Dr. Ashley Gannon
Manufacturing Science Division
Oak Ridge National Laboratory

Simulation Calibration and Process Optimization for Directed Energy Deposition

ABSTRACT: This seminar will present a two-stage framework that calibrates high-fidelity thermal simulations against in-situ infrared (IR) measurements and optimizes interpass dwell and reheat parameters using Bayesian methods. In the calibration stage, spatially resolved IR temperature fields are projected onto the simulation mesh, and key model parameters (laser absorption and penetration depth, convective heat-transfer coefficient) are adjusted to minimize the discrepancy between simulated and measured thermal histories. Building on the calibrated model, the optimization stage partitions the part geometry and injects candidate dwell durations and reheat power settings into the toolpath. An ensemble of Adamantine simulations evaluates each candidate by quantifying the cumulative time points spent within critical phase-transformation windows for 17-4PH stainless steel. Bayesian optimization then guides successive iterations toward parameter combinations that maximize time in the phase-transformation window across partitions. This pipeline demonstrates how coupling IR-driven calibration with data-driven optimization can improve process development in directed energy deposition and lays the groundwork for future closed-loop control.

BIOGRAPHY: Dr. Ashley Gannon earned her Ph.D. in Computational Science from Florida State University, a Master of Science in Science, Technology, Engineering, and Mathematics Teaching, and dual Bachelor of Science in Chemical & Biomedical Engineering and Biological Sciences. Her work focuses on combining finite element simulations, in-situ monitoring, and toolpath optimization strategies to optimize additive manufacturing processes by reducing residual stress and geometric distortion in printed parts.