Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3218-3222 3218 www.etasr.com Yerrawar & Arakerimath: Parametric Analysis of Magnetorheological Strut for Semiactive Suspension … Parametric Analysis of Magnetorheological Strut for Semiactive Suspension System Using Taguchi Method R. N. Yerrawar Department of Mechanical Engineering D Y Patil Institute of Engineering & Technology Pune, Maharashtra, India rahul.yerrawar@gmail.com R. R. Arakerimath Department of Mechanical Engineering G H Raisoni College of Engineering & Management Pune, Maharashtra, India rrarakerimath@gmail.com Abstract—Magnetorheological (MR) strut is among the leading advanced applications of semi-active suspension systems. The damping force of MR damper is controlled by varying the viscosity of MR fluid. In this work, the viscosity of MR damper varies by changing the current from 0.5A to 0.7A. The design of experiments is taken into account in concert with the product/process development as one completely advanced tool. The parameters used for ride comfort optimization are sprung mass, spring stiffness, tire pressure, current, and cylinder material with two levels of each. Taguchi orthogonal array method is used to select the best results by parameter optimization with a minimum number of test runs. In this paper, from Taguchi L16 array and S/N ratio analysis, it is observed that the cylinder material with Al and CS for damper cylinder is a key parameter for performance measure of semi-active suspension system. From regression analysis, a linear mathematical model is developed for Al and CS as cylinder materials. The interaction of cylinder materials with all four parameters is plotted. The methodology implemented for measurement of acceleration as a ride comfort is as per IS 2631- 1997. The more economical model of magnetorheological damper will motivate Indian auto industry to broader applications. Keywords-magnetorheological damper; parametric analysis; semiactive I. INTRODUCTION An effective automotive mechanical system has to give comfort and good road holding ability, once the vehicle is moving on any road disturbance. The vehicle body is supposed to have considerable damping or resistance to avoid massive oscillations. Comfort optimization is predicated on spring stiffness, damping and sprung mass. Normally, such system types are developed by experimental approach. The procedure for designing and modeling of a novel large-stroke MR damper is tested in [1] under impact load and it is observed that the inertia damping force could not be ignored like in the common MR damper due to large acceleration. 2D axisymmetric model was built up in [2] to investigate and analyze the MR damper qualities predicting the damping force characteristics. The Bingham model of MR damper was proposed and compared with experimental results for magnetic saturation. This work inferred that the obtained magnetic saturation model can show effect on current, amplitude and excitation frequency. The phenomenon of magnetic saturation is simulated in [3] and obtained results are compared with the experimental results. Optimization technique is used to maximize magnetic flux density which affects the force developed by damper [4]. The MR fluid structure and properties, damper description and parametric damper models which are connected in various ways with the actual behavior of the MR fluid were presented in [5]. Recommendations for the design of main-spring and shock absorber of motor vehicle suspensions were given in [6]. Indications acquired for the shock absorber allow us to obtain the damping coefficients in the compression and rebound strokes and to calculate the power dissipated during the vehicle oscillatory movement. An extensive work has been done on the transmissibility of vibration isolator in [7], where authors proposed the concept of output frequency response function in order to examine the force transmissibility of MDOF structures with a cubic non-linear viscous damping device. Control strategies can be divided into two main categories: active and semiactive. In the case of active systems the performance is superior and optimum transmissibility has no resonance. Semiactive systems provide reliability comparable to passive suspension systems [8]. In [9], it is observed that the vertical acceleration of body mass is substantially reduced by using a controlled MR damper. Sensitivity analysis for leaf springs with different stiffness was done in [10] on roll bar of front suspension system by using Adams simulation software. Static test results were compared with the dynamic test ones. From the simulation results, it is observed that the parametric effect on suspension system is not satisfactory by using static analysis. The experimental results carried out in [11] regarded optimization of suspension system used for vehicle occupant seat for a quarter car model. The Taguchi method was implemented in order to reduce vibrations. From the simulation results it is observed that stiffness and damping preliminary values are the control variables of the suspension system. Authors in [12] studied the dynamic loads enforced on suspension components through bushings and joints. In order to predict the vehicle com mi pro me exp ach tha opt pro sus stif dam Th for des ma and dam vis A. FƮ wh cyl Engineerin www.etasr mponent life e inimizing the oposed in [13 ethod and use periment was hieve suspensi at both suspe timization. T ocesses and a spension para ffness, sprung mper are cons his standard re r vehicle which TABLE I. Accelera Less th 0.315 0.5 0.8 t 1.25 Greate The MR dam signed accordi athematical m d magnetic ci mping force scous compone Fig. 1 Damping Fo The damping and viscous co FD=FƮ+Fƞ = (2.07 (1 here, Q= Ap × The calculati linder gap as ng, Technology r.com estimation, Ta stress value 3] robust opt es Adams sof s designed ba ion Pareto solu ension and ro The design of lot of time. In ameters such g mass and mo sidered to achi ecommended h is represente COMFORT REAC tion (m/s2) han 0.315 to 0.63 5 to 1 to 1.6 to 2.5 er than 2 II. DAMP mper damping ingly. For calc model suggeste ircuit shown i is divided in ents . . Magnetic ci orce Calculatio g force FD is d omponent Fƞ a ƞƞ . 2Ʈ 12ƞQLtAp wh3 v , Ap= (D 2 ions are obtain s 1.2mm and y & Applied Sci Yerrawar & aguchi method of the lower timization is ftware to mod ased on the T ution set and t obustness imp f car robustn n the current w as tire pressu st important, c ieve good ride the accelerati ed in Table I. CTIONS TO VIBRAT L Not Unc A little Un Fairly Un Uncom Very Unc Extremely U PER FABRICATI force is calcu culation of the ed by the Bing in Figure 1 is nto an induced ircuit design of M ons divided into a y as in (1) and (2 Ʈ p 2 - do 2) , w = π ned by conside velocity as ience Research Arakerimath: P d is implement r control arm based on Ta del suspension Taguchi meth the result illus proved a lot ness needs v work, the semi ure, current, s cylinder mater e comfort (RC) ion as RC me TION ENVIRONMEN Level comfortable ncomfortable ncomfortable mfortable comfortable Uncomfortable ION ulated for 250k e damping forc gham plastic m s used in whic d yield stress, F MR damper yield stress in 2). (1) (2) π ( ) ering the pisto 10 kmph. Ta h V Parametric An ted by m. The aguchi n. An hod to strated after arious active spring rial for ) [14]. easure NT kg and ce, the model ch the FƮ and nduced on and able II sho can roa use test mas min (co test setu acc mo acc pc acc acc Fig. cycl Vol. 8, No. 4, 20 alysis of Magn ows the dimen n see the fabric TABLE II. Param Piston rod and c dista Cylinder t Length o Piston and cylin Piston r Distance amo Radius of p III. EXP In order to st ad holding, th ed. Test set up t the suspens ss is connecte nimize frictio mpatible for ting. Figure 3 up is provide cording to con tor. It contain celerometer, a and connectin celerometer. T celeration of w 3. Experimen linder 018, 3218-3222 etorheological nsions of the M cated damper. Fig. 2. Fabr . DIMENSION meter oil width radial ance thickness of Pole nder clearance radius ong the poles piston rod PERIMENTAL P tudy the suspe he Quarter Ca p has arrangem ion on differ ed to the verti on. Double McPherson ty 3 demonstrate ed with a ca nstant wave f ns instruments loaded pc, M ng cables for The model i wishbone and o ntal setup of quar 2 Strut for Semia MR damper a icated MR dampe NS OF FABRICATED Notations H t L h R l r PROGRAM FOR ension effects ar Model exp ments for varyi rent loading c ical column b wishbone su ype of suspen es the experim am exciter w form and pow s like data acq MR damper, po r the DC curr is tested for on sprung mass rter car model wi 3219 active Suspensi and in Figure er D MR DAMPER Dimensions(m 5 3 9 1.5 18.03 42 9 R MR DAMPER and study RC perimental setu ing sprung ma conditions. Sp y linear bearin uspension sy nsion) is use mental setup. which is desi wered by a PM quisition hardw ower supply fo rent controller r measuremen s. ith MR damper w ion … 2 we mm) C and up is ass to prung ng to ystem d for The igned MDC ware, or the r and nt of with CS Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3218-3222 3220 www.etasr.com Yerrawar & Arakerimath: Parametric Analysis of Magnetorheological Strut for Semiactive Suspension … IV. EXPERIMENT DESIGN The experiment design is taken into account in concert with the product/process development as one completely advanced tool. The experiment design takes into consideration wide- ranging applications in the development of product/process. There are the two approaches of DOE. A. Full Factorial Design In full factorial experimental approach, two (at least) or more factors are used in a feasible level and the required units take all probable combinations of all levels of these factors. For 9 factors with 2 levels for each, 512 runs are required for a full factorial design. In this method, each parametric effect on response variables and the interaction between factors is studied. B. Taguchi Method The major shortcoming of full factorial design is that the number of experiments to be carried out is large which means more laborious work on a large number of parameters. To avoid this troublesome work, Taguchi recommended a specially intended method called orthogonal array to learn the whole parameter effect with lesser quantity of experiments. C. Selection of Control Factors It is observed that spring stiffness and electric current need to be appropriately considered under positive range of sprung mass and road conditions. For exceedingly rigid suspension the vehicle system will be vastly stable but acceleration of occupant body will be high and comfort will be low. For a smaller amount of hard suspension, passenger comfort will increase but vehicle stability decreases. The selected parameters for suspension are described in Table III. TABLE III. TAGUCHI PARAMETER TABLE Parameter Unit Levels Level 1 Level 2 Tire Pressure Kgf/cm2 (PSI) 2 34 40 Spring Stiffness N/mm 2 14000 8000 Damper Cyclider Material --- 2 Al (+1) Structural Carbon Steel (-1) Sprung Mass Kg 2 72 97 Current (A) 2 0.5 0.7 D. Orthogonal Arrays Taguchi developed the orthogonal array method to study systems whose performance is affected by different factors, in a more convenient and rapid way. This method can be used to select the best results by optimization of parameters with a minimum number of test runs. Application of Taguchi method can achieve the best results out of various tests by selecting the optimum combination of different factors. Final product quality can be improved in terms of process optimization, product design and system analysis. To study a sufficient number of experiments with a large number of variables, Taguchi recommended the orthogonal array method for experimental design. The conclusions made after a sufficient number of experiments are usable over the whole experimental area covered by control factors and their ranges. Orthogonal array method was discovered before Taguchi implemented it for DOE. In this method, the columns of tabulated sets are orthogonal and consequently linear graphs are plotted to fit the specific project applications. In the current study, five parameters, tire pressure, spring stiffness, damper cylinder material, sprung mass and electric current, each one with two levels are used. Table IV defines the L16 array. The number following the “L” indicates the number of runs in the design, and 16 means the array requires 16 runs (25), which indicates that the design estimates up to 5 main parameters at 2 levels each. TABLE IV. TAGUCHI L16 ORTHOGONAL ARRAY Run Air Pressure Spring Stiffness Cylinder Material Sprung Mass Current 1 34 14000 1 72 0.5 2 34 14000 1 97 0.7 3 34 14000 -1 72 0.7 4 34 14000 -1 97 0.5 5 34 8000 1 72 0.7 6 34 8000 1 97 0.5 7 34 8000 -1 72 0.5 8 34 8000 -1 97 0.7 9 40 14000 1 72 0.7 10 40 14000 1 97 0.5 11 40 14000 -1 72 0.5 12 40 14000 -1 97 0.7 13 40 8000 1 72 0.5 14 40 8000 1 97 0.7 15 40 8000 -1 72 0.7 16 40 8000 -1 97 0.5 E. Analysis of S/N Ratio To measure the performance characteristics, Taguchi recommends the use of the loss function that is deviating from the desired target value. The value of this loss function is further transformed into signal-to-noise (S/N) ratio or SNR. The preferred parameter settings are determined through SNR analysis where the optimized parameters are considered those that maximize the appropriate SNR [15-17]. To analyze SNR three types of performance characteristics are considered. 1) The Smaller the Better (for making the system response as small as possible) =-10log( ∑ ) (3) 2) The Nominal the Best (for reducing variability around a target) =10 log( ) (4) 3) The Larger the Better (for making the system response as large as possible): =-10 log( ∑ ) (5) Lower values of passenger seat displacement and displacement settling time are the prime requirements to achieve passenger RC and safety, thus in present work, SNR with lower-is-better characteristic is selected. Table V represents the L16 orthogonal array with acceleration as output. Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3218-3222 3221 www.etasr.com Yerrawar & Arakerimath: Parametric Analysis of Magnetorheological Strut for Semiactive Suspension … TABLE V. TAGUCHI L16 ARRAY WITH ACCELERATION AND SN RATIO R u n A ir P re ss u re S p ri n g S ti ff n es s C y li n d er M a te ri a l S p ru n g M a ss C u rr en t A cc el er a ti o n (R C ) S N R 1 34 14000 1 72 0.5 0.7250 2.79324 2 34 14000 1 97 0.7 0.6820 3.32431 3 34 14000 -1 72 0.7 1.1150 -0.94550 4 34 14000 -1 97 0.5 0.8852 1.05917 5 34 8000 1 72 0.7 0.9980 0.01739 6 34 8000 1 97 0.5 0.9250 0.67717 7 34 8000 -1 72 0.5 1.2500 -1.93820 8 34 8000 -1 97 0.7 0.9960 0.03481 9 40 14000 1 72 0.7 0.6520 3.71505 10 40 14000 1 97 0.5 0.8250 1.67092 11 40 14000 -1 72 0.5 0.8524 1.38713 12 40 14000 -1 97 0.7 0.7920 2.02550 13 40 8000 1 72 0.5 0.5240 5.61337 14 40 8000 1 97 0.7 0.9152 0.76968 15 40 8000 -1 72 0.7 1.2100 -1.65571 16 40 8000 -1 97 0.5 1.1600 -1.28916 F. Optimum Level Prediction The results obtained are depicted in Table VI. It is observed that cylinder material ranks first and Figure 4 shows the optimum parameter level. From Table VI it is observed that cylinder material performs ranks first and from the main effect plot of SN Ratios (Figure 4), P34K8000CM-1M97I0.7 shows optimum level of parameter. TABLE VI. SNR RESPONSE TABLES Level Air Pressure Spring Stiffness Cylinder Material Sprung Mass Current 1 0.6278 0.2787 -0.1652 1.1233 1.2467 2 1.5296 1.8787 2.3226 1.0341 0.9107 Delta 0.9018 1.6001 2.4879 0.0893 0.3360 Rank 3 2 1 5 4 Fig. 4. Main effect plot for SNR V. SNR INTERACTION PLOT Taguchi designs conventionally focus on the main effects, but it is important to test the supposed interactions. In order to determine the effect of one on another parameter with respect to its response, Taguchi uses a method known as interaction plot. To study the interaction plot, the interpretation of Table VII is considered. According to SNR response table, ranks are established. From Table VI, cylinder material takes first rank. In order to check the cylinder material’s effect on the other four parameters, its interaction plot with tire pressure, spring stiffness, sprung mass and current is plotted. Table VII represents the interpretation of interaction lines. The regression equation (6) is generated from regression analysis. _ 1 1.718 0.0134 0.000030 0.1259 0.00073 0.134 RC P K CM M I       (6) TABLE VII. INTERPRETATION OF INTERACTION LINES Position of Interaction Lines Interpretation Parallel No interaction between two factors Not Parallel Interaction between two factors VI. CONCLUSION Taguchi L16 array and SNR analysis show the cylinder material with Al and CS for damper cylinder is a key parameter for performance measure of semi active suspension with MR damper. It is observed that damper cylinder material is a key parameter to design the MR damper. From the interaction plot, it is concluded that cylinder material interacts more with spring stiffness and tire pressure compared to sprung mass and current. The low price model developed will have wide range of applications to Indian auto industry. Fig. 5. Interaction plot for SN ratios ACKNOWLEDGMENT This work is supported by BCUD, Board of College & University Development, Savitribai Phule Pune University, Pune, India under University Research Grant Scheme and we would like to thank MES College of Engineering, Pune for providing the research facilities. REFERENCES [1] X. Jiang, J. Wang, H. Hu, “Designing and Modeling of a Novel Magneto-rheological Fluid Damper Under Impact Load”, 3rd International Conference on Mechanical Engineering and Mechanics, Beijing, China, pp. 203-207, October 21−23, 2009 [2] S. K. Mangal, A. Kumar, “Experimental and Numerical Studies of Magnetorheological (MR) Damper”, Chinese Journal of Engineering, Vol. 2014, Article ID 915694, 2014 [3] Z. D. Xu, D. H. Jia, X. C. Zhang, “Performance tests and mathematical model considering magnetic saturation for magnetorheological damper”, 4034 2.5 2.0 1.5 1.0 0.5 0.0 140008000 1-1 9772 0.70.5 P M ea n of S N r at io s K CM M I Main Effects Plot for SN ratios Data Means Signal-to-noise: Smaller is better 3.0 1.5 0.0 140008000 9772 3.0 1.5 0.0 3.0 1.5 0.0 4034 3.0 1.5 0.0 0.70.5 3.0 1.5 0.0 1-1 P K CM M I 34 40 P 8000 14000 K -1 1 CM 72 97 M 0.5 0.7 I Interaction Plot for SN ratios Data Means Signal-to-noise: Smaller is better Engineering, Technology & Applied Science Research Vol. 8, No. 4, 2018, 3218-3222 3222 www.etasr.com Yerrawar & Arakerimath: Parametric Analysis of Magnetorheological Strut for Semiactive Suspension … Journal of Intelligent Material Systems and Structures, Vol. 23, No. 12, pp. 1331–1349, 2012 [4] H. Krishna, H. Kumar, K. Gangadharan, Journal of The Institution of Engineers (India): Series C, Vol. 98, No. 4, pp. 533-539, 2017 [5] B. Sapinski, “Analysis Of Parametric Models Of MR Linear Damper”, Journal Of Theoretical And Applied Mechanics, Vol. 41, No. 2, pp. 215- 240, 2003 [6] I. Preda, “About the preliminary design of the suspension spring and shock absorber”, in: IOP Conf. Series: Materials Science and Engineering, Vol. 147, IOP Publishing, 2016 [7] Z. K. Peng, Z. Q. Lang, L. Zhao, S. A. Billings, G. R. Tomlinson, P. F. Guo, “The force transmissibility of MDOF structures with a non-linear viscous damping device”, International Journal of Non-Linear Mechanics, Vol. 46, No. 10, pp. 1305-1314, 2011 [8] S. Segla, J. Kajaste, P. Keski-Honkola, “Optimization of Semi-active Seat Suspension with Magnetorheological Damper”, in: Vibration Problems ICOVP 2011, Springer Proceedings in Physics, Vol. 139, pp. 393-398, Springer, 2011 [9] H. F. Lam, C. Y. Lai, W. H. Liao, Automobile Suspension Systems with MR Fluid Dampers, Technical Report, Smart Materials and Structures Laboratory, Department of Mechanical and Automation Engineering, The Chinese University of Hong Kong, 2002 [10] E. Sert, “Improvement of the Vehicle Stability Using Suspension Optimization Methods”, International Journal of Automotive Engineering and Technologies, Vol. 6, No. 2, pp. 70–84, 2017 [11] M. B. S. Sreekar Reddy, P. Vigneshwar, D. Raja Sekhar, K. Akhil, P. L. Reddy, “Optimization Study on Quarter Car Suspension System by RSM and Taguchi”, in: Proceedings of the International Conference on Signal, Networks, Computing, and Systems, Lecture Notes in Electrical Engineering, Vol. 396, pp. 261-271, Springer, 2016 [12] J. H. Kim, H. C. Sin, B. J. Kang, N. W. Kim, “Characteristic study of bushing compliance in consideration of stresses in a vehicle suspension system by the Taguchi method”, in: Proceedings of the Institution of Mechanical Engineers, Part D: Journal of Automobile Engineering, Vol. 220, No. 10, pp. 1383-1399, 2006 [13] S. Liao, K. Cao, H. Wei, J. Liu, C. Xie, “Robust Optimization of Multi- link Suspension Compliance Characteristics Based on Taguchi Method”, Advances in Computer Science Research, Vol. 74, pp. 726-132, Atlantis Press, 2017 [14] ISO2631-1, Mechanical vibration and shock – Evaluation of human exposure to whole-body vibration – Part 1: General requirements, International Organization for Standardization. Geneva, pp. 1-31, 1997 [15] G. M. Reddy, V. D. Reddy, B. S. Kumar, J. Shyamsunder, “Experimental Investigation on Radial Ball Bearing Parameters Using Taguchi Method”, Journal of Applied and Computational Mechanics, Vol. 4, No. 1, pp. 69-74, 2018 [16] M. B. Silva, L. M. Carneiro, J. P. A. Silva, I. S. Oliveira, H. J. I. Filho, C. R. de Oliveira Almeida, “An Application of the Taguchi Method (Robust Design) to Environmental Engineering: Evaluating Advanced Oxidative Processes in Polyester-Resin Wastewater Treatment”, American Journal of Analytical Chemistry,Vol. 5, pp. 828-837, 2014 [17] C. F. J. Kuo, M. Y. Dong, “Application of the Taguchi Method in Analyzing the Impact of Modified Gemini Surfactants on TiO2 Nano- Suspension”, Journal of Surfactants and Detergents, Vol. 15, No. 4, pp. 471–476, 2012