Content area
This dissertation presents a unified framework for the topology optimization of structural assemblies composed of multiple components, materials, and manufacturing processes. While structural topology optimization (TO) enables the generation of innovative and high-performance structures by material distribution within a given domain, its designs are often monolithic, complicated -- making them impractical for industrial manufacturing.
Structural products are often assembled from simpler components made using different materials and processes. In conventional design practice, engineers typically begin with the optimization of a monolithic (one-piece) structure and then sequentially partitioning it based on manufacturing and joining constraints. However, this two-step approach often leads to suboptimal outcomes, as the optimal decomposition in the second step is decoupled from the optimal structural geometry obtained in the first step.
To address this gap, this work proposes a multi-component, multi-material, and multi-process topology optimization (M3TO) framework. The proposed method simultaneously optimizes the structural topology, component partitioning, material selections, and manufacturing constraints. This method accounts for process-specific geometric constraints and material properties, enabling for more manufacturable yet still structurally optimal solutions.
In the context of die casting, this dissertation discusses a method incorporating arbitrarily curved nonplanar parting surfaces and lateral side dies. These improvements extended the possibility of castable TO designs to include more intricate geometries.
The M3TO framework models each component as a unique material-process pair, enabling the simultaneous consideration of material properties and process-specific manufacturing constraints. This formulation has been applied to hybrid assemblies combining additive manufacturing (AM) and die casting, accounting for geometric constraints such as undercut avoidance and bounding radius. Additionally, the method introduces a joint stiffness constraint based on directional loading behavior, enabling the consideration of anisotropic joint performance between components.
Several numerical examples are presented to demonstrate the capabilities and versatility of the proposed M3TO framework. These examples showcase the proposed method's ability to generate high quality designs balancing structural performance and manufacturability.