Designing Porous Structure with Optimized Topology using Machine Learning

Biomedical engineering relies on topology optimization to refine material distribution, crucial for lightweight, high-performance prostheses and orthoses. Advanced manufacturing techniques like additive manufacturing can then be used to create these intricate designs layer by layer, ensuring precision and customization. However, conventional numerical simulation-based topology optimization methods can be timeconsuming and resource-intensive, especially as the design domain expands. To overcome this issue, machine learning models are investigated for their ability to perform topology optimization. The results indicate a significant decrease in computation time, along with comparable optimization performance to conventional methods.