| About this Abstract |
| Meeting |
TMS Specialty Congress 2025
|
| Symposium
|
8th World Congress on Integrated Computational Materials Engineering (ICME 2025)
|
| Presentation Title |
A Framework for Efficient Part-Scale Microstructure Prediction in Laser Powder Bed Ti-6Al-4V Using Combined Physics-Based Modeling and Machine Learning Surrogate Methods |
| Author(s) |
Anthony G. Spangenberger, Bonnie Whitney, Diana Lados |
| On-Site Speaker (Planned) |
Anthony G. Spangenberger |
| Abstract Scope |
Microstructure formation in additive manufacturing (AM) spans lengths from centimeter-scale components to micrometer-scale grain sizes, making their simultaneous resolution in computational simulations intractable with current high-fidelity physics-based methods. These models are needed to simulate microstructure sensitivity to process parameters, as the basis for subsequent property prediction models, and to mitigate process defects and anisotropy. A three-part modeling framework for laser powder bed Ti-6Al-4V is proposed to address this deficit: (i) coupled continuum heat transfer and kinetic Monte Carlo simulation of β grain morphology and texture at the part-scale, (ii) phase field (PF) modeling of the β→α/α’ solid-state transformation at the subgrain-scale, and (iii) a scale-bridging surrogate of the PF model for transient α/α’ phase fraction and width predictions. Model calibration and validation are supported by in-situ synchrotron heat treatment studies and electron backscatter diffraction data that inform transformation kinetics and grain morphological/textural variations across a wide range of processing parameters. |
| Proceedings Inclusion? |
Undecided |