Surrogate-Based Global Optimization of Composite Material Parts under Dynamic Loading

This work presents the implementation of the Efficient Global Optimization (EGO) approach for the design of composite materials under dynamic loading conditions. The optimization algorithm is based on design and analysis of computer experiments (DACE) in which smart sampling and continuous metamodel enhancement drive the design towards a global optimum. An expected improvement function is maximized during each iteration to locate the designs that update the metamodel until convergence. The algorithm solves single and multi-objective optimization problems. In the first case, the penetration of an armor plate is minimized by finding the optimal fiber orientations. Multi-objective formulation is used to minimize the intrusion and impact acceleration of a composite tube. The design variables include the fiber orientations and the size of zones that control the tube collapse. The results show the versatility of the algorithm in the design of composite parts, which involve constrained, mixed-integer and multi-objective optimization problems. In the case of single objective problems, the algorithm finds global solutions. When working with multi-objective problems, an enhanced Pareto is provided.