| About this Abstract |
| Meeting |
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Symposium
|
2024 ASC Technical Conference, US-Japan Joint Symposium, D30 Meeting
|
| Presentation Title |
Uncertainty Quantification in Advanced Aerospace Composite Manufacturing Through Stochastic Finite Element Analysis and Probabilistic Machine Learning |
| Author(s) |
HUILONG FU, Kendall A. Johnson, Navid Zobeiry |
| On-Site Speaker (Planned) |
HUILONG FU |
| Abstract Scope |
For aerospace composite manufacturing and assembly, high confidence-level models are desired to predict and mitigate defects such as process-induced-deformations (PIDs). These are typically predicted with deterministic high-fidelity Finite Element (FE) process simulations, calibrated and validated by conducting numerous experiments at different scales. However, this approach is very time-consuming, cost-intensive, and neglects the variabilities in the material and process, and their effects on manufacturing defects. To quantify such uncertainties, we propose a novel framework combining fast stochastic FE simulations with probabilistic machine learning. Results from fast FE models are used for Global Sensitivity Analysis (GSA), Neural Network (NN) surrogate modeling, and Markov Chain Monte Carlo (MCMC), and Bayesian Network (BN) modeling to effectively quantify and visualize the effect of uncertainties in formation of manufacturing defects. This paper will demonstrate the capabilities of the proposed framework with case studies concerning manufacturing parameters such as layup and cure cycle. Results contribute to better understanding effects of uncertainties in the manufacturing process, therefore facilitating the prediction and mitigation of manufacturing defects. |
| Proceedings Inclusion? |
Definite: Post-meeting proceedings |