Prediction of fatigue life for spot welds using back-propagation neural networks

In this research, the authors developed back-propagation neural networks (BNNs) to predict the fatigue life of spot welds subjected to various geometric factors and loading conditions. This paper described the developing procedures of the BNNs in detail for the spot weld fatigue. Then, the BNNs developed in this study were employed to predict the fatigue life of the spot welds subjected to combined tension and shear loading and tensile-shear loading. The BNNs showed reasonable accuracy for the cases tested in this study. The BNNs can be employed as a practical method for calculating fatigue life of spot welds under various geometric dimensions and loading conditions. (c) 2006 Elsevier Ltd. All rights reserved.