Study On Multi-Objective Optimization Of A Coach Retrofit Design Based On The Mesh Morphing Method

CAD data need to be changed repeatedly according to CAE simulation results duringthe traditional method of coach body design, it takes a long time. Morphing technology canrealize mesh deformation of the finite element model of the original models directly andgenerate new models of CAE models without modifying the CAD data. Morphingtechnology and optimization theory are used to realize multi-objective optimization of coachretrofit, which greatly shorten vehicle development cycle.As coach currently tends to be more high-end、more large-scale and more safe, thispaper intends to realize size retrofit design of a low-floor coach, aimed at developing a newcoach with longer, taller and wider overall dimensions on the foundation of the original finiteelement modelFirstly, the change of length, height and width were selected as design variables, basedon original coach model, modification scheme was put forward to determine design variablespace including60sample points generated by optimal Latin Hypercube. Finite elementmodels of60bank new models were created quickly by software MeshWorks/Morpher.Then FE simulation analysis of60new models were conduct to get torsional stiffness、open deformation rate、first-order torsional mode and body mass. According to fourresponses, approximation model of high reliability was established to replace simulatedprogram. Finally, NSGA_II was applied to multi-objective optimization with the change oflength, height and width selected as design variables, bending stiffness and Sheddingdeformation rate selected as objectives，mass and first order frequency taken as constrains.Pareto optimal set was obtained through optimization, and one of Pareto optimal set wastaken as modification design scheme to product new FE model taking into account themodifications goal and both torsional stiffness and the opening deformation rate request.Rollover simulation analysis was conducted with finite element model of the modifiednew model. When rollover safety of the new model is guaranteed, the new models were used to superstructure lightweight design. Five groups bar thickness were selected in the upperstructure of the bus as design variable of orthogonal array experimental design to get16groups of sample points. Rollover simulation analysis of16groups test programs offersrelative intrusion of living space and mass as objects to build neural network model andoptimize coach upper structure, which reduces the mass15.05kg of coach based rolloverregulations.The method of multi-objective optimization of a coach retrofit design based on themesh morphing method of this paper provides a feasible way to guide the retrofit design ofactual coach. This method can continually be explored and improved to really make itthrough the process of developing new models.