PSI - Issue 19

Théophane Vanlemmens et al. / Procedia Structural Integrity 19 (2019) 610–616 Vanlemmens, Elbel, Meneghetti/ Structural Integrity Procedia 00 (2019) 000 – 000

3

612

Fig. 2: Compared modelling of the weld for PSM and R1 concept

Fig. 3: Compared mesh of the weld for PSM and R1 concept

3. Results of the tests

3.1. Description of the test

The test was done with ANSYS software (V19.1) using two formulations suggested by Meneghetti. First hexahedral elements have been used. The results presented in §3.2 are computed with this formulation. In a second step, tetrahedral elements have been tested and proved to be more efficient to implement on complex geometries. The results presented in §3.3 were computed with this second formulation. The post treatment consisting in collecting the stress components in the right local coordinate systems has been done manually. The lives are computed using FAT225 for the R1 concept and using FAT156 for the PSM ([2]), which are both the recommended FAT classes for these methods with a probability of failure from 2.5%. The geometries selected for this purpose are simple probes where fatigue results are available, some of them have been already computed by Meneghetti and documented in one of the papers. For those geometries, the results could be retrieved. For all these geometries, the results were compared to the results from the R1 concept. The mean value of the quotient between the life PSM and the life R1 concept is 0.64 for the weld toes and 0.76 for the weld roots. This shows that the PSM is a little bit more pessimistic than the R1 concept on those examples. 3.2. Tests on simple geometries