Mixed-mode fatigue from stress concentrations: an approach based on equivalent stress intensity

Abstract

This paper describes the analysis of fatigue failure in an automotive crankshaft as a result of loading under test conditions in bending and torsion. Failures occurred from stress concentrations whose location varied with loading type. In order to describe the severity of the stress concentration, an equivalent K method was used which is known as `crack modelling'. This method uses a modelling approach to find effective stress-intensity ( K) values from static-load finite element (FE) analysis. Subsequently the cyclic, Δ K, value can be found; failure is assumed to occur if this is greater than the threshold, Δ K th, for the material. Previously used for simple tensile and bending stresses (Taylor D, Lawless S. Prediction of fatigue behaviour in stress concentrators of arbitrary geometry. Engng Fract Mech 1996;53:929–39; Taylor D. Crack modelling: a technique for the fatigue design of components. Engng Failure Analysis 1996;3:129–36; Taylor D, Ciepalowicz AJ, Rogers P, Devlukia J. Prediction of fatigue failure in a crankshaft using the technique of crack modelling. Fatigue Fract Engng Mater Struct 1997;20:13–21; Taylor D. Crack modelling: a novel technique for the prediction of fatigue failure in the presence of stress concentrations. Computational Mechanics 1997;20:176–80), the method was extended in this study to consider other loading modes. Initial analysis for the bending case, using a rather coarse FE mesh, gave a predicted fatigue limit within 2% of the experimental value. A more refined mesh, which gave a much higher `hot-spot' stress, had no significant effect on the accuracy of the prediction. The same methodology applied to torsional loading also gave an accurate prediction (within 3%). It is therefore predicted that the method will work for any in-phase combination of bending and torsion.