Document Type



Master of Science


Structural Engineering

First Adviser

Roy, Sougata


After a comprehensive study of the site specific loading, it was decided to replace the 45 year old distressed concrete filled steel grid deck at the upper level of the signature suspension bridge with a steel orthotropic deck that is integral with the floor system and the stiffening truss. In view of the high volume of truck traffic on the deck, and the concerns of increased possibility of fatigue cracking from a large number of welded connections in the orthotropic deck, the design and fabrication of the replacement orthotropic deck details needed to be verified for infinite fatigue life, i.e., 75 years' service life without any fatigue cracking under site specific loading.A full size prototype of a quarter of the deck between the panel points of the stiffening truss, and including one floor beam and two stringers was fatigue tested in at the ATLSS Engineering Research Center, Lehigh University. This study, identified as Phase 1 and reported by Alapati (2012), demonstrated that the proposed replacement deck design did not meet the fatigue design requirements of 75 years' service life. The prototype exhibited premature fatigue cracking of the intermediate subfloor beam-to-rib weld at the cutout termination, and the internal bulkhead plate-to-rib welds at the intermediate subfloor beam. Based on these findings, it was recommended to reduce the stresses at the critical details below their respective constant amplitude fatigue thresholds (CAFT) for achieving a 75 years' service life under site specific loading. Accordingly, deck design was revised where the cutout for the ribs at the intermediate subfloor beams was enlarged at the rib termination, and the thickness of the subfloor beam web was increased from 5/8 in. (16 mm) to 7/8 in. (22 mm). In addition, the number of subfloor beams between the floor beams was also increased from four to five, which reduced the spacing between the subfloor beams from 12 ft. 41/2 in. (1.257 m) to 8 ft. 3 in (2.515 m). The thickness of the full depth bulkhead plates inside the ribs was increased from 5/16 in. (8 mm) to 5/8 in. (16 mm). The welded connection between the bulk-head plate and the rib over the top 7 in. (175 mm) critically stressed region was fabricated as a complete joint penetration weld, which ensured adequate penetration at the weld root and increased the fatigue resistance of this cruciform detail.The fatigue performance of the refurbished deck was evaluated by full size prototype testing of the modified Phase 1 specimen, where the intermediate floor beam was replaced to incorporate the improved design. Similar to the Phase 1 testing, the test setup simulated the global boundary conditions including the supporting floor framing. The deck was fatigue tested using four stationary overhead hydraulic actuators positioned centrally in the transverse direction. In the longitudinal directions the actuators were positioned in concentric and eccentric configuration about the intermediate subfloor beam and were loaded in paired sequence simulating the passage of the tandem axle of an AASHTO fatigue truck across the intermediate subfloor beam. Three additional under deck actuators were used to simulate continuity boundary conditions. The deck was tested for 5 million cycles at a load level of 3.45×HS15 (=3×0.75×HS20 or the HL-93 fatigue truck) including impact, which was considered equivalent to 75 years' service life under site specific loading. The deck was instrumented to evaluate its response, with majority of the instrumentation concentrated around the critical welded details in the intermediate subfloor beam.No fatigue cracking was found in the deck upon completion of the testing. The Phase 2 study verified that the refurbished deck would provide a 75 years' service life under site specific loading. With the design improvements, the stresses at the fatigue critical details reduced significantly compared to the original deck tested in Phase 1. The measured stress ranges at all details except at the cutout edge were at or below the constant amplitude fatigue threshold (CAFT) of their respective detail categories. 3D Finite Element analyses of the deck were performed to understand the complex behavior of the deck under localized and moving wheel loads, and to verify the measured response under static and dynamic loading conditions. Multi-level submodel analyses were performed to assess the fatigue performance of the critical connection details, and to assess the fatigue strength of the rib-to-subfloor beam and rib-to-bulkhead plate weld connection at the cutout termination using local stress based approaches. The hot spot stress obtained at the rib-to-subfloor beam and the rib-to-bulkhead plate weld toes were compared against the FAT 100 curve as per IIW recommendations. The notch stress approach used by Roy and Fisher (2005) for assessing constant amplitude fatigue threshold (CAFT) of welded connections was also implemented to predict the fatigue performance of the rib-to-subfloor beam connection detail.