Observations on assessing ductile instability margins in the presence of fatigue

Abstract

The current situation in respect of analysing a structure against ductile tearing instability in the presence of fatigue is reviewed. Although in its main sections, R6 advises against such analysis under normal operating loads, in its status notes, it does allow the analysis under certain provisos. A brief review of the real behaviour of structures is performed. It is apparent that ductile instability is determined by the material's resistance curve, while the crack growth to that point is determined by the influence of tearing on the fatigue laws. It is shown that restricting an analysis to ductile initiation is arbitrary and gives no indication of the real margins. Indeed, if this is done, margins could be very small without knowing it. Some of the issues associated with assessing for ductile instability in the presence of fatigue are addressed, particularly those associated with variable amplitude loading. It is shown that although it is not possible to assess the instantaneous effect of any load applied to a structure where cracking has entered the tearing-fatigue regime, because of the unknowns associated with variable amplitude load excursions, the limiting case can be assessed, and this can be used to determine margins and associated factors. Uncertainties may arise because of the effects of temperature and hold times on plasticity, and where the crack growth and instability laws have not been properly verified. So long as steps are taken to safeguard the situation against these uncertainties, possibly by means of verification tests where needed, it is more appropriate to use tearing under normal operating conditions than to limit the analysis to initiation.