Evolution of Fatigue-Resistant Steel Bridges

Fatigue cracking was seldom found in welded highway and railroad bridges from the time of their introduction in the 1950s until the late 1960s. The fatigue design specifications used in that era were developed from a limited knowledge base and largely with small-scale specimens that simulated welded details. During the AASHO Road Test in 1960 fatigue cracks were observed to develop in cover-plated steel bridge beams as a result of the heavy loads and high stress ranges. This observation subsequently resulted in a series of experimental studies supported by NCHRP starting in 1967. The laboratory studies with full-scale details were designed to evaluate the significance of many factors thought to influence fatigue resistance, including loading history (and associated stress states including residual stresses), type of steel, design details, and quality of fabrication. These studies indicated that small-scale specimens overestimated fatigue resistance and that only the stress range for a given detail was critical. As a result fatigue resistance design provisions in use since the 1950s were inadequate and overly optimistic, particularly at longer lives, because the assumption of a fatigue limit of 2 million cycles proved to be incorrect. The results of laboratory studies with full-size specimens and their impact on changing the concept of fatigue design and the bridge fatigue design provisions used for highway and railroad bridges today are reviewed. During the 1970s and 1980s fatigue cracking associated with low-fatigue-strength details (Categories E and E′), such as cover plates and lateral gusset plates, increased. Cracks were also found in transverse groove welds, particularly in attachments such as longitudinal stiffeners, gusset plates and even flange splices. These groove weld cracks generally occurred because large defects were inadvertently fabricated into the welded joint. The occurrence of these cracks was found to be predictable and in agreement with the laboratory fatigue resistance results. The 1970s also exposed an unexpected source of cracking due to the distortion of small web gaps that were frequently used in welded bridge structures. Web gap cracking continues to develop in a wide range of bridge types. It is the source of most fatigue cracks in steel bridges. Existing bridges that are susceptible to fatigue cracks or that develop fatigue cracks at primary details or from web gap distortion are easily repaired or retrofitted to ensure long-term performance. Examples of such repairs are reviewed. The future is bright for welded bridges because the knowledge base and current design provisions make it possible to design and build fatigue-resistant bridges.