Prediction of the Fatigue Strength Defining Mechanism of Nodular Cast Iron Based on Statistical Microstructural Features

Nowadays, casting simulations can predict local properties, such as matrix ferrite-pearlite ratio and nodule count. Thus, there is a need for a model capable of using these to predict the fatigue of nodular cast irons. In this paper, we derive the necessary methods for predicting the fatigue of nodular cast irons, where graphites act as crack initiating defects. The graphite and ferrite size distributions were derived based on elementary inputs using stochastic processes to emulate the microstructure patterns. The √area model was used to check crack arrest at the ferrite-pearlite interface of mixed grades, with respective microhardness. We address ferrite and graphite clustering by stochastic processes that have not been addressed before properly. Furthermore, we provide a solution to an important problem that has not been raised in the literature: the largest ferrite defect containing a crack initiating graphite. We propose a model to take into account changes in ferrite hardness by solid solution strengthening. The model predictions were compared to a large amount of literature data with various parameters. Finally, an in-depth analysis of the mechanisms was performed to provide a clear overview of the problem. The method was compared to some of the other methods proposed in the literature.