Document Type



Master of Science


Civil Engineering

First Adviser

Roy, Sougata


A steel orthotropic deck integrated with steel box girders is proposed for a replacement vertical lift bridge. In contrast to the orthotropic bridge decks implemented in the United States over the past two decades with ribs passing through matching cutouts in floor beams with extended cutout under the rib soffit, the subject deck incorporates a 3/4 in. (19 mm) deck plate stiffened by 5/16 in. (8 mm) thick rounded bottom ribs (U-shaped) passing continuously through matching cutouts in the floor beams without any extended cut-out under the rib soffit. Perceived to be cost-effective, the fitted rib-to-floor beam connection is proposed to be fillet welded, however, performance of this connection requires careful fit-up, which can incur additional fabrication costs. In addition, the lift bridge serves as a key element of a portway corridor carrying the main trucking route and hence the expected high volume of Average Daily Truck Traffic (ADTT) on the lift bridge was a concern. Accordingly, the connection details of the subject deck was evaluated for infinite fatigue life by testing a full-scale prototype of the part-bridge deck at the ATLSS Engineering Research Center of Lehigh University. This thesis pertains to the fatigue performance of the proposed rib-to-floor beam connection design.Multi-level 3D linear elastic finite element analyses (FEA) of the proposed bridge deck were performed that identified the rib-to-floor beam connection adjacent to a box girder as the most critically stressed region of the deck, when the rear tandem axle of the AASHTO fatigue design truck was symmetric with the floor beam and the rib was located in the shear span of the floor beam. Based on the analyses, a full-scale prototype of the part bridge deck comprising 5 ribs and 3 floor beams, and a test setup that would adequately replicate the boundary conditions were decided for assessing the in-service fatigue performance of the proposed deck by testing in the laboratory under simulated conditions. The prototype deck was fabricated in two panels, which were spliced (transverse to the ribs) in the laboratory by a complete joint penetration (CJP) weld at the deck plate and bolted splices at the ribs and the girder, simulating the field splice in the actual bridge construction. During assembly of the deck panels in the laboratory, significant lack of fit was noted between the panels due to the distortion from welding heat effects on asymmetric cross section of the specimen. The CJP deck plate splice was performed using a brass backing bar to facilitate removal of after the welding. Use of a brass backing bar, however, resulted in significant lack of fusion (LOF) at the weld root, which could not be effectively repaired due to the access restrictions.The fatigue testing was performed using a pair of above-deck hydraulic actuators, which were attached to spreader beams and loading pads, simulating the rear tandem axles of the AASHTO fatigue truck for orthotropic deck design and the tire contact with the deck plate. The deck was loaded as per the Fatigue I limit state load of the AASHTO LRFD Bridge Design specifications (BDS) 6th edition to verify infinite life performance, which resulted in a total load range of 82.8 kip (368.4 kN) or 41.4 kip (184.2 kN) per axle (or actuator). In addition, an under-deck actuator provided at the inner floor beam was cycled for a displacement range of 0.1 in. (2.5 mm), synchronous with the above-deck actuators to simulate the global displacement boundary condition. The deck was extensively instrumented with strain gauges and displacement transducers to evaluate the response of the deck and the various connection details, with majority of the instrumentation concentrated at the critical connection between the floor beam at the midspan and the rib adjacent to the girder. The deck response was also determined under a rolling tandem axle bogie load moved across the deck at a slow rate.The fatigue testing was run-out at 8 million cycles without any detectable fatigue cracking in the deck. The measured stress ranges at all critical connections were less than the CAFT of their respective detail categories. The test results would indicate infinite life performance of the rib-to-floor beam connection detail of the proposed deck design, as long as the site specific overloads do not exceed the AASHTO Fatigue I limit state load more than 1 in 10000 occurrences. The test results also demonstrated that the large deviation from the specified fabrication tolerances and the rejected welding procedures, which were noted during the specimen fabrication and installation, did not affect the fatigue resistance of the connection details. This suggests that the specified fabrication tolerances are arbitrary. In addition, the study provided critical information on issues related to fabrication and installation of the proposed orthotropic deck design for the lift bridge.