PSI - Issue 57

Jacques BERTHELLEMY et al. / Procedia Structural Integrity 57 (2024) 872–903 J. Berthellemy / Structural Integrity Procedia 00 (2023) 000 – 000

883

4. Fatigue evaluation of details using the “effective notch stress method” 4.1 Principes of the method

In order to evaluate details including fillet welds, the “effective notch stress method” is more adapted, as explained by A. Hobbacher, H.P. Lieurade et all. [9] . The way how to calculate the stresses using Code_Aster is the same as in the previous described method.

Figure 19: Fictitious rounding of weld toes and roots

However, the mesh changes because an artificial rounding with a radius of one millimetre is introduced at the hot spot. In this rounded area the mesh is much finer than in the case of the hot spot method. Experience shows that 8 segments on the arc of a quarter circle are needed to obtain the best results with the unstructured mesh of tetrahedrons well suited to Code_Aster. This results in tetrahedrons with sides of 0.0034 millimetres instead of 4 millimetres using the hot-spot method. It should be noted that the results vary very few with the precision of this mesh. However, the fineness of the mesh leads to heavier models: the required addressable memory and the computation time are greater than with the hot-spot method. 4.2 Probing calculation of a category of detail already known experimentally During this cycle presented in Fig. 5, a time varying factoris applied to the set of stresses that constitute the loading and for instance for the longitudinal stress, the amplitude of stress is of 50 MPa which means in other words that the  stress range is of 100 MPa. The quality of the fillet weld surface is assumed to be 125 MPa and Code_Aster presents directly as a result the probability of crack initiation.