Estimation of fatigue life for aluminium welded joints with the application of artificial neural networks

The aim of this investigation was determining the fatigue behaviour of welded aluminium joints and so the appertaining SN-lines by application of Artificial Neural Network (ANN) architectures. For this, fatigue data obtained with aluminium welded joints subjected to constant amplitude loading were used. The main benefit of ANN is the good description of the effects of different factors on fatigue life. The results determined by the ANN method for four aluminium alloys are displayed in scatter bands of SN-lines. It is observed that the trained results are in good agreement with the tested data and enable the estimation of SN-lines. © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim