Integrating GA-BEM and polynomial fitting for efficient structural shape optimization

Abstract

This paper presents a novel simplified approach to achieving smooth boundaries on structural shape optimizations when combining Genetic Algorithms (GA) with the Boundary Element Method (BEM) by applying a simple polynomial fitting technique for boundary smoothing. The methodology focuses on the challenges of reducing material usage while maintaining constructability. The integration of polynomial fitting for boundary smoothing mitigates the complexity often resulting from GA-based optimizations, while keeping the complexity of implementation low when compared to other boundary smoothing techniques. Case studies are used to demonstrate the effectiveness of this approach in reducing material usage while following stress and displacement constraints. Comparative analysis with existing methods, such as Isogeometric Analysis-BEM with Particle Swarm Optimization, highlights the efficiency and implementation simplicity of the proposed technique. The results show significant improvements in structural performance and material reduction, demonstrating that the method can be used as a valid tool for structural shape optimization.