Session: 26-03 Probabilistic Lifing and Damage Tolerance

Submission Number: 178827

Probabilistic Low Cycle Fatigue Predictions for Nickel Based Alloy 

Extensive experimental observations demonstrated that the variation in the fatigue life of polycrystalline nickel-based superalloy used for rotating parts of gas turbines is due to the microstructure variability, especially to the grain size distribution and micro-level features like annealing twins [1]. Moreover, unintentional features like pores, microcavities, metallic and ceramic inclusions resulting from fabrication process and acting as local stress risers may decrease the number of cycles to failure, potentially increasing the LCF standard deviation [2].

To account for this variability, a low cycle fatigue methodology, sensitive to the microstructure was developed for a series of polycrystalline high temperature nickel-based superalloys. The approach starts with generation of a digital twin of the fatigue specimen or part subjected to cyclic loading. The model evaluates next, the life associated with each microstructural feature. It accounts for mechanisms like crack initiation within each grain, small (microstructure sensitive) and long crack propagation.  The use of multiple possible instances of the fatigue specimens or parts mimic the fatigue life experimental procedure where each specimen/part has a unique predicted life. Thus, the predicted population has a distribution that can be statistically compared with the experimental life population.

The efficiency is assured by replacing explicit physics-based simulations with surrogate models. The surrogate modes are determined from independent finite element-based crystal plasticity simulations, specially designed to incorporate representative microstructures, using robust data interpretation methodologies (e.g. analytic expressions, deep learning algorithms, etc.).  

The model successfully predicted the life for production nickel-based superalloys at temperatures higher than 400^oC, and for stress amplitudes above and below the macroscopic yield.

The current study expands the application of the existent framework to a broader range of microstructural features, like metallic or ceramic inclusions with different stiffnesses, volumes and shapes.

 

[1] Stinville, J.C. et Al. (2018). Fatigue deformation in a polycrystalline nickel base superalloy at intermediate and high temperature: Competing failure modes. Acta Materialia, 152, pp.16-33.

[2] Huron, E., & Roth, P. (1996). The Influence of Inclusions on Low Cycle Fatigue in a P/M Nickel-Base Disk Superalloy. Superalloys 1996 (pp. 359-368). The Minerals, Metals and Materials Society.

Presenting Author: Simone Romano Avio Aero, GE Aerospace

Presenting Author Biography: Dr. Simone Romano is currently Senior Engineer (CTH) in life methods at Avio Aero, A GE Aerospace Company. He has over six years of experience at GE Aerospace, following a research role in Politecnico di Milano. Dr. Romano holds an MSc and PhD in Mechanical Engineering from Politecnico di Milano, where he focused on fracture mechanics and probabilistic fatigue approaches. At Avio Aero, he is part of the lifing team supporting engine and gearbox programs across all phases of product development, and he contributes to the development of new life assessment methods with a particular focus on fracture mechanics and probabilistic approaches.

Authors:

Monica Soare GE Aerospace Research
Umit Ozkan GE Aerospace Research
Simone Romano Avio Aero, GE Aerospace
Rajiv Sampath GE Aerospace
Kevin Eck GE Aerospace