High-dimensional Bayesian inference for crystal plasticity parameter identification of Hastelloy X produced via laser powder bed fusion under cyclic loading

Journal article

Crystal plasticity (CP) modeling is a vital tool for predicting the mechanical behavior of materials but requires calibration of numerous (>8) constitutive parameters, often through tedious trial-and-error methods. This paper proposes a robust approach using Bayesian inference to identify optimal CP model parameters under fatigue loading conditions. Utilizing cyclic data from additively manufactured Hastelloy X specimens at 500 °F, the inference framework, integrated with Gaussian process surrogate model, significantly reduces the number of required simulations. The framework significantly reduces computational cost, achieving accurate calibration with only 50 initial simulations and 75 optimization iterations. The results show that the optimized parameters align well with the experimental stress-strain data, and the use of the newly develpoed objective function may lead to an improved representation of the hardening behavior between two consecutive cycles. Sensitivity analysis reveals the influence of CP parameters at various locations on the stress-strain curve. It shows that the stress-strain response is primarily governed by yield-related parameters, while the kinematic hardening parameter influences the stress at peak strain. In summary, this approach can enable efficient and accurate calibration of CP models under cyclic loading, facilitating improved predictive capabilities for fatigue performance and accelerating the design of new materials with tailored fatigue resistance.

Authors: Kushwaha, A; Demir, E; Basak, A

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Elsevier
• Journal: Mechanics of Materials
• Volume: 211
• Article number: 105500
• Publication date: 2025-09-12
• Acceptance date: 2025-09-11
• DOI: 10.1016/j.mechmat.2025.105500
• EISSN: 1872-7743
• ISSN: 0167-6636
• Language: English
• Keywords: microstructural features; Bayesian optimization; sensitivity analysis; Hastelloy X; strain-gradient crystal plasticity; parameter identification