Abstract
In this paper, a method to evaluate the cumulative ductility till fracture in steel members subjected to random amplitude loading is presented. The method predicts fracture when the damage variable, evaluated through a combination of Manson-Coffin's law and Miner's rule, equals unity. The effect of mean plastic deformation is taken into account through the method of Sachs and Weiss and the underestimation of Miner's rule in evaluating damage of loading histories which include relatively small amplitudes is remedied by a modification of the loading history through separation of cycles. The proposed method is verified with experimental data from steel members with various material, geometry, and mode of fracture (ductile, brittle, pseudo-brittle). The relative error in the predicted cumulative ductility till fracture in most of the specimens is within 25%.
