You are on page 1of 9

Study of the response time of MR dampers

Xinchun Guan*aPengfei GuoaJinping Oua,b a School of Civil Engineering, Harbin Institute of Technology, Harbin, China, 150090; b School of Civil and Hydraulic Engineering, Dalian University of Technology, Dalian, 116024, China
ABSTRACT
Response time is an important parameter which determines the applied fields and practical vibration reduction effects of magnetorheological (MR) dampers. However, up to now, only a few papers discuss the test and analysis of response times. In this paper, the response time of a large-scale MR damper at different velocities and currents was firstly tested. Then, the transient magnetic field excited by the time-variant excitation current was simulated by finite element method (FEM). Based on the variation of the shear yield stress of magnetorheological fluids in the gap between the cylinder and the piston, the response time of the MR damper was investigated. Influences of eddy current and excitation current response time on the dampers response were also explored. Results show that by utilizing finite elements method, the calculated average effective shear yield strength can be used to predict the response time of a MR damper. Electromagnetic response is the predominant factor influencing the response time of a MR damper, and reducing eddy currents is the key to accelerate the response of a MR damper. Moreover, influence of eddy currents is much larger under stepping down excitation currents than stepping up currents, and with a same magnitude of step, no matter when the current increases or decreases, the smaller the initial current, the greater the eddy current affects a dampers response and the longer the response time of damping force is. A fast response excitation current may induce large eddy currents which reduce the response of the damper instead. Keywords: magnetorheological damper, eddy current, response time, finite element analysis, magnetic field

1. INTRODUCTION
With the ability that quickly responds an external magnetic field with a dramatic change in rheological behavior, magnetorheological fluids (MRF) have received a lot attention over the past several decades. A variety of MRF based devices are developed, among which MR dampers have been widely used in vibration controls of civil engineering structures, automobile suspension systems, etc [1,2]. Unfortunately, response of force output of a MR damper is far more slowly than MRF itself and response time of a damper is a key parameter that determines its applied fields and practical damping effects. However, presently most studies mainly focus on analysis of static magnetic field, modeling of damping force in control strategy, and investigation of controlling effects, etc. Few papers discuss the response time of MR dampers and published results are also different because of different testing methods and structural parameters of MR dampers. For example, the response time of a MR damper made by Lord Company is between 7.5 ms and 85ms tested by Koo [3], and that of a valve MR damper tested by Milecki is between 30 and 180ms [4], and that of a MR damper fabricated and tested by Soda is about 300 to 400ms [5]. Besides, variation of magnetic field in the damper which directly determines the changing of the damping force is not discussed in these studies. In this paper, the response time of a large scale MR damper whose maximum damping force is about 270kN will be tested. By utilizing the finite element software ANSYS, magnetic field in the gap between cylinder and piston will be calculated. Then based on the calculated shear stress of MRF in the gap, response time of the MR damper will be analyzed theoretically. Finally, possible influencing factors of the response time of MR dampers such as eddy currents and response of excitation current will be explored.

2. EXPERIMENTAL STUDY
2.1 Experimental instruments
guanxch@hit.edu.cn; phone 86 451 8628-2367; fax 86 451 8628-2367; Harbin Institute of Technology

Second International Conference on Smart Materials and Nanotechnology in Engineering, edited by Jinsong Leng, Anand K. Asundi, Wolfgang Ecke, Proc. of SPIE Vol. 7493, 74930U 2009 SPIE CCC code: 0277-786X/09/$18 doi: 10.1117/12.840217 Proc. of SPIE Vol. 7493 74930U-1

The damper was tested with a servo-hydraulic test facility (MTS) whose maximum driving force is about 2500kN. Main parameters of the MR damper are listed in Table 1. Experimental results show that its maximum damping force at velocity of 40 mm/s is about 270kN.
Table 1. Main parameters of the MR damper

Maximum Stroke (mm) 50 2.2 Testing method

Number of coils 4

Turns of each coil 900

Gap (mm) 1.5

Diameter of cylinder (mm) 200

Effective length of piston (mm) 120

The testing method and experimental set up are shown in Fig.1. Driven by MTS, the piston of the MR damper is pulled out and back at a constant velocity. Two voltage signals are sent out at the same time by an industry computer. One was sent to the current driver which then supplies direct currents to coils of the damper. Meanwhile the other voltage was sent to the MTS system to record times of current changes. Consequently, four signals operation time of MTS, time of current driver, displacement and damping force of the MR damper were stored in MTS.

Fig.1 Experiment set up for testing of response time of the MR damper

As shown in Fig.2, when damping force gradually changes from one stable state Fi to another, Ff, response time is defined as the time span from the end time of Fi to the time instant that damping force reaches Fi+0.632 (Ff-Fi). A high frequency sampling was adopted to ensure that the whole changing process of the damping force accompanied with varying currents could be captured.

(a) Damping force under stepping up current Fig.2 Definition of the response time

(b) Damping force under stepping down current

Proc. of SPIE Vol. 7493 74930U-2

2.3 Experimental results The response time of the damper is tested with piston at velocities of 1cm/s, 2cm/s and 4cm/s and step currents 0.0A0.8A-0.0A, 0.4A-1.2A-0.4A, 0.0A-0.4A-0.0A, 0.4-0.8A-0.4A are applied for each velocity case. Time history cure of damping force under 4 cm/s piston velocity and 0.4A-1.2A-0.4A step current is shown in Fig.3 and results under all other conditions have similar characters.

(a) Stepping up current 0.4A-1.2A Fig.3 Typical experimental curves

(b) Stepping down current 1.2A-0.4A

(a) Stepping up current Fig. 4 Statistical results of the tested response time of the MR damper

(b) Stepping down current

In each operation condition, the response time is tested 3 times and average value is obtained, as plotted in Fig.4. It can be observed that: (1) Although with a same stepping magnitude, 0.4A, response time under the step current 0A-0.8A is much longer than that of 0.4A-1.2A, no matter whether the current is stepping up or down. This indicates that with a same stepping magnitude, a larger initial current leads to a shorter response time. (2) With a same stepping magnitude, response time under the stepping down current is longer than under the stepping up current, especially for the current decreasing to zero. (3) It still takes as long as 160ms-240ms for the MR damper to response under stepping up currents. (4) A fast piston velocity results in a fast response.

Proc. of SPIE Vol. 7493 74930U-3

3. THEORETICAL INVESTIGATION BASED ON MAGNETIC FINITE ELEMENT ANALYSIS


The adjustable damping force induced by modifying excitation current in a MR damper mainly comes from the changing of shear yield strength of MRF. In addition, given that the damping force is mainly predicted based on quasi-static model [6, 7] at present, in this paper, the response time of the MR damper was investigated by observing the variation of the shear yield strength of MRF, instead of damping force directly. 3.1 Magnetic finite element model of the MR damper A 2D axial symmetric finite element model of the MR damper is constructed in ANSYS, as shown in Fig.5. Elements types used for the piston, rod, cylinder, MRF and air are PLANE53, and the whole model is enclosed by a layer of INFIN110 elements. CIRCU124 elements are used for the independent current source (ICS) and four parallel connected coils. A small element size applied for the gap, piston, and cylinder which together constitute the magnetic circuit of the MR damper. It should be noted that in ANSYS an arbitrary transient excitation can be implemented by modifying REAL CONSTANT within load steps.

Fig.5 Finite element model of the MR damper

3.2 Average effective shear strength of MRF Since the adjustable damping force of a MR damper results from the changing of shear yield strength of parts of MRF in the gap which is named effective MRF (shown in Fig.6) hereafter for convenience, transient damping force should be predictable by calculating time-variant shear yield strength of the effective MRF.

Fig. 6 Effective parts of magnetorheological fluid

However, magnetic fields in elements of effective MRF differ from each other, the concept of the average effective shear strength *(t) is put forward and given by

Proc. of SPIE Vol. 7493 74930U-4

* (t ) = 3632 B + 206593.4 B 2
178444.2 B 3 + 5023.3 B 4 + 23023.4 B 5

(1)

Here, relationship between magnetic flux density and shear yield strength of MRF is fitted from the experimental data (see Fig.7); B,the average effective magnetic flux density, is time dependent and given by.

B (t ) =

B (t ) V
i =1 i

V
i =1

(2)

Here, Vi and Bi(t) are volume and radial magnetic flux density of effective MRF element i at time instant t respectively.

Fig.7 Magnetic flux density vs. shear yield strength of MRF

Fig.8 Testing method of excitation currents

The calculating procedure for* at time t, *(t), is as follows. Firstly, implement time history curve of the excitation current in the finite element model in section 3.1. When the solution is done, Bi (t) and Vi (t) are retrieved by using APDL language of ANSYS. Then they are substituted into equation (2) to get the average effective magnetic flux density B(t). At last, substituting B(t) into equation (1) will obtain *(t). In the above procedure, the time history curve of the electric current is experimental data and measured by placing the damper on the ground. As shown in Fig.8, the entire electric circuit in this test is composed of four parallel-connected coils of the damper, one standard resistance and one current power supply. The voltage signal sent out by the industrial computer is proportionally changed by current driver into a current signal which is then used as the dampers excitation current. Time-variant current in the circuit is obtained by dividing the voltage drop on the stand resister by its resistance and this current can be used to check that whether the current is loaded as expected. 3.3 Results and discussion Tested damping force (4cm/s) and calculated shear yield strength are shown in Fig.9. Same time scale and range are used for these two time-history curve to make them comparable. Due to the influences of system stiff of the whole experiment set up, tested response time of a MR damper is greatly depend on the dampers velocity. The larger the velocity, the smaller the influence is [1]. It can be observed from statistical results in Fig.4 that 4cm/s is large enough to ignore such influence, thus the corresponding test data are used in Fig.9. Time history curve of damping force is in well agreement with that of average effective shear yield strength, as shown in Fig.9. This indicates that the calculating method for response time of MR dampers in this paper is correct, i.e. transient damping fore can be predicted by average effective shear yield strength.

Proc. of SPIE Vol. 7493 74930U-5

In fact, response of excitation current and magnetic field are both taken into account in the above calculation in which the former is considered by using test data and the latter by transient magnetic finite element method. Agreement in Fig.9 indicates that the response time of a MR damper is mainly composed of the response of excitation current and that of magnetic field.

4. INFLUENCES OF EDDY CURRENT AND EXCITING CURRENTS


4.1 Influence of eddy current By comparing tested history-curves of currents and damping forces as shown in Fig10, it can be found that response of magnetic field lags far behind its exciting current. Given the fact that MRF response to an applied magnetic field within several milliseconds, the lag of damping force can be considered as the lag of magnetic field which is caused by eddy currents according to the Faraday electromagnetic induction law. In condition that electric currents step up or down from 0A to 0.8A, the response time of a magnetic field for the former is about 300ms and 520ms for the latter. The response time of the magnetic field for the current stepping from 0.4A to 1.2A is about 20ms, 150ms conversely. This means that, with a same stepping magnitude, eddy currents have larger effects on the magnetic field for stepping up excitation currents than stepping down. Furthermore, the lower the initial excitation current, the larger the effects are.

(a) Current steps up from 0.0A to 0.8A

(b) Current steps down from 0.8A to 0.0A

(c) Current steps up from 0.4A to 1.2A Fig. 9 Tested damping force and calculated shear yield strength

(d) Current steps down from 1.2A to 0.4A

Proc. of SPIE Vol. 7493 74930U-6

(a) Current steps up from 0.0A to 0.8A

(b) Current steps down from 0.8A to 0.0A

(c) Current steps up from 0.4A to 1.2A Fig. 10 Experimental time-history curves of current and damping force

(d) Current steps down from 1.2A to 0.4A

4.2 Influence of exciting currents Results in section 3.3 have shown that the response time of damping force mainly depends on the electromagnetic response of the damper and the lag of magnetic field is also studied in section 4.1. This section will continue to explore the influence of response time excitation currents on that of damping force, i.e. whether a fast changing current will lead to a fast response damper. The time-history curve of the average effective shear stress of MRF are computed as the time for excitation current stepping from one level to another is shorten to 50ms., i.e. in 50ms, electric current complete the process of jumping from 0.0A (0.4A) to 0.8A (1.2A). The calculated average effective shear stress of MRF and the tested damping force in section 2 are plotted together in Fig.11. Fig.11 shows that after excitation current shortened to 50ms, the response time of the damping force is shortened only a little, except for the current stepping from 0A to 1.2A. The reason for this perhaps is larger eddy currents that caused by faster variations of excitation currents. Excitation currents response time has bilateral effects on the damping force. A fast electric current change, on one hand, will directly shorten the response time of the damping force; on the other hand, it will induce a larger eddy current which slows the response of the damping force. Consequently, a fast excitation current response doesnt necessarily means a fast response of the damping force and the fastest response should be reached at the balance point of bilateral effects. As discussed in sections 4.1 and 4.2, reducing the eddy current is the key to accelerate the response of a MR damper.

Proc. of SPIE Vol. 7493 74930U-7

(a) Current steps up from 0.0A to 0.8A

(b) Current steps down from 0.8A to 0.0A

(c) Current steps up from 0.4A to 1.2A

(d) Current steps down from 1.2A to 0.4A

Fig. 11 Time-history curves of calculated shear strength and experimental damping force

5. CONCLUSIONS
In this paper, experimental study and finite element method based theoretical analysis of response time of a MR damper are conducted and the following conclusions are achieved. (1) By utilizing finite elements method, the calculated average effective shear yield strength can be used to predict response time of a MR damper. (2) The response time of a MR damper mainly depends on its electromagnetic response, and reducing the eddy current is the key to develop a fast response MR damper. (3) With a same stepping magnitude, eddy currents have larger effects on the magnetic field for stepping up excitation currents than stepping down currents and response of damping force is longer in the former condition. Furthermore, the lower the initial excitation current, the larger the effects are. (4) Due to the bilateral influences of excitation current, a fast current change may not result in a fast response of the damping force.

Proc. of SPIE Vol. 7493 74930U-8

6. ACKNOWLEDGEMENTS
This research is financially supported by the National Natural Science Foundation of China under grant number 90815027High-Tech Research and Development of China under grant number 2006AA03Z103 and National Basic Research Program of China under grant number 2007CB714204, Commonweal Vocation Scientific Research Program of China under the grant number 2008419073, National Science and Technology Support Plan under number 2006BAJ03B06.

REFERENCES
1. J. D. Carlson, D. M. Catanizarite, K. A. Clair, "Commercial magneto-rheological fluid device," Proc. 5th Int. Conf. ER Fluids, MR Suspensions and Associated Technology, Singapore, 2857-2865(1996). 2. Wojciech Szelag, "Finite element analysis of the magnetorheological fluid brake transients," The International Journal for Computation and Mathematics in Electrical and Electronic Engineering, 23(3), 758-766(2004). 3. J. H. Koo, F. D. Goncalves, Mehdi Ahmadian. "Investigation of the response time of magnetorheological fluid dampers," Proc. SPIE 5386, 63-71(2004). 4. Andrzej Milecki. "Investigation of Dynamic Properties and Control Method Influences on MR Fluid Dampers Performance," Journal of Intelligent Material Systems and Structures, 13, 453-458(2002). 5. Satsua Soda, Haruhide Kusumoto, et al., "Semi-active seismic response control of base-isolated building with MR damper," Proc. SPIE 5052, 460-467(2003). 6. Young Tai Choi, Norman M. Wereley. "Assessment of time response characteristics of electrorheological and magnetorheological dampers," Proc. SPIE 4331, 92-102(2001). 7. Yang G. "Large-scale magnetorheological fluid damper for vibration mitigation: modeling, testing and control," Notre Dame, Indiana. Ph.D. Dissertation University of Notre Dame, 2001.

Proc. of SPIE Vol. 7493 74930U-9

You might also like