Estimating the S-N Curve by Machine Learning Random Forest Method

Abstract

Fatigue limit is well predicted by tensile strength or hardness, and the relationship is often discussed by the linear regression by the minimum squared approximation. However, the prediction of the number cycles to failure at given stress amplitude, which means the prediction of S-N curve, has not been realized. The present study, therefore, aims to investigate the estimability of the S-N curve by the random forest method based on the data described in NIMS fatigue data sheet. The random forest method is one of the machine learning algorithms and is an ensemble learning algorithm that integrates weak learners of multiple decision tree models to improve generalization ability. It was clarified that the machine learning of multiple decision tree model is excellent in fatigue limit prediction. The S-N curve can be accurately estimated by combining the prediction of fatigue limit and that of the number of cycles to failure at given stress amplitude.