Multiaxial fatigue life prediction of metals considering loading paths by image recognition and machine learning

Abstract

The loading path of multiaxial fatigue has an important and direct impact on the life prediction, and machine learning can predict the multiaxial fatigue life. However, machine learning cannot directly accept input from the loading path because it is graphic data. To solve this problem, the image recognition method is used to extract the loading path features and transform them into numerical data. These data are placed in the first layer and together with the second layer of identifiable strain amplitude data to construct a machine learning method based on the bilayer features. This method can predict multiaxial fatigue life of different loading paths. Experimental data of PA38-T6 and E355 under multiaxial fatigue loadings are used to verify the feasibility and applicability of the proposed method. The results show that the proposed method can accurately predict the multiaxial fatigue life and has good extrapolation abilities, and image recognition technology is suitable for the feature extraction of loading paths.