About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Verification, Calibration, and Validation Approaches in Modeling the Mechanical Performance of Metallic Materials
|
| Presentation Title |
Bayesian Calibration of Crystal Plasticity Finite Element Model Parameters |
| Author(s) |
Ramesh Babu Jangala, Pranav Karve, Sankaran Mahadevan, Kishore Appunhi Nair, Tawqeer Nasir Tak, Somnath Ghosh |
| On-Site Speaker (Planned) |
Pranav Karve |
| Abstract Scope |
We perform Bayesian calibration of crystal plasticity finite element model (CPFEM) parameters for α/β dual-phase Ti-6Al-4V alloy. The high-fidelity CPFEM model accounts for anisotropic elasticity and plasticity; rate and size dependency; as well as isotropic and kinematic hardening. Laboratory test data from only a few experiments is typically available for calibrating this complex model. A Bayesian calibration methodology that incorporates prior information through prior distributions and provides posterior distributions of calibrated model parameters is desirable. It is computationally challenging to compute data likelihood and perform Markov Chain Monte Carlo (MCMC) simulations to obtain posteriors for all the parameters. We select a few important parameters for calibration and train a CPFEM surrogate model to expedite likelihood computation. By integrating prior knowledge with experimental data, the Bayesian approach accounts for uncertainty and adapts to new information. This enhances the model’s reliability in predicting the behavior of Ti-6Al-4V under complex loading conditions. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Computational Materials Science & Engineering, ICME, Modeling and Simulation |