Phase Field Simulations for Fatigue Failure Prediction in Manufacturing Processes

Abstract

Fatigue failure is one of the most crucial issues in manufacturing and engineering processes. Stress cycles can cause cracks to form and grow over time, eventually leading to structural failure. To avoid these failures, it is important to predict fatigue crack evolution behavior in advance. In the past decade, the phase field method for crack evoluation analysis has drawn a lot of attention for its application in fracture mechanics. The biggest advantage of the phase field model is its uniform description of all crack evolution behaviors by one evolution equation. The phase field method simultaneously models crack nucleation and crack propagation which will be particularly useful manufacturing problems. In this work, we show that the phase field method is capable to reproduce the most important fatigue features, e.g., Paris' law, mean stress effect, and load sequence effects. For efficient computing, a "cycle"-"time" transformation is introduced to convert individual cycle numbers into a continuous time domain. In order to exploit the symmetry property of the demonstrated examples, a phase field model in cylindrical coordinates is presented. Finally, the fatigue modeling approach presented is applied to study a cold forging process in manufacturing.