Thermomechanical fatigue life prediction of metallic materials by a gradient‐enhanced viscoplastic damage approach. (24th January 2022)

• Record Type: Journal Article
• Authors: Yin, Bo; Zreid, Imadeddin; Zhao, Dong; Ahmed, Raasheduddin; Lin, Guoyu; Kaliske, Michael
• Is Part Of: International journal for numerical methods in engineering. Volume 123:Number 9(2022)
• Page Start: 2042
• Page End: 2075
• Publication Date: 2022-01-24
• Subjects: continuum damage theory -- gradient‐enhanced formulation -- nonlinear viscoplasticity -- thermomechanical fatigue failure; Numerical analysis -- Periodicals; Engineering mathematics -- Periodicals; 620.001518
• DOI: 10.1002/nme.6926
• Languages: English
• ISSNs: 0029-5981

Abstract

Abstract: Fatigue failure plays an important role in engineering applications, especially when structural components experience significant cyclic thermal loading and complex force loading simultaneously. During the last decades, several post‐processing techniques have been developed based on empirical investigations of experimental evidence to predict the fatigue life of materials. The work at hand postulates a conventional continuum damage theory for thermomechanical fatigue failure modeling. In particular, an implicit gradient‐enhanced approach is employed to address the ill‐posedness of the partial differential equation system when the damage onsets. An internal fatigue variable is phenomenologically defined based on the accumulation of viscoplasticity. In the sequel, a regularized fatigue variable is obtained to further yield the damage softening function, which straightforwardly applies to the stress, material tangent, and viscoplastic dissipation. A multi‐field problem, consisting of the strain field, the temperature, and the non‐local variable, is taken into consideration, leading to a fully coupled system. This numerical methodology is consistently derived and implemented into the context of the finite element method. Several representative and demonstrative examples are performed, which yield good numerical stability and agreement with experimental data. Conclusive findings and further perspectives close this article.