Fatigue Behavior of spot welded high-strength sheet steels

The fatigue behavior of spot welds was examined in three high-strength steels - HSLA 50, dual-phase 590, and TRIP 590 - and a lower-strength DQSK steel. Two sample geometries chosen to represent different loading conditions on the spot weld were evaluated. The specimen configurations included tensile-shear and cross tension. Spot welding was performed using standard industrial practices, although a two-step procedure was used for the higher-alloy TRIP grade. In both the tensile-shear and cross-tension tests, low- and intermediate-cycle fatigue lives increased with increased base material strength. Fatigue performance was found to be independent of strength and microstructure at high-cycle fatigue lives. In the cross-tension samples, which subjected the weld to crack opening (i.e., Mode 1) loading, the TRIP material exhibited some interfacial fracture (brittle fracture along the weld centerline), and the presence of the interfacial fracture reduced the low-cycle (i.e., high-load) fatigue performance. At high cycles, the TRIP material performed equal to the other materials in both the cross-tension and tensile-shear samples. For all four materials, crack initiation occurred at a tongue structure at the sheet interface that results from molten material forced out from the weld nugget. The tongue structure was weakly bonded to the sheet surface, served as a precrack for fatigue, and nearly the entire fatigue life was spent in crack propagation. Crack propagation occurred through the heat-affected zone, with the crack path determined by stress distribution around the weld. No preferred microstructural path was evident. Fatigue data were analyzed using current fracture mechanics theories, and the results provided an accurate means of interpreting spot weld fatigue data between specimens of different geometries. Using stress intensity equations, spot weld fatigue data from other studies on steels with a wide range of strengths were also analyzed, and it was found that fatigue performance at high cycles was independent of base material strength and microstructure. It was confirmed that high-cycle fatigue performance is controlled by sample geometry alone, with no significant effect due to strength or microstructure.