Field instrumentation and measured response of the I-295 cable-stayed bridge.

27784070Final report.PDFTech Reporthttp://ntl.bts.gov/lib/36000/36800/36869/93-R12.pdfVTRC 93-R12VTRC 93-R13VirginiaVirginia Transportation Research CouncilDuemmel, Paul SVirginia Transportation Research CouncilUS Transportation CollectionThis first report describes the results of a field study of the live load responses of a segmentally constructed prestressed concrete cable-stayed bridge. The main span of the test structure consists of twin box girders connected by delta frames. Known vehicular loadings were placed statically at various points along the bridge. Strains measured during this loading were compared with those obtained from a finite element model of the bridge. Strain trends predicted by the finite element model were in good agreement with the measured strain trends. Quantitative agreement was fair, at least in part because of the high stiffness of the bridge and the limitations on the magnitude of load that could be applied. The second report describes the results of a field study of the thermal responses of a cable-stayed bridge. Data were gathered from the I-295 James River Bridge, a precast segmental concrete bridge with a cable-stayed main span consisting of twin box girders connected by delta frames. The thermal gradient and associated thermal strains in the box girders and pylons were measured using an extensive array of thermocouples and strain-gaged reinforcing bars installed at selected locations in the main-span box girder and south pylon. The temperature and strain response data were compared with that predicted from detailed finite element models of the structure using both frame and plate elements. Comparison revealed a complex three-dimensional strain pattern dependent on the wind direction and the angle of solar incidence. Simplified beam element models were unable to predict many of the observed local variations in thermal strain, which are influenced by wind direction, solar heating direction, proximity to the web, and the existence of parapets monolithic with the deck. Three-dimensional finite element models appear to be more capable of predicting the kind of three-dimensional strains observed, but quantitative agreement with the observed thermal strains was limited at best