Fatigue analysis of spot welds subjected to combined tension and shear loading

In this research, fatigue tests of spot welded coupons subjected to combined tension and shear forces were conducted using a Chrysler multiaxial spot weld test fixture. Fatigue cracks initiated at the edge of the weld nugget on the centerline of the faying surface for all specimens. The direction of the crack propagation depended on the combination of the applied axial loading direction, mean load, load ranges, and specimen stiffness. The fatigue life of a single resistance spot welded specimen increased as shear loading increased for all the tested specimens. The larger diameter specimens showed longer fatigue life for all loading angles. As the applied load range increased, the fatigue life decreased.; The test results were used to evaluate Swellam et al.'s, Sheppard's, and Rupp et al.'s approaches. All the approaches were reasonably accurate for multiaxial test data. However, the calculated fatigue lives by using Sheppard's and Rupp et al.'s approaches showed more scatter for Swellam's data because the approaches are sensitive to the variation of sheet thickness.; Therefore, a fatigue damage model was proposed based on the local structural stress at the spot welded joints. The forces and moments were calculated using linear elastic finite element analysis with coarse meshed models. The equivalent stress was calculated by Von Mises' equivalent stress equation. The proposed model using structural equivalent stress was well correlated to multiaxial and uniaxial test data. The results of this model also compared with those of Rupp et al.'s and Sheppard's models. The results of the structural equivalent stress model showed less scatter than those of Sheppard's and Rupp et al.'s models.; A numerical interpolation technique developed by Rule has also been applied to calculate the fatigue life of spot welded coupons. The technique works reasonably well in calculating fatigue life of specimens subjected to multiaxial or unaxial loads when some readily identifiable independent variables of spot welded coupons are taken into account. This technique does not require a detailed mechanics model of the local stress state or material constants for semi-empirical equations since it only depends on the test data.