimal sprung mass acceleration with a safe tyre deflection. The value of k should be chosen according to the oscillations frequency. Rewriting the system (3) with the equation (9) it is obtained the close loop system (10).

xT = zs

zs

zu

. z u (4a)

yT = z s
..

zu z r

z s z u (4b)

u T = [Fa

z r ] (4c)

Proceedings of the 2008 International Conference on Electrical Machines

. .. m s z s = k z u (10) .. . m u z u = K t ( z u z r ) k z u

The simulation frequency response obtained is presented in figure 3, which shows the differences between the tree control laws.
10
3

7. The suspension system is controlled using a DSP, dSpace ACE1102 with the processor TMS320C31/60MHz. This controller receives the sensors signals, processes the information according to the specified control law and generates the actuator reference force. The suspension system experimental result with passive control law is shown in the figure 4, and the respective frequency spectrum in figure 5. Signals frequency spectrum analysis is a very important tool in order to better understand the results and synthesise the controller.
40 0.4 0.2 Zs [gs] 0
0.1 0 -0.1

(a)
10 Zs/Zr
2

(b) (c)
(Zs-Zu) [mm] 20 0

10

-20 -40 0
10 0 10
0

0.2 5 10 Time [s] 15 20 0.4 0 2 1 5 10 Time [s] 15

20

10

10 Frequency [rad/s]

10

500

0.2 0 -0.2

Figure 3: Frequency response: a) Passive b) Active c) Pseudo-Levitating.

Fa [N]

Zu [gs]

1 -500 0 5 10 Time [s] 15 20 20 5 10 Time [s] 15

4.5

5.5

20

III. SIMULATION AND EXPERIMENTAL RESULTS The pseudo-levitating workbench is composed components: suspension system by the following

Fig. 4. Experimental result for a passive control and a step in Bs [0 2000N/(ms-1) ]and a oscillation frequency of 2,3Hz.
6 0,04 0,035 0,03 Zs'' [g's] 4 3 2 1 0 0 5 10 15 20 25 Frequncia [Hz] Frequency [Hz] 0,025 0,02 0,015 0,01 0,005 0 0 5 10 15 20 25 Frequncia [Hz] Frequency [Hz]

1. The road perturbation generator, where the mechanical torque is provided by a 7,5kW ac motor coupled to a 1/6 gearbox. An inverter is used to power the motor. 2. The unsprung mass that includes the wheel and the respective support. The unsprung mass weight (mu) is 36,5kg. 3. The spring, installed between the sprung and the unsprung masses in order to support the sprung mass. The spring was dimensioned to have a stiffness (KS) of 16000Nm-1. 4. The electromagnetic linear actuator, which is a double phase, with permanent magnets, linear motor. A power electronic current controller converter feeds the linear motor that has a maximum force of 980N. 5. The sprung mass that represents a quarter of the vehicle body and passengers weight. Its value is 200kg. 6. The sensors, which are two accelerometers installed in the sprung (Zs) and unsprung masses (Zu) and one lvdt between the two masses in order to measure the (Zs-Zu) value.

5 (Zs-Zu) [mm]

160 140 120 Zu'' [Hz] 0 5 10 15 20 25 Fa [N] 100 80 60 40 20 0

0,07 0,06 0,05 0,04 0,03 0,02 0,01 0 0 5 10 15 20 25

Frequency [Hz] Frequncia [Hz]

Frequency [Hz] Frequncia [Hz]

Fig. 5. Frequency spectrum from the experimental results of fig. 4. In the above figure it is possible to see the frequencies in the experimental suspension system work bench, which are the typical frequencies around 1,4Hz and 10Hz. There are also other frequencies because of the system nonlinearities such as static and friction coefficient. However, there is a clearly benefit on the sprung mass acceleration, that is related with the passengers comfort.

Proceedings of the 2008 International Conference on Electrical Machines

Simulation and experimental results for a pseudo-levitating suspension control are presented in figures 6 and 7. Where is shown the advantage on the sprung mass acceleration, related with the passengers comfort. The tyre deformation have a dynamic response that is equivalent to the others suspension systems solutions.
200
20 10 0 -10 -20 27

Moreover, the electromagnetic linear actuator based suspension system has the advantage of being particularly flexible. It can be easily implemented a various kind of controllers, and switch between them, in order to achieve, in real time operation, the optimal conditions of safety and comfort the two main goals of an automobile suspension system. Also, with this kind of automobile suspension system, one electromagnetic linear actuator per wheel, there is also the possibility of building integrated controllers in order to compensate the vehicle roll balance.

2
0.1

100 (Zs-Zu) [mm] 0 Zs [gs]

1 0 1 0 10 20 Time [s] 30 20 4 10 20 Time [s] 30

0.05 0 -0.05 -0.1 27

-100
28 29 30

-200

28

29

30

1500
30 0 20 0 10 0 0 -100 -200 -300 27 28 29 30

0.5

1000 Fa [N] Zu [gs] 500 0 -500 -1000 0 10 20 Time [s] 30

2 0 2 40 10 20 Time [s] 30
0

REFERENCES
28 29 30

-0.5 27

Fig. 6. Simulation result for pseudo-levitating control, for t>10s and an oscillation frequency of 1,4Hz.
10 0.1 5 0 0 -5 -10 -0.1 -0.2 0 1 Time [s] 100 50 0 -0.1 -50 -100 0 1 Time [s] 2 3 0 1 Time [s] 2 3 -0.2 0.2 0.1 0 2 3 0 1 Time [s] 2 3

[1] Martins, I. Esteves, J. Pina da Silva, F. Tom, A. (2000). Automobile Suspensions Using Electromagnetic Linear Actuators. Proceedings of the International Conference IFACMechatronics 2000, Darmstadt. [2] Martins, I. Esteves, J. Pina da Silva, F. Verdelho, P. (1999). Electromagnetic Hybrid Active-Passive Vehicle Suspension System. Proceedings of the International Conference IEEE VTC99, Houston. [3] Bastow, D. (1987). Car Suspension and Handling, Pentech Press Limited London. [4] Kassakian, J.G. Wolf, H.-C. Miller, J.M. Hurton, C.J. (1996). Automotive Electrical Systems Circa 2005, IEEE Spectrum, August. [5] Ismenio Martins, Jorge Esteves, Gil D. Marques, Fernando Pina da Silva, Permanent-Magnets Linear Actuators Applicability in Automotive Active Suspensions IEEE Transactions on Vehicular Technology, vol. 55, no. 1, January 2006.

Fig. 7. Experimental result for pseudo-levitating control with an oscillation frequency of 1,4Hz. IV. CONCLUSIONS In this paper was analysed a pseudo-levitating suspension system, which was implemented in a workbench. This suspension system uses an electromagnetic actuator and a spring between the sprung and the unsprung masses. The spring supports the body weight and the electromagnetic actuator is controlled in order to compensate the spring deformation force. In the system there are two critical frequencies, the first one is associated with sprung mass and the other with the unsprung mass. The simulations and the frequency response diagram shows that for low frequencies the pseudo-levitating control system can bring more benefits in the passengers comfort when comparing with other suspension systems solutions.