The application of neural computing methods to the modelling of fatigue in Ni-base superalloys

The current financial climate is driving a move towards increased use of computer modelling techniques in alloy design and development in order to reduce cost. In this paper the potential for use of neural computing methods in the prediction of fatigue resistance in Ni-base superalloys is assessed. Initial work has been conducted on the Stage II (Paris regime) behaviour, as the literature indicates that this is the simplest region of the fatigue crack growth curve to predict, with an approximately linear relationship existing between log(da/dN) and log(Delta K), and the crack growth rates being principally affected by temperature, Young's modulus and yield strength. These three parameters were chosen for initial data collection and modelling. The predictions made are of fatigue life, calculated from the slope and intercept values of the linear portion of the log-log fatigue crack growth curve. A test dataset has been successfully predicted along with the trends in the data. The effect of adding ultimate tensile strength and electron valencies as inputs to the model is assessed. It is shown that validation of models produced against metallurgical experience, and careful construction of the database are important conditions for effective use of neural network models for fatigue life predictions.