Because of increasing fuel costs and environmental concerns, the automotive industry is under enormous pressure to reduce vehicle weight. One strategy, downgaging, substitutes a reduced gage (thickness) steel in place of a thicker one, and is usually accompanied by a material grade change to a higher strength steel. Thus, Advanced High Strength Steels (AHSS) are increasingly used for lightweight automotive body structures. The critical durability concern with steels is the spot welds used to join them, since fatigue cracks in body structures preferentially initiate at spot welds. Hence, the Auto/Steel Partnership (A/SP) Sheet Steel Fatigue Taskforce undertook an investigation both to study the fatigue performance of AHSS spot welds, and to generate data for OEM durability analysis. The study included seven AHSS grades and, for comparison, mild steels and a conventional High Strength Low Alloy grade, HSLA340. In addition, smaller self-contained investigations were conducted: adhesive bonds and weld-bonds of DP600 and HSLA340, the effects of processing parameters (such as button size, hold time, and a paint bake cycle); spectrum (or variable amplitude) loading vs. constant amplitude loading; and mean stress effects.
In all, over 500 fatigue specimens were tested. As a result of this testing, it was concluded that fatigue performance of spot welds is not sensitive to the strength of the parent steel, nor is it affected greatly by mean stress. Neither a typical automotive paint bake cycle nor changes in weld hold time have very much effect on fatigue performance. Fatigue strength is largely related to the geometric factors (steel thickness, button size, and coupon width) which affect the stress conditions around the welds. Adhesively bonded and weld-bonded joints always exhibited much higher fatigue strength than spot welds alone, in line with their increased joint area. This performance benefit diminished with increasing load amplitude.
Lastly, using the fatigue data generated in this study, a comprehensive analytical study was conducted to evaluate the performance of five commonly used spot weld fatigue damage prediction parameters - Rupp and co-workers, Sheppard, Dong, Kang, and Swellam. The resulting correlations indicated that, with the possible exception of Swellam's parameter (which had slightly greater scatter), all worked satisfactorily and provided equivalent performance when used on this dataset and with these specimen configurations.