A Multi-Scale Computational Scheme for Prediction of High-Cycle Fatigue Damage in Metal Alloy Components

Abstract

<div class="section abstract"><div class="htmlview paragraph">High-cycle fatigue damage causing micro-crack initiation is a critical concern in aerospace structural components and alloys due to intense thermo-mechanical stress and vibration. Vibration or overload/impact can initiate small cracks near the stress concentration zones. These cracks may expand erratically before being detectable in subsequent inspections, emphasizing the need to predict the effects of usage and aging on components. This predictive ability would significantly aid material refinement, design enhancements, and inspection planning. Prediction of fatigue damage leading to the formation of cracks is a great challenge for many reasons, including microstructure anisotropy and uncertainties in complex stress states compared to design stress used in testing and qualifying a component. These uncertainties undermine inspection reliability and effectiveness. The elastic moduli of the material are considered isotropic and homogeneous at the macroscopic level of continuum plasticity. Effective properties at the microscopic level are anisotropic and are strongly correlated to constituent phases, interphases, and geometric factors like shape, size, and orientation, which are the reasons for anisotropy in elastic moduli. Statistical modeling of microstructure is vital to identify the scatter in the properties, which involves the generation of synthetic microstructure, that is statistically equivalent to experimental microstructure. A multi-scale computational scheme and tool are developed to accurately estimate adequately resolved fatigue damage-induced plastic strain. The damage evolution model developed from constitutive properties at the microstructure level is the precursor for predicting continuum damage. The model developed correlates the damage accumulation and life (in terms of number of cycles). A polygonal finite element scheme developed recently in our previous studies employing a numerical integration scheme is used for modeling complex grain geometries.</div></div>