Studies of low cycle fatigue and creep-fatigue-oxidation interaction of material and constitutive modeling

Abstract

The structural materials used in power plants can exhibit initial hardening/softening, prolonged softening/ hardening, saturation, and secondary hardening behavior before the final fracture under cyclic loading. The cyclic deformation of materials also causes the hysteresis loop to show Masing/non-Masing behavior. In the open literature, the non-Masing behavior is recently classified into Type-I (only isotropic stress changes with strain amplitude) and Type-II (both isotropic stress and strain hardening rate changes with strain amplitude). Constitutive models in the literature deal with material fatigue behavior without regard for the Masing/non-Masing behavior. Thus, the constitutive modeling of the above four phenomena with accurate prediction of Masing, non-Masing Type-I, and Type-II is essential for successfully designing and developing components and structures. The constitutive models available in the open literature are primarily developed by modifying the isotropic and kinematic hardening laws. The classical and modified models developed to date cannot predict the material’s behavior exhibiting significant secondary hardening, as shown by some structural steels undergoing low cycle fatigue loading. Moreover, no article in the literature has demonstrated (or validated) any constitutive model for simulating materials' Type-I and Type-II non-Masing behavior. A modified constitutive model is proposed to consider all the cyclic features mentioned above as exhibited by the material under cyclic loading. A new non-dimensional function, 𝜁 (representing cyclic hardening or cyclic softening rate), is introduced in the isotropic and kinematic hardening equations to account for the significant change in softening and hardening behavior with strain amplitude through maximum plastic strain range memory (𝑞). Moreover, to take into account the Masing and non-Masing behavior of Type-I and Type-II, another function, 𝜑𝐾𝐻 (representing the dependence of maximum back stress on strain amplitude), is introduced. The components of power plants undergo creep-fatigue loading at high temperature every day, causing the material to undergo a combination of creep, fatigue, and oxidation damage. In open literature, many methods are available to predict the life of the material. Most of these methods predict the life of material considering creep and fatigue loading without any consideration of oxidation damage. Moreover, most of these methods use data from pure fatigue, creep-fatigue, and pure creep tests. To overcome these problems, the tensile hysteresis strain energy density method (THSED) is investigated, and a modified THSED method is proposed that takes into account the parameter 𝛾𝑑 that depends on oxidation damage, temperature, strain rate of cyclic loading, and hold time. Further, the proposed modified isotropic and kinematic hardening laws are coupled with the unified Chaboche viscoplastic flow rule to predict the time-dependent creep effect. The developed viscoplasticity model further takes into account the complex microstructural degradation effect due to a synergistic combination of creep, fatigue, and oxidation through the incremental scalar damage lifetime rule, which can predict the component's mechanical state and the material's behavior under creep-fatigue loading. The proposed constitutive models are implemented in ABAQUS as user subroutines (UMAT) for simulating the LCF behavior of 304L SS and 321 SS materials and the creep-fatigue behavior of 304L SS material under peak tensile and peak compressive hold for 60sec. The excellent agreement between experimental data and simulated results suggests that the proposed model works well in predicting the cyclic deformation behavior of the materials.