Multivariable Crashworthiness Optimization Of Vehicle Body By Factorial Design

Crashworthiness is an important content for automobile passive safty research. Itis known that the automobile is consist of thousands of components, the geometricand material parameters of these components may affect the automobile passive safty.However, all the potential parameters have different degree of influence on theautomobile carshworthiness. How to single out the most remarkable factors for theautomobile carshworthiness is a problem to be solved during multivariablecarshworthiness design for the vehicle body. There are two common phenomenaexsiting in the available research, one is that a limited number of factors are selectedto be design variables for carshworthiness optimization design, which may miss someimportant factors; another is that researchers blindly regard all the potential factorsas design variables without factor screening process, which may lower the computingefficiency. However, the unreplicated saturated factorial design is a factor screeningmethod with high efficiency, it can study any level of experimental design problem,and analyze the complicated relationship between objective indicators and analysisfactors, all the potential combinations of analysis factors can be considered, includingthe interaction effects. This method has been widely used in biological, medicine,chemistry and other fields.Therefore, this paper, based on the regulatory requirements of frontal and sideimpact, employs unreplicated saturated factorial design to define the most remarablefactors for frontal and side impact load at the initial automobile design stage, and then,the selected factors are regarded as design variables for the following carshworthinessoptimization design. Thus, the design variable number is rationally reduced, thewhole design cycle is effectively shortened, and the development costs are largelylowered.The method of using unreplicated saturated factorial design to solve the factorscreening problems in the automobile crashworthiness optimization design isproposed in this study, the limitations of selecting design variables in terms of theresearchers’ experience are overcome, and inefficiencies of selecting all the potentialfacrors without factor screening process are successfully solved. So this studyproposed an effective factor screening method for the crashworthiness optimizationdesign in the future.