PSI - Issue 19

Vanlemmens, Elbel, Meneghetti/ Structural Integrity Procedia 00 (2019) 000 – 000

6

Théophane Vanlemmens et al. / Procedia Structural Integrity 19 (2019) 610–616

615

Fig. 8: 3D models from both samples

The results of the test showed two different crack initiation: - For samples 1, the cracks initiated at the end of the root of the weld, in the direction of the base sheet for the 30 samples tested. - For samples 2, the crack initiated at the same point but in the direction of the welded sheet for the 23 samples tested. The crack initiation location correlates with the test as shown in Fig. 9. The red arrows point the maximal damage location in the simulation. The direction of the crack is well predicted as well in the simulation: for sample 1 the maximal damage is on the horizontal edge of the weld end, which would lead to a crack in the base sheet, for sample 2, the maximal damage is located on the vertical edge. In order to have an idea about the effect of the idealized geometry on the result, three different angles at the weld end have been tested, 90°, 120° and 135°.

Test 1 Test 2 Fig. 9: Crack initiation in the test and in the simulation (illustration with the 90° model)

The results in terms of absolute number of cycles are shown in Table 1. The results from the three different angles vary within a factor 2.4 for samples 1 and within a factor 4 for samples 2. The geometry of the samples has been 3D scanned and the mean angles have been measured. The result shows that the angle is always higher than 90°, the mean value is around 110° and for some of the samples, the angles goes up to 135° and even more. The comparison between the number of cycles with PSM at an angle of 120° and those from the tests, shows that the PSM is pessimistic (between a factor3.1 for samples 1 and 2.1 for samples 2). This result can be compared to the comparison done with R1 concept at §3.3, where the PSM has been shown to be pessimistic as well.