Neural Topology and Manufacturing Optimization
Published:
This framework employs three implicit neural fields to represent geometric shape, layer sequence, and fiber orientation. It enables direct formulation of both design and manufacturability objectives into an integrated and differentiable optimization process.
Key Features
- Co-optimization: Simultaneous optimization of structure and manufacturing parameters
- Anisotropic strength: Achieves strong directional strength in fiber-reinforced composites
- Manufacturability: Ensures manufacturability constraints for multi-axis 3D printing
- Performance: Up to 33.1% improvement in failure loads compared to traditional methods
Links
- GitHub Repository: NeuralTOMO
- Publication: ACM TOG 2025