Optimization of a Vehicle A-pillar in a Rollover Crash
Year: 2009
Editor: Chakrabarti, A.
Author: Paranjape,Rajesh; Bawab,Sebastian; Unal, Resit; Hou,Gene; Ringleb, I.Stacie
Section: Design for Environment, Safety, Manufacture and Cost
Page(s): 144-151
Abstract
A response surface method is used in this paper to optimize the structural dimensions of a vehicular A-pillar considered under a rollover crash impact. The objective is to minimize the weight, while the constraints include the yield stress and the maximum allowable de ection. Different cross sections of an A-pillar were subjected to a NHTSAstandard impact test. The impact test was simulated using mathematical dynamic model (MADYMO) software by TNO Automotive Safety Systems (TASS), Netherlands. Latin hypercube method was used to generate the sample points. A total of 125 simulations were generated to capture the complete spectrum of three design variables with ve levels for each design variable. Response surfaces were obtained by performing regression analysis of the weight, the maximum stress, and displacement in the A-pillar obtained from the results of the simulations. Linear, quadratic, and cubic functions were tted for the stress, the weight,and the displacement. The best t was chosen based upon the R2 and the p-value of the regression analysis.Optimization was then performed to minimize the weight of the A-pillar subjected to the limiting conditionsof the stress as well as the maximum allowable displacement. The optimum set of variables obtained was used for the A-pillar in MADYMO to generate the stress and displacement results. These results produced the optimal solution compared to all 125 original sets of data and closely match the solution obtained from the design of experiments technique
Keywords: A-pillar, Rollover, Design of Experiments, Optimization, Response Surface Method