PSI - Issue 54

Jenny Köckritz et al. / Procedia Structural Integrity 54 (2024) 423–430 J. Köckritz/ Structural Integrity Procedia 00 (2019) 000 – 000

427

5

The fatigue material data for the SWF in the form of S-N curves for membrane stress of welded EN AW 7020 are derived from Mensinger (2018). Values of both butt joint (SWF Mat1) and cross joint (SWF Mat2) are applied and compared. The bending stress S-N curve was derived from the pure bending tests. The fatigue material data for the SWF model are listed in Table 2, where SR1 corresponds to the fatigue coefficient and 1 =−1/ is the fatigue exponent according to the equation of Basquin with the stress amplitude and cycles to failure N in (1). = 1 ∙ 1 (1) The material data are implemented as a two segmented S-N curve, in which the fatigue exponent for the second segment is set as −0.015 and the number of cycles at the transition point is 10 7 for both bending and membrane S-N curve. Damage is not considered below the fatigue limit of 70 . Table 2. Fatigue material data for the seam weld fatigue method Parameter Bending SN curve Membrane SWF Mat1 (Mensinger, 2018) Membrane SWF Mat2 (Mensinger, 2018) SR1 1987 1670 1740 b 1 -0.167 -0.17 -0.199 A parameter analysis was conducted for the SWF analysis with load case 1 to show the effect of weld thickness t , bending ratio r and the material coefficients SR1 and 1 . In the presented investigations, the default bending ratio of =0.5 (Altair, 2023) was used. A variation of r between 0.1 and 0.9 in intervals of 0.1 as presented in Fig. 3 shows a substantial influence on the assessed life. The green highlighted lines represent the original value, no changes occur above a upper limit of r . The weld thickness varies greatly in reality. However, a variation in the modelled thickness from = 3…7 in intervals of 1 mm shows almost no influence on the assessed fatigue life. In contrast, a variation in the material data for bending and membrane shows a higher influence on fatigue life, see Fig. 4. The influence of SR1 and 1 was studied separately for membrane and bending material data, with the other data set remaining unchanged. In general, changes in the membrane material data have a higher influence than changes in the bending material data. However, the difference in the influence depends on the bending ratio r . Within this scope, variations of the fatigue exponent 1 show the highest influence on the assessed life of all varied parameters. Load case 1 Load case 2

Bending ratio variation

Fig. 3. Parameter variation of bending ratio and influence on assessed fatigue life with the SWF method

The cargo bicycle frame is modelled with shell elements on the midplane of the tubes. While freeze contacts or rigid body connectors are feasible for static load case assessment, they prove to be too inaccurate for fatigue assessment. The welds are modelled with quad shell elements at the tube edge for the SWF method. One critical weld joint was modelled with all presented modelling methods to compare them regarding their applicability. There, a solid body frame section with the solid weld and the effective notch radius was used for EN method. The HSS method only required a remeshing to provide the extrapolation points. The load cases as described in chapter 2.1 are applied separately with a constant amplitude. The measured service life forces were introduced at the driver contact points with their individual load time histories. Evaluation of the expected life was performed for all welds for each load case.