The overall objective of this research is to develop a general model of fatigue crack propagation in resistance spot-welded joints. An important feature of this development is that the model and accompanying methodology should be accessible to designers evaluating fatigue response of structures containing multiple welds. This objective is achieved by examining the stress state around a resistance spot weld. A general expression for the structural stress around the weld is formulated that is dependent only on the loading immediately surrounding the weld; as such, it is specimen independent.
In previous work  it was found that structural stress could be successfully used to estimate life for crack initiation and growth to a length of 0.01 in. (0.254 mm) in resistance spot welds, and that this period represents less than 30% of the total life in as-welded joints. It is important to note that estimation of this period is highly dependent on fatigue-related material properties. In Ref 2 it was shown that structural stress can be related to crack propagation life through linear elastic fracture mechanics. Using the resulting relationship between structural stress and life, life estimations were made for a variety of HSLA steel specimens (e.g., tensile-shear, coach peel) in a number of conditions (e.g., as-welded, prestressed) and were compared with experimentally measured lives. Life estimations were within a factor of 3 of measured lives.
In the current work, the methodology developed in Ref 2 for estimating propagation life is extended by explicitly considering the effects of axial loading and of load ratio on life estimates. In addition, issues related to creating an appropriate finite element model are addressed, and a number of modeling guidelines are established. Total life estimations are made and compare favorably with experimentally measured lives of welds in mild steels in a variety of specimen configurations.