Abstract: Variable-thickness rolled blanks(VRBs)is an advanced lightweight manufacturing process, the VRB thin-walled structure by which has the advantages of light weight and good energy-absorption effect. Taking the transverse impact process of thin-walled structures with top-hat section as the research object, the high-precision finite element model of VRB thin-walled structures with top-hat section is established. A comprehensive learning particle swarm optimizer (CLPSO) is proposed to handle the optimal thickness distribution problem of VRB thin-walled structures with rolling constraints based on the particle swarm optimizer (PSO). The algorithm can handle the optimal thickness distribution problem of VRB thin-walled section with large-scale design variables under rolling constraints and performance constraints. The effectiveness of CLPSO is verified by carrying out the optimal thickness design of VRB thin-walled structures with or without manufacturing constraints respectively. The optimization results show that the CLPSO algorithm with manufacturing constraints is easier to obtain design results that can be used in manufacturing than the CLPSO algorithm without manufacturing constraints; and under the conditions of meeting the performance constraints and VRB rolling constraints, the weight of thin-walled structures with VRB top-hat structure is reduced by 21.7% compared with the structure with uniform-thickness.

Key words: lightweight, variable-thickness rolled blank (VRB), thickness distribution optimization, comprehensive learning particle swarm optimizer