International Conference on Bridge Management, Safety and Management (IABMAS 2010)
July 11-15, 2010
Dubbs, N., F.L. Moon, A.E. Aktan
The Burlington-Bristol Bridge spans 3,000 ft across the Delaware River and has a main lift span composed of a 540 ft through-truss. In 2007, this bridge was load rated using both analytical methods and traditional truck load tests, and the results were found to be inconclusive. Following the load tests, there were several models that matched the live load strains equally well and predicted significantly different rating factors. The key uncertainty was related to the distribution of dead load actions, which represented over 95% of the demand on the critical main span members and was completely unobservable during tradition truck load tests. In an attempt to estimate dead load demand, a three phased approach was developed and carried out. First, the critical top chord member was instrumented and monitored with a series of both high speed and vibrating wire strain gages to capture the variation of dead load strain due to temperature, radiation, and other seasonal effects. The second phase included detailed ambient vibration monitoring study and a subsequent physics-based structural identification to estimate total mass, mass distribution and boundary/continuity conditions. The third and final phase included the use of a portable x-ray diffraction technology to estimate the intrinsic stress at various locations around the critical upper chord member. This paper and will present the details of this study and discuss the proposal of how uncertainty can be reduced using this technology while the presentation will discuss the completion of these techniques on the Burlington-Bristol Bridge and the resulting effect on the load capacity estimation.