PhD candidate in Mechanical Engineering at Northwestern University focused on HPC, GPU acceleration, numerical simulation, and machine learning. Creator of GO-MELT, a GPU-accelerated thermal simulation framework delivering a 678x speedup for metal additive manufacturing. My work bridges physics-based modeling, surrogate modeling, and workflow automation to build scalable, reproducible tools for scientific and engineering applications.
GO-MELT integrates GPU-optimized multilevel FEM to efficiently resolve melt-pool and part-scale thermal behavior across complex AM geometries.
HASTE integrates data-driven surrogate models into GO-MELT, enabling seamless switching between FEM and ML predictions during near-steady-state melt pool evolution.
LinkedIn: linkedin.com/in/joseph-p-leonor