Development of a hybrid topology optimization framework and its application on concurrent mold and part design

In mechanical component design, balancing structural lightness and high performance with manufacturability poses a fundamental engineering challenge. Traditional design methods often fall short in simultaneously addressing these multidisciplinary requirements. This study aims to bridge this gap by developing a hybrid optimization platform that integrally optimizes structural performance and manufacturing constraints. The proposed approach combines density-based topology optimization with advanced steps such as submodeling, shape optimization, element removal, and residual stress simulation. This integration ensures that the final design is both mechanically efficient and practicable for casting and polymer-based manufacturing methods (3D printing, injection molding). At the core of the methodology lies the Submodeling–Density–Shape–Element Removal (S-D-S-ER) algorithm, which focuses computational resources on critical regions, enabling high-resolution stress analysis and shape optimization. This strategy reduces computational time while enhancing design accuracy. To model manufacturing-induced residual stresses, the developed AETDPA (Adjacent Element Temperature Driven Prestress) algorithm integrates thermal stress fields with mechanical loads, providing a realistic simulation environment. Another significant outcome of the research is a greedy search-based algorithm for the automatic placement of runners and ingates. This allows for the simultaneous optimization of part and mold geometry, directly incorporating manufacturing constraints like parting direction and material flow into the design loop. Experimental validation was performed on 3D-printed and injection-molded polymer specimens. Bending, tensile, and fatigue tests revealed a high correlation between simulation predictions and experimental observations. Furthermore, the primary influence of micro-macro geometric parameters on fatigue life was emphasized, demonstrating the framework's ability to co-optimize these parameters through shape optimization. In conclusion, this work provides a comprehensive, process-oriented, and geometry-driven design methodology for digital casting engineering and advanced manufacturing systems, effectively bridging the gap between digital design and physical production realities.