This report contains the findings from an extensive laboratory investigation of fatigue effects in welded steel beams subjected to variable-amplitude loadings similar to those that occur in actual bridges. This report is based on the results of an experimental program that included constant- and variable-amplitude fatigue tests of both small specimens and relatively large beams of various steels, with structural details similar to those tested in NCHRP Project 12-7. Extensive test results showed that variable amplitude random-sequence stress spectrums can be conveniently represented by a single constant amplitude effective stress range that would result in the same fatigue life as the variable amplitude stress range spectrum. A review of available field data on stresses in actual bridges showed that the passage of a truck across a bridge usually produces a single major stress cycle with small superimposed vibration stresses. The field data also showed that Raleigh probability-density curves can be used to approximate the frequency of occurence of major stress cycles in most highway-bridge stress spectrums. A possible new approach for bridge-design fatigue specifications is outlined in this report. This new approach is based on a fatigue-design truck placed in realistic positions to calculate a design stress range. To provide realistic fatigue specifications, the fatigue limit on the fatigue behavior at low stress ranges must be accurately known for various details. Effective methods of performing variable-amplitude random-sequence fatigue tests on large specimens were developed as part of the program and are described in the report. (Author/publisher)
Abstract