No abstract

For the purpose of defining the initial crack size used in estimating the fatigue life of fillet welded joints, fatigue tests of fillet welded joints with a longitudinal attachment were carried out under constant and computer simulated highway variable amplitude loadings, and the initiation and early-stage propagation behavior of fatigue cracks at the fillet weld toe was investigated. Facet-like fracture surfaces were observed where fatigue cracks initiated and the depth of them was independent of stress range and welding material. Fatigue life analyses applying the fracture mechanics concept were performed on the assumption that such a fracture surface is the initial crack. Good agreement was obtained between the estimated fatigue life and the experimental one.

In order to examine the fatigue strengths of welded bridge details subjected to traffic loading, four large-scale welded girders were tested with various gusset details, transverse stiffener and cover-plates. The results were compared with the fatigue strengths of joint specimens and the allowable stresses in some specifications. Furthermore, the applicability of drilling holes as a retrofitting technique was investigated.

Fatigue behavior of fillet welded joints with a scallop under highway and railroad loading was studied. Variable amplitude stress fluctuations were generated by computer simulations of highway and railroad live loads. Fatigue failure occurred even if the equivalent stress range was less than the constant amplitude fatigue limit. Therefore, fatigue life evaluation assuming that stress ranges below the fatigue limit did no damage was unconservative. It was verified that the fatigue life under variable amplitude loading could be estimated accurately by applying the fracture mechanics approach with the linear damage rule.