Issue 37

C. Riess et alii, Frattura ed Integrità Strutturale, 37 (2016) 52-59; DOI: 10.3221/IGF-ESIS37.08 57 the weakening, the critical locations are shifted towards the middle of the component. Hence, effects due to the load introduction and the bolted connections are minimized. R ESULTS esults of the optimization can be found in Tab. 1. Beside the solution parameters opt opt /   the value of the objective function np opt f , is also listed. In the case without weakening of the structure, the results for the different versions are quite similar. Independent of the damage parameter, the same critical node is identified. It is not possible to attain a NPF higher than 0.16 at the potential crack initiation site. damage parameter opt  opt  np opt f , original structure f 3.113 1.516 0.160 a cp ,  3.139 1.441 0.153 a v ,  0.0 1.536 0.159 weakened structure f 0.983 0.904 0.812 a cp ,  0.882 0.688 0.831 a v ,  0.946 0.776 0.819 Table 1 : Results of the optimization. The selective weakening results in the desired effect. High non-proportionalities ( np f 0 8 .  ) are obtained for all versions. Similar critical nodes are identified using the signed von Mises approach and the a cp ,  approach. A slightly different solution is found by the Findley parameter approach. In this solution the critical node crit x is located some nodes away in contrast to both other versions. D ISCUSSION here is a high non-proportionality in many regions of the component. However, stresses are often small in these regions and failure will not occur there. On the other hand, the non-proportionality is negligible at many crack initiation sites. Independent of the external loads the stresses are highly oriented at the crack initiation site in those cases. Typical examples are ribs of housings made by die-cast or a cross-hole in a shaft. These are locations, where the local stress states are nearly proportional. In order to get a high non-proportionality the following two conditions have to be fulfilled: The unit load cases have to create stress states with different planes of maximum shear (different principal axis) and the forces have to be in an appropriate ratio. A systematic planning of component tests with high non-proportionality is only feasible using an automated optimization process. Figure 4 : ߪ ௔,௩ for the result of the optimization (left) and the tradeoff (right). R T