期刊名称：Journal of Computational Science and Technology
电子版ISSN：1881-6894
出版年度：2011
卷号：5
期号：1
页码：1-12
DOI：10.1299/jcst.5.1
出版社：The Japan Society of Mechanical Engineers
摘要：Recent earth environment problem is accelerating rapid popularization of hybrid or electric vehicles replacing conventional vehicles powered by fossil fuels. The electric vehicles don't need exhaust pipes and power train system like propeller shaft under the floor, so their floors can be flattened. Flat floor may also contribute to crash safety and suitable structural material for this purpose will be required. In this paper, authors studied optimal shape to improve energy absorption ability of truss core panel which the usage is considered as vehicle structure. Effective optimization technique especially for nonlinear problem, Radial Basis Function (RBF) network with Response Surface Method (RSM) is adopted in the paper and applied to optimize the shape of truss core panel. Energy absorption quantities for x, y, and z directions were treated as independent objectives and multi-objective optimization was performed. A commercial preprocessor HyperMorph is used for morphing, explicit FEM software LS-DYNA is used to solve crash analysis and LS-OPT is used to handle multiple jobs and optimization process. As the result, 7.1 % of improvement for the energy absorption for the crash in x direction was achieved and Pareto curves for the objectives were obtained.
