Geometric optimization and performance analysis of radial MR valve using Taguchi orthogonal experiment method
Guoliang Hu,Feng Zhou,Wencai Zhu,Lifan Yu,Gang Li
DOI: https://doi.org/10.1007/s12206-022-0822-2
IF: 1.81
2022-09-03
Journal of Mechanical Science and Technology
Abstract:Aiming at the problems of increasing external dimensions and deteriorating key performance indicators in the design process of magnetorheological (MR) valve by using structural optimization method, a geometric optimization design methodology for the optimal design of a MR valve structure under specific volume constraints is proposed in this article. The optimization methodology couples the finite element model (FEM) constructed in COMSOL software with the Taguchi orthogonal experiment and response surface technology to build an approximate response surface function for the identified independent variables. Suitable optimization algorithms are then utilized to determine the optimal geometry of the MR valve, thereby maximizing the valve performance. Firstly, a radial MR valve with a single excitation coil was presented, and its structure and working principle were also elaborated. A mathematical model of the pressure drop was derived on the basis of the Bingham-Papanastasiou non-Newtonian constitutive model of MR fluid and the magnetic circuit had been analyzed with the FE analysis methodology. Then, a second-order response surface model (RSM) had been fitted for the magnetic flux density in the radial flow channel and spool region of the radial MR valve based on the Taguchi orthogonal experimental design. The fitted model was a function of the four independent variables of the radial MR valve, and the accuracy of the developed response surface function over the entire design space had also been estimated. Meanwhile, predictions made by the RSM and FE models were evaluated by analysis of variance and it was exhibited that the RSM model's results agree with FE result fairly. Subsequently, the geometric optimization problem had been formulated for the constructed RSM exploiting the genetic algorithm to find the global optimum geometrical parameters of the radial MR valve. Furthermore, the experimental test rig was setup to explore the pressure drop and the response time characteristics of the initial and optimal radial MR valve as well as the dynamic performance of the MR valve controlled cylinder system under different excitation conditions. The experimental results show that under the applied current of 2 A, the pressure drop and adjustable coefficient of the optimal radial MR valve observably increased with values of 3.15 MPa and 5.40, respectively, when compared to 2.11 MPa and 4.22 of their respective initial values. Also, at the applied current of 1.25 A, the damping force of the MR valve controlled cylinder system enlarged by 46 %, with its optimal value being 3.65 kN and initial value as 2.50 kN, which was an excellent verification of the correctness of the RSM and the effectiveness of the optimal design.
engineering, mechanical