Multiobjective optimization and sensitivity analysis of honeycomb sandwich cylindrical columns under axial crushing loads

Abstract

This paper firstly investigates the energy absorption characteristics of honeycomb sandwich cylindrical columns such as square, triangle, kagome and diamond core under axial crushing loads by nonlinear finite element analysis. The interaction effects between the honeycomb and column walls greatly improve the energy absorption efficiency. The response surface method with cubic basis functions is employed to formulate specific energy absorption and peak crushing force which reduces considerably the computational cost of crush simulations by finite element method. Both the single objective and multiobjective optimizations are performed for columns under axial crushing load with design variables inner, outer and core thickness. Models are optimized by multiobjective particle swarm optimization algorithm to achieve maximum specific energy absorption capacity and minimum peak crushing force. Furthermore, local and global sensitivity analyses are performed to assess the effect of design variable values on the specific energy absorption and peak crushing force functions in design domain.