ABCD Pittsburgh Chapter

The presentation focused on three different aspects of fatigue performance in steel bridge structures.

First, full-scale prototype tests of an orthotropic bridge deck was reviewed.  The full-scale laboratory testing of cantilevered sections of the prototype Williamsburgh Bridge orthotropic bridge deck demonstrated that the 1994 AASHTO LRFD fatigue provisions were incorrect in their classification of fatigue resistance.  In addition, the tests demonstrated that the fatigue resistance of the welded rib-to-diaphragm detail was dominated by in-plane bending stresses.  This resulted in better performance for combination groove-welded/fillet-welded connections than load-carrying fillet welds alone. Field testing of the final constructed south roadways of the Williamsburg Bridge indicated that the laboratory test accurately reflected the in-service behavior and boundary conditions.

Second, the ongoing experience with local distortion and out-of-plane bending was reviewed with a variety of examples.  Focus was on the most susessful methods of retrofitting fatigue-damaged details.

Finally, the results of recent tests to assess improvement methods were reviewed.  One study involved retrofitted Category E cover-plated bridge girders on the I-95 Yellow Mill Pond Bridge which was retrofitted by air hammer peening and gas-tungsten arc (GTA) remelting in 1976.  All retrofitted details had small detectable fatigue cracks at that time.  No further fatigue cracking was observed at those details after 20 years of service and an estimated 56 million truck passages at the time that beams were removed and laboratory tested. These tests demonstrated that the residual fatigue strength exceeded Category D.  Another study was a pilot study on HPS-485 W plate girders with as-welded and ultrasonic impact treated (UIT) details.  This technique offers a means to enhance low fatigue resistant details so that the fatigue limit state does not control the application of higher strength steels.