Design for Machinability: Topology Optimization for Subtractive Machining Using Trimmed Neural Fields
ACM Transactions on Graphics, 2026
Neural fields that bring tool accessibility and efficient flank milling into structural optimization.
Postdoctoral Associate · MIT Mechanical Engineering
I am a Postdoctoral Associate in the DeCoDE Lab at the Massachusetts Institute of Technology, working with Faez Ahmed. My research brings together computational design, machine learning, and manufacturing. I develop neural representations and generative models for topology optimization, with a focus on efficient, controllable, and manufacturable structural design.
I received my Ph.D. in Mechanical Engineering from Carnegie Mellon University in 2025, advised by Levent Burak Kara, and my B.S. in Mechanical Engineering from the University of Wisconsin–Madison in 2020.
ACM Transactions on Graphics, 2026
Neural fields that bring tool accessibility and efficient flank milling into structural optimization.
ASME IDETC/CIE, 2026
An agent-based workflow translating qualitative design preferences into optimized structures.
ASME IDETC/CIE, 2026
Controlling member thickness, length, and connectivity through guided latent diffusion.
arXiv preprint, 2026
Fast, physics-aware topology warping, lattice replacement, and no-design region editing.
Journal of Mechanical Design, 148(10):101705, 2026
Multiscale neural representations for connected metamaterials with spatially varying properties.
Computers in Industry, 168:104286, 2025
Connecting design requirements and manufacturing resources to generate feasible parts.
Engineering with Computers, 41:3313–3327, 2025
Direct optimization of microstructures with smooth connectivity across neighboring cells.
Additive Manufacturing, 84:104095, 2024
Predicting melt pool geometry from path-level thermal history with a generative model.
Engineering with Computers, 40:2227–2240, 2024
Coupled density and displacement networks for direct, mesh-free topology optimization.
Additive Manufacturing, 83:104057, 2024
Inferring melt pool behavior and defects from acoustic and thermal sensing.
Journal of Mechanical Design, 146(6):061702, 2024
Strain-energy conditioning fields improve the efficiency of neural topology optimization.
Journal of Mechanical Design, 145(9):091702, 2023
Jointly optimizing geometry, segmentation, and print direction to reduce support material.
ASME IDETC/CIE, 2021
Geometry-aware generative learning for realistic heterogeneous material microstructures.
ASME IDETC/CIE, 2020
Density-adaptive toolpaths and continuous Fermat curves for fused filament fabrication.
SPE ANTEC, 2020
Combining printed structures and composites in the design of a lightweight bike saddle.
Research code and example implementations from my GitHub.