PSI - Issue 28

Michael Jones et al. / Procedia Structural Integrity 28 (2020) 2078–2085 Author name / Structural Integrity Procedia 00 (2019) 000–000

2084

7

triaxiality value implies that the ratio of mean to equivalent stress is decreasing, or specifically in this case of a tensile test, that the equivalent stress is increasing faster than the mean stress is. Figure 7 shows the stress components of interest in the region of maximum triaxiality in the small notch sample. This corresponds to ‘FEA fourth’ in Figure 6(c). This actually shows that the mean stress increases faster than the equivalent stress and is the most extreme case where the triaxiality increases slightly throughout the test. The evolution with time will be similar in the other tests. Ultimately all of the stress components continue to increase at similar rates, and as such the ratio of mean to equivalent stress does not vary significantly. Subsequently the triaxiality stays near constant, even as the notch radius decreases and the radius of curvature increases.

0 100 200 300 400 500 600 700 800

Equivalent Mean Radial

Axial Hoop

Stress (MPa)

0,0

0,5

1,0

Normalised load

Figure 7 – Stress components in the region of maximum triaxiality in the small notch sample.

Although quantitively there is little agreement, some qualitative similarities can be drawn. The Bridgman theory predicts that a larger notch radius of curvature leads to lower values of triaxiality. There may be little variation throughout the test, but the triaxiality values are generally lower in the FEA models for the blunt notch, with the small notch having the highest triaxiality values and the medium notch sitting in between the two. 6. Conclusion The analysis undertaken categorically shows that Bridgman expressions for strain and triaxiality are not valid for use in the testing of notch bar specimens. The triaxiality values in the FEA models appeared to be independent of the notch radius and radius of curvature, and the strain distribution across the radius is not uniform. The disparities are likely due to the difference in material flow and stress distribution between a post necked geometry from an initially smooth uniaxial test and that of a notched specimen. Determination of a ductile failure locus is of great importance when characterising the behaviour of a material. It is therefore recommended that finite element analysis is performed for accuracy when testing notch bars. Acknowledgements The authors thank EDF Energy and EPSRC (project reference 2067508) for providing the funding to conduct the research for this publication. This paper is published with permission from EDF Energy.