Non-parametric optimization for lightweight and high heat conductive structures under convection using metaheuristic structure binary-distribution method

In this paper, we propose a novel Metaheuristic Structure Binary-Distribution (MSB) method for attaining lightweight and high thermal conductive structure. MSB combines metaheuristic search with gradient descent optimization, offering a robust and efficient approach. The paper also introduces a connectivity filtering approach to enhance additive manufacturability and eliminate segregated materials in thermal structures, improving integrity and minimizing materials wastage. The performance of the MSB method is compared to other topology optimization methods, namely bi-directional evolutionary structural optimization (BESO), solid isotropic materials with penalization (SIMP), and the parametrized level set method (PLSM). Results consistently demonstrate the better performance of MSB in weight reduction and minimizing thermal compliance. Even at significant weight reductions (50%, 60%, and 70%), MSB outperforms other methods, achieving lower thermal compliance. Investigations on mesh impact and the effectiveness of MSB with suggested connectivity filtering show a successful reduction of islands to solid elements to 0%, improving additive manufacturability and decreasing thermal compliance. The feasibility and suitability of the optimized designs for additive manufacturing are validated through 3D printing, confirming the practicality of the MSB method.