| About this Abstract |
| Meeting |
MS&T24: Materials Science & Technology
|
| Symposium
|
Computational Materials for Qualification and Certification
|
| Presentation Title |
Quantification of Microstructure-Induced Uncertainty in Fatigue Nucleation in Polycrystalline Materials |
| Author(s) |
Caglar Oskay, Xiaoyu Zhang |
| On-Site Speaker (Planned) |
Caglar Oskay |
| Abstract Scope |
Prediction of fatigue initiation is a stochastic multiscale problem since initiation location and the number of cycles to initiation are strongly affected by both structural configuration and loading, as well as the material microstructure. We present a computational framework for uncertainty quantification and prediction of fatigue crack nucleation in polycrystalline materials. The framework relies on a reduced order model for “forward” modelling of a statistical volume element (SVE) subjected to fatigue. The reduced order model is efficient enough that we employ classical Monte Carlo techniques to sample the parameter space tp quantify uncertainty. The evolution and statistics of fatigue nucleation process in an SVE are tracked based on the fatigue indicator parameter (FIP) approach. A statistical time acceleration that predicts long time evolution of FIP statistics based on a first few hundred cycles is proposed. The uncertainty quantification framework was calibrated and validated with experiments on titanium alloy (Ti-6242S). |