Fatigue strength estimation of butt welded joints in magnesium AZ31 alloy using the genetic algorithm

Abstract

Genetic Algorithm is applied to calculate the substitute structural length (SSL) and obtain the fatigue strength of welded joints in magnesium. Fatigue tests with unnotched and notched specimens of the base material metal (BM), the weld metal (WM) and the heat-affected zone (HAZ) of magnesium AZ31 (MgA13) were carried out in order to derive the unknown substitute structural length. Results of fatigue tests with geometrically similar welded joints with full and partial penetration were available. The comparison of the experimental and predicted substitute structural length using Genetic Algorithm Substitute Structural Length Estimation Model (GASLEM) shows that the developed models seem to be capable of predicting the SSL values. Fatigue strength values were also obtained using predicted SSL. © 2008 WILEY-VCH Verlag GmbH & Co. KGaA.