Generating a Digital Twin of the Laser Powder Bed Fusion Process and for Predicting Nonlinear Pseudoelasticity in NiTi Lattice Structures

This work examines dual computational frameworks designed to optimize metal additive manufacturing. The first section addresses process control by generating a digital twin of the Laser Powder Bed Fusion (LPBF) process. By integrating numerical simulations, machine learning (ML), and in-situ monitoring, this digital twin models printed part quality to enable real-time closed-loop quality control, achieving "first-time-right" fabrication without costly empirical parameter tuning.

The second section highlights recent research on lattice property control, specifically predicting nonlinear pseudoelasticity in NiTi lattice structures. By combining Finite Element Analysis (FEA) with ML, complex mechanical responses—including Young's modulus, yield stress, Poisson's ratio, and stress distribution—are mapped into comprehensive structural databases. These datasets train ML algorithms to evaluate and optimize novel lattice configurations, dramatically accelerating the design cycle for high-performance, smart material NiTi actuators.