Microstructurally-Based Simulation of Multiaxial Low-Cycle Fatigue Damage of 316L Stainless Steel in Terms of the Behaviour of a Crack Population

Abstract

A statistical simulation of low-cycle fatigue damage is presented. In this two-dimensional model, microcraks are nucleated randomly according to a density which was determined experimentally. These microcracks propagate with a velocity which was also derived from metallographical observations. They subsequently coalesce to lead to final fracture when the calculated “plastic zone” sizes at the tip of two neighbouring cracks are overlapping. This model which was developed for 316 stainless steel tested at RT or at elevated temperature is shown to be able to reproduce very well the results of continuous LCF tests carried out either under tension/compression or under fully-reversed torsion loading. This model accounts also very well for the large deviations from the Miner linear damage cumulative rule which were observed for tension/compression → torsion sequentials tests at 600°C. The scatter as well as the size effects observed in simulated tests can be reprented by a Weibull statistical distribution.