Jtam.dvi


JOURNAL OF THEORETICAL

AND APPLIED MECHANICS

45, 1, pp. 133-145, Warsaw 2007

MR DAMPER PERFORMANCE FOR SHOCK ISOLATION

Bogdan Sapiński

Maciej Rosół

Department of Process Control, AGH – University of Science and Technology

e-mail: deep@uci.agh.edu.pl; mr@ia.agh.edu.pl

The paper is focused on the shock isolation performance of a drivers
seat whose suspension is completed with a linear magnetorheological
fluid damper (MRdamper). The aim of experimental investigations was
to recognize the MR damper performance against shock effects. The
experiments were performed on a linear damper of RD-1005-3 series
manufactured by Lord Corporation operating in open loop and closed-
loop system configurations under shock displacement-inputs (rounded
pulses and squarewaves). In thefirst case, theMRdamperwasoperating
as a passive damper and in the second case, as a controllable damper for
which real-time controllers were developed in the MATLAB/Simulink
environment. The system performance for shock isolationwas evaluated
basing onmeasured system responses.

Key words: magnetorheological fluid damper, drivers seat, shock iso-
lation, controller

1. Introduction

In recent years,MRfluid technology has spread rapidly andmanyMRdevices
and systems have been commercialized. In the automotive industry such pro-

ducts include: Delphi Automotives MagneRideTM shock absorbers, Carreras

MagneShockTM automotive racing shocks and Motion Master Ride Mana-
gement System (Carlson, 2003). These systems offer good performance for
suppressing of unwanted vibrations usingMRdampers. Experimental investi-
gations reported in Choi et al. (2000), Sapiński ( 2003), Sapiński and Rosół
(2003) confirmed the effectiveness ofMRdamper operation for vibro-isolation
in a suspension system of a driver’s seat support. AnMR seat damper acts as
an interface between the electronic control unit and the mechanical structure
of the suspension.Thedamping characteristics of anMRdamper canbevaried
continuouslywhich allows full control of drivers seat vibrationswith the use of



134 B. Sapiński, M. Rosół

magnetic fields. Such a semi-active system uses external power only to adjust
the damping and operates as a controller and set of sensors attached to the
seat (Ahmadian, 1999). The controller determines the required damping for-
ce and automatically commands the damper to generate an appropriate force
to reduce the amount of energy transmitted from the source of vibrations or
shocks to the suspended equipment.

The aim of this paper is to evaluate the system performance for shock
isolation through experimental investigation of shock isolation behaviour of a
driver seat supported by the RD-1005-3 damper. For this purpose, the seat
was tested under shock excitations (rounded pulses and square waves) in an
experimental setup. It appears, however, that the seat vibrations are difficult
to measure hence only their temporal effect could be quantified. Tests were
run in open loop and closed-loop system configurations.

2. Operating principle of MR damper

TheMR seat damper employed in the tested seat suspension is a compact de-
vice fabricated by Lord Corporation. It was developed for drivers seats in ve-
hicles (trucks, buses, agriculture tractors).Themainbenefits of theRD-1005-3
damper are: low voltage and current demands, precise and instantaneous con-
trol, continuously variable damping, simple electronics, real time control and
long service life. A schematic diagram for theMR damper is shown in Fig.1.

Fig. 1. Schematic diagram forMR damper

The inherent feature for the MR damper is high non-linear dynamic be-
haviour including hysteresis and step-like effects (Sapiński, 2002). This is due
to special properties of the magnetorheological fluid (MRF) (Sapiński, 2004).
The performance curves of velocity and control for theRD-1005-3 damper are
provided in Fig.2. It illustrates 3D-force-velocity-control curves captured in
the range of applied current 0 to 0.2A and piston velocity −0.4 to +0.4m/s.

The MR damper is controlled using the pulse width modulation (PWM)
method (Sapiński and Rosół, 2003). It is known that the RD-1005-3 response
time is less than25ms(time to reach90%of themaximal level duringa0 to1A



MR damper performance for shock isolation 135

Fig. 2. Performance curves for RD-1005-3 damper

step input at velocity 51 ·10−3m/s, RD-1005-3 Product Bulletin, 2003). This
time is dependent on an amplifier and power supply.

3. Shaping of excitation signals

In the experiments, displacement-input excitations were applied as rounded
pulse shocks and squarewaves. The roundedpulse shock is analytically expres-
sed by the formula

x0(t)=X0
e2

4
(γωnt)e

−γωnt (3.1)

where γ is the parameter expressing the time of pulse duration in relation to
the half-period of natural system vibrations. The parameter γ is given by the
formula

γ=
T

2τ
=
π

ωnτ
(3.2)

where
τ – duration of the square impulse with the area equal to that of

the rounded pulse,
ωn – pulsation of natural vibrations of the system,
X0 – rounded pulse amplitude.

The chief advantage of the rounded pulse excitation is that its first and
second derivatives assume limited values for all time instants t.



136 B. Sapiński, M. Rosół

The frequency of natural vibrations of the seat obtained experimentally
is 5Hz (Sapiński, 2003). The desired amplitude of the rounded pulse was set
to be X0 = 2.57 ·10

−3m. After re-scaling associated with the signal passing
through aC/A converter of the RT-DAC4 board and amplifiers of the control
cubicle shaker, themaximum value of the rounded pulse equals 1.75 ·10−3m.

Fig. 3. Rounded pulse displacement for various values of γ

Time patterns of simulated roundedpulse displacements for various values
of γ are provided in Fig.3. Time patterns of simulated andmeasured rounded
pulse displacements for γ =1 and γ =3 are compared in Fig.4. It is readily
seen that the roundedpulse signals are not perfectly transmittedby the shaker
system.Thedistortions are apparent in thepulsepatterns, particularly at high
values of γ.

Fig. 4. Rounded pulse displacement: (a) γ=1, (b) γ=3

The square wave excitations were applied to find the properties of the
system when shaker displacements underwent step changes. The square wa-
ve frequency was assumed to be 0.5Hz which guarantees steady-state seat
vibrations in between two subsequent wave edges.



MR damper performance for shock isolation 137

4. Experimental setup

A schematic diagram of the experimental setup prepared for testing of the
driver’s seat against shock excitations is shown in Fig.5.

Fig. 5. Experimental setup – schematic diagram

The power supply circuit consists of an electro-hydraulic shaker (EHS)
with a hydraulic pump (P) and a control cubicle (CB). Input-output data
were acquired using a data acquisition and control system based on a PC
(Pentium III/1GHz) with a multi I/O board (RT-DAC4), operating in the
Windows 2000,MATLAB/Simulink andRealTimeWindowsTarget (RTWT)
environment.The shaker-basedisplacement x0 and the framedisplacement x1
were measured with linear displacement transducers (LVDTs). The output
signal from the controllers developed for theRD-1005-3 damper is the voltage
in the range 0 to 5V which, after leaving the power controller engineered
by the authors is fed to the damper coil. Velocity and acceleration signals
were reproduced by using derivative blocks. In Fig.5, the RD-1005-3 damper
is represented by an electrically controlled damping element parallel to the
spring.
A general view of the drivers seat to be tested in the experimental setup

is shown in Fig.6.
A Simulink diagram developed for testing of the system is shown in

Fig.7. It enables us to: use the SA1 or SA2 controller, choose the type of
displacement-input excitation for the shaker base (roundedpulse, squarewave
and others), measure and process signals of the shaker-base displacement and
frame displacement.



138 B. Sapiński, M. Rosół

Fig. 6. Seat with RD-1005-3 ready for tests

Fig. 7. Simulink diagram for shock isolation testing of driver’s seat



MR damper performance for shock isolation 139

The block unit designated as the ”Rounded pulse generator” implements
rounded pulse displacement–input excitations, in accordance with formula
(3.1).

5. Experiments

Experiments were conducted in open loop and closed-loop systems. The aim
of open loop testswas to obtain the system response to various types of signals
exciting the shaker base at various levels of current applied to the RD-1005-3
damper coil.Theaimof closed-loop testswas to check thedampingperforman-
ce using the RD-1005-3 damper controlled by controllers developed specially
for that purpose.

5.1. Open loop system

Results obtained in the open loop system testing are shown as time pat-
terns in Figs 8-13. Figure 8a provides seat responses for the current I = 0A
and γ =1, γ=2, γ =3. It appears thatwhen thevalue of γ increases (i.e. the
time of impulse duration gets shorter), the maximum value of the seat frame
displacement will decrease. It follows from Fig. 9a that the higher the value
of γ, the greater themaximum value of vibration acceleration. Moreover, the
time required for the system to reach the steady state gets shorter.

Fig. 8. Frame displacement in response to rounded pulses for: (a) I =0A and
various values of γ, (b) γ=1 and various applied currents

The influence of the applied current level on the displacement and ac-
celeration of the seat frame is shown in Fig.8b and Fig.9b. As the current
level increases, themaximum values of the seat frame displacement and acce-



140 B. Sapiński, M. Rosół

leration increase too. Throughout the investigated range of the applied cur-
rent, there were no major changes in the time required to reach the steady
state.

Fig. 9. Frame acceleration in response to rounded pulses for: (a) I =0A and various
values of γ, (b) γ=1 and various applied currents

Figures 10a and 10b present the seat vibration acceleration obtained for
square wave displacement-input excitations applied to the shaker base for
I = 0A and I = 0.10A. The base excitation signal (doted line) is scaled
500:1 and expressed inmeters. Figs 10a and 10b provide sections of time pat-
terns covering the period between the occurrence of ascending or descending
wave edges and the instant the steady state is reached. As the current level
increases, the amplitude of seat vibration acceleration rises too, and the time
required to reach the steady state gets shorter. The variations of acceleration
amplitude for various current levels are nonlinear.

Fig. 10. Frame acceleration in response to square wave for: (a) I =0A,
(b) I =0.10A



MR damper performance for shock isolation 141

5.2. Closed-loop system

It is assumed that the operation of developed controllers for the RD-1005
damper, i.e. on-off controller (SA1) and continuous controller (SA2), is gover-
ned by equations (5.1) (Liu et al., 2002)

SA1 : I =

{

c1 for ẋ1(ẋ1− ẋ0)­ 0

0 for ẋ1(ẋ1− ẋ0)< 0

(5.1)

SA2 : I =

{

c2|ẋ1− ẋ0| for ẋ1(ẋ1− ẋ0)­ 0

0 for ẋ1(ẋ1− ẋ0)< 0

where
ẋ0 – base velocity,
ẋ1 – frame seat velocity,
ẋ1− ẋ – relative velocity,
c1,c)2 – constants.

The values of c1 and c2 depend on themaximumvalue of the applied cur-
rent. The term ẋ1(ẋ1−ẋ0) is called the velocity product.Timepatterns of the
seat vibration acceleration obtained in the closed-loop system completed with
the controllers SA1 and SA2 and the time patterns in the open-loop system
are compared in Figs 11a and 11b. In both cases, the rounded-pulse displa-
cement input excitations were applied. It is readily seen that the controllers
SA1 and SA2 do not provide any reduction to the seat vibration acceleration
while compared to the open loop system.

Fig. 11. Frame acceleration in response to rounded pulse for: (a) γ=1, (b) γ=3

The experiments when square wave excitations were applied to the shaker
base allow similar conclusions to be drawn as in the case of rounded pulse



142 B. Sapiński, M. Rosół

excitations. That was confirmed by the time patterns of seat vibration accele-
ration obtained for the closed–loop system with the controllers SA1 and SA2
(Fig.12).

Fig. 12. Frame acceleration in response to square wave for closed-loop systemwith:
(a) SA1, (b) SA2

A comparison between the performance of the open loop and closed-loop
systems (with the controllers SA1 and SA2) is shown in Fig.13.

Fig. 13. Closed-loop systemwith: (a) SA1, (b) SA2

It is seen that the control signals for the controllers SA1 andSA2 are gene-
rated in accordance with formulas (5.1). Nonzero values of the control signal
(i.e. current in the MR damper coil) occur only when the term of velocity
product is positive. It is reasonable to expect, therefore, that at the zero cur-
rent, the time patterns of seat acceleration obtained for the open loop and
closed-loop system configurations and for the same initial conditions ought to



MR damper performance for shock isolation 143

be similar. Actually, the acceleration of seat vibrations is much greater in the
closed-loop system, which is best seen in Fig.13a (time from 0.026 to 0.038s)
and in Fig.13b (time from 0.026 to 0.037s).

6. Conclusion

The paper summarises results of an experimental programmewhere a drivers
seat equippedwith anMRdamper was subjected to shock displacement exci-
tations. Responses of open-loop and closed-loop systems (with the controllers
SA1 and SA2) are compared, leading us to the conclusion that the action of
these two controllers fails to reduce the vibration acceleration in the system.
The reasons for this state of affairs are attributable to the properties and ope-
rating principles of the electromagnetic circuit of theRD-1005-3 damper.This
problem is illustrated graphically in Figs 14a and 14b, showing the current
levels and damping force under triangle displacement-input excitations of the
base and PWMvoltage signal across coil clamps.

Figure 14a shows base excitations (frequency 1Hz, amplitude 10 ·10−3m)
andPMWexcitations (frequency 4Hz, amplitude 0.8V,width factor 0.2/0.8).
It is apparent (Fig.14b) that the time required for the current and damping
force to stabilise is still considerable. The times tI and tF are understood as
times required to reach 95%of the steady value in conditions of step variations
of voltage across the coil clamps. The measurement results reveal that the
times required for stabilisation of the current in the coil and the damping
force are:

• for the ascending edge of the voltage signal:

tI =65 ·10
−3 s, tF =150 ·10

−3 s,

• for the descending edge of the voltage signal:

tI =101 ·10
−3 s, tF =78 ·10

−3 s.

Figure 14b shows base excitations (frequency 4Hz, amplitude 1.5·10−3m)
andPWMexcitations (frequency20Hz, amplitude0.8V,width factor 0.2/0.8).
It appears that as the frequency of the PWMcontrol signal increases, neither
the current in the coil nor the damping force will reach the steady value.

In the case considered in this study, the current level would vary from 0 to
0.2A. One has to bear inmind that the times required to stabilise the current
and thedamping force chieflydependon theparameters of thepower controller
and on the current level in the coil. Themeasurement results obtained for the
current varied from 0 to 1.0A reveal that the times required for stabilisation
of the relevant parameters changed significantly, and now are equal to:



144 B. Sapiński, M. Rosół

Fig. 14. Force and current responses in RD-1005-3 damper: (a) PWM frequency
4Hz, triangle displacement-input 1Hz, (b) PWM frequency 20Hz, triangle

displacement-input 4Hz

• for the ascending edge of the voltage signal:

tI =48 ·10
−3 s, tF =56 ·10

−3 s,

• for the descending edge of the voltage signal:

tI =148 ·10
−3 s, tF =173 ·10

−3 s.

This is a consequence of themagnetic circuit reaching the state of saturation.

Concluding remarks why the system time of response to changes of the
control signal is prolongedmightbeas follows: nonzero timeof theMRdamper
force stabilisation, magnetic residues in the MR damper components, delays
due to current signal stabilisation at the output of the power controller.

Acknowledgement

The research work has been supported by the State Committee for Scientific

Research as a part of the scientific research project no. 5T07B02422.



MR damper performance for shock isolation 145

References

1. Ahmadian M., 1999, On the isolation properties of semi-active dampers, Jo-
urnal of Vibration and Control, 217-232

2. Carlson J.D., 2003, Engineering with magnetorheological fluids, Proc. of
Smart03 Workshop, Warsaw, to be published

3. Choi S.B., Lee B.K., Nam M.H., Cheong C.C., 2000, Vibration control
of a MR seat damper for commercial vehicle, Smart Structures and Integrated
Systems, N.Werely (Edit.),Proc. of SPIE, 3985

4. Liu Y., Mace B., Waters T., 2002, Semi-active dampers for shock and
vibration isolation: algorithms and performance, Proceedings of International
Symposium on Active Control of Sound and Vibration, UK, 1121-1132

5. Lord Corporation: RD-1005-3 Product Bulletin, 2003

6. SapińskiB., 2002,Parametric identificationofMRlinearautomotive sizedam-
per, Journal of Theoretical and Applied Mechanics, 40, 703-722

7. Sapiński B., 2003,Autonomous control systemwith fuzzy capabilities forMR
seat damper,Archives of Control Sciences, 13 (XLIX), 2, 115-136

8. SapińskiB., 2004,LinearMagnetorheological FluidDampers forVibrationMi-
tigation: Modelling Control and Experimental Testing, Monography 128, Cra-
cow (to be published)

9. Sapiński B., Rosół M., 2003, Real-time controllers for MR seat damper,
Proc. of Smart03 Workshop, Warsaw (to be published)

Efektywność zastosowania tłumika magnetoreologicznego w izolowaniu

obciążeń udarowych

Streszczenie

W artykule rozważono problem wibroizolacji zawieszenia fotela kierowcy z linio-
wym tłumikiem magnetoreologicznym (MR). Celem badań była ocena efektywności
działania tłumika MR w układzie zawieszenia fotela poddanego działaniu wymuszeń
impulsowych. Eksperymenty przeprowadzono dla układu z otwartą i zamkniętą pę-
tlą sprzężenia zwrotnego.Wbadaniachwykorzystano tłumikMR (model RD-1005-3)
firmy Lord Corporation. Fotel poddawano wymuszeniom kinematycznym typu pro-
stokątnego i symulującego wstrząsy. Zaprojektowane regulatory zaimplementowano
w programieMATLAB/Simulink, a następnie uruchomiono w środowisku czasu rze-
czywistego oferowanego dla systemu operacyjnegoWindows 2000/XPprzez przybor-
niki RTW/RTWT. Oceny efektywności układu wibroizolacji fotela z tłumikiem MR
dokonano na podstawie pomiarów odpowiedzi dynamicznych zawieszenia.

Manuscript received August 3, 2006; accepted for print August 18, 2006