Modelling by artificial neural network of high temperature fatigue life of oxide dispersion strengthened nickel-based superalloy PM 1000

In this study, an artificial neural network model was developed to predict the thermal-mechanical fatigue life and pure isothermal low-cycle fatigue life of oxide dispersion strengthened nickel-based superalloy PM 1000. The input parameters to the model consisted of the concentration of five inputs: mean temperature, temperature amplitude, mean total strain, total strain amplitude, and heating/cooling rate. The calculated results fit perfectly with the experimental data in both types of fatigue experiments. Furthermore, the interactions between heating/cooling rate and thermal-mechanical fatigue life were estimated based on the obtained artificial neural network model.