About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Computational Materials for Qualification and Certification
|
| Presentation Title |
Bayesian Modeling for Concurrent Process and Part Design
for Large Scale Additive Manufacturing |
| Author(s) |
Christopher C. Bock, Masoud Rais-Rohani, Brett Ellis |
| On-Site Speaker (Planned) |
Christopher C. Bock |
| Abstract Scope |
Batch processes such as Large-Scale Additive Manufacturing (LSAM) present design challenges due to process-dependent structures and properties, uncertainties, and expensive operation. This work seeks to address these challenges via a Bayesian model of process-structure-property (PSP) relations for carbon fiber reinforced polyethylene terephthalate glycol (cfPETg) trained by optical microscopy, computed tomography, 4-point-bending, short-beam shear, and density data. The trained Bayesian model results in distributions of processes, structures, and properties responses, thus allowing PSP uncertainty to be quantified and utilized for design. The usefulness of this approach is demonstrated via a benchmark problem for the reliability-based design of a cfPETg LSAM-manufactured boat hull. This work is important because it demonstrates a method to account for uncertainty in the concurrent design of processes and part geometry. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
ICME, Additive Manufacturing, Composites |