Artificial Neural Network – An Application to Predict the Fatigue Crack Growth Rate of HSLA Steel and Copper

Abstract

In the present work , a prediction method has been used to describe the life of High Speed Low Alloy steel (HSLA Steel ) and Copper under constant load ratio by using Artificial Neural Network (ANN). Therefore a methodology has been developed to determine the fatigue crack growth rate (da/dN) of HSLA steel and Copper under constant amplitude loading at different load ratios i.e. R = 0, 0.2, 0.4, 0.5, 0.6 and 0.8 by adopting an exponential model to raw experimental a – N data. A soft-computing technique, i.e. Artificial Neural Network (ANN) has been formulated and implemented to estimate the fatigue life at R = 0.5. A comparison has been made with experimental data obtained by earlier researchers and found to be within limits and in good agreement. It is observed that percentage deviations from the experimental values for HSLA steel and Copper are 4.14 and 4.574 respectively. The error values are well within limits of -0.06% and -0.09% for both the materials.