Impact Performance Prediction and Optimization of a Circumferentially Corrugated Tube with Variable Wall Thickness Using Support Vector Machine

Based on the hypothesis that multi-corner and multi-cell structures can effectively improve energy absorption behavior, this paper designed a circumferentially corrugated tube (MCCT) for absorption. The MCCT was as variable thickness form to study influence of materials distribution cross section performance investigated under impact condition with finite element simulation validated by drop hammer test. Support vector machine, machine learning technique, used predict further optimization MCCT. results show same mass, decreasing wall (corners thicker than other regions) shows 4.81%, 30.67% 37.70% improvement, respectively, in PCF, SEA CFE, compared increasing thinner regions). Moreover, most samples Pareto front lie region tc > tm. These all indicate performs better regards In conclusion, arranging more corner area characteristics thin-walled tube. This highlights importance designing tubes configurations