Discrete thickness optimization of an automobile body by using the continuous-variable-based method

Abstract

Design optimization of an automobile body for dynamic stiffness improvement is presented. The thicknesses of plates consisting of a monocoque body of an automobile are employed as design variables for optimization whose objective is to increase the first torsional and bending natural frequencies. By allotting one design variable to each plate of the body, compared to previous works based on element-wise design variables, the design space of optimization can be reduced to a large extent. Because the present optimization is based on continuous-variable-based algorithms, considering manufacturability of the optimized result, the converged values of plate thicknesses should be approximated to commercially available discrete values. A new straightforward thickness discretization scheme considering design sensitivities and employing a subsequent reduced optimization problem is proposed. The validity of the proposed thickness discretization scheme is verified through numerical experiments.

Multi-objective crashworthiness optimization of vehicle body using particle swarm algorithm coupled with bacterial foraging algorithm

“…1 Generally, the performance of BIW can be improved by using novel materials, advanced processing techniques and structural optimization methods. However, in the structural optimization of vehicle components, considerable literature has been limited to the topology optimization and size optimization, [1][2][3][4] which achieve lightweight and performance improvement mainly through the reasonable distribution of material and the thicknesses optimization of sheet metal. Fairly complex BIW geometrical shapes are primarily composed of thin-walled beam structures, and thus have greater potential to improve BIW performance by optimizing the shape of the cross-section of components.…”