The Theory of Critical Distances to estimate the static strength of notched samples of Al6082 loaded in combined tension and torsion. Part II: Multiaxial static assessment

This paper is concerned with a novel reformulation of the Theory of Critical Distances (TCD) suitable for estimating static strength of notched ductile materials subjected to multiaxial loading. The main feature of the method proposed and validated here is that the static assessment is performed by directly post-processing the linear-elastic stress fields in the vicinity of crack initiation sites. In more detail, our theory is proposed to be applied in the form of both the Point Method (PM) and the Line Method (LM) and it is formalised so that it can be used in conjunction with any classical equivalent stress (i.e., Von Mises’ equivalent stress, Tresca’s equivalent stress, maximum principal stress criterion, etc.). The accuracy and reliability of such an approach was checked by using the experimental results we generated by testing cylindrical samples, containing notches of different sharpness, made of Al6082 and loaded in combined tension and torsion. Observations of the failure modes in these specimens (reported in Part 1 of this two-part series of papers) informed the development of the above approach. Our method proved to be capable of estimates characterised by the usual level of accuracy shown by the TCD when used in other ambits of the structural integrity discipline, that is, of predictions falling within an error interval of about ±20%. This result is very encouraging especially in light of the fact that our simple linear-elastic approach was used to estimate static strength even when the material in the vicinity of crack initiation sites experienced multiaxial plastic deformations and the observed cracking behaviour was rather complex. Thanks to its particular features, the approach formalised in the present paper can be considered as an interesting engineering tool suitable for performing the static assessment of notched ductile materials by directly post-processing the outputs from simple linear-elastic Finite Element (FE) models.