Low Cycle Fatigue Model Selection by Performance Analysis

Abstract

A reliable low cycle fatigue (LCF) model requires the selection of appropriate damage mechanism components necessary to accurately approximate experimental life data. In particular, two fundamental tasks have to be performed: firstly, identify the most appropriate model among those available in literature, and secondly, verify that the model is appropriate in relation to the operational conditions of the component whose life is under evaluation (e.g., check if the model accounts for all the relevant damage mechanisms and phenomena). The European Creep Collaborative Committee (ECCC) developed a procedure that supports the researcher in evaluating performances and reliability of creep models, known as Post Assessment Tests (PATs). At the moment, there is no equivalent procedure for low cycle fatigue and ECCC work may provide the LCF researcher with useful guidelines. This paper is intended to investigate, compare and suggest which kind of verification is appropriate to identify any possible misbehavior in a fatigue model or in the LCF tests that supported the model. This procedure involves an analysis of the model performances in terms of comparison with the experimental population, supported by a deep knowledge of the damage mechanisms of the given material. Particular attention will be paid to materials with a particularly high dispersion, such as cast nickel-based superalloys. This paper is also meant to stimulate the fatigue data user community to propose and share methodologies in the perspective of the creation of a recommendation code, similar to what has been done by ECCC for creep.