Lightweight design optimization of two-layer corrugated cored sandwich panel under blast loading using surrogate-assisted different evolution for mixed-integer variables

Abstract

The sandwich structure stands out as an ideal candidate for blast-resistant armor due to its excellent performance, of which the optimization typically involves time-consuming finite element simulations. Recently, surrogate models have emerged as a promising alternative for time-consuming simulations to accelerate the optimization process of blast-resistant sandwich panels. However, existing methods are limited to optimizing a handful of continuous variables, and often focus on single-layer core sandwich panels. This article introduces a novel constrained optimization model for Two-layer composite Corrugated cored Sandwich Panels (TCSP). The design variables include not only continuous variables such as structural component heights and thicknesses, but also integer variables such as material selection and the number of core unit-cells. To tackle this complex optimization problem with mixed-integer variables, a surrogate-assisted differential evolution (SADE-TCSP) is proposed. This algorithm utilizes the predictive capabilities of surrogate models to guide the optimization process towards a high-quality feasible solution within a very limited computational budget. In order to validate our approach, an optimization framework integrating MATLAB and ABAQUS is developed to conduct numerical experiments. Experimental results show that the proposed method can achieve a 51.04 % reduction in total structural mass while satisfying performance constraints, which is significantly better than existing methods.