ABSTRACT
This paper summarizes studies of a prestressed concrete bridge at both service and ultimate load levels. The objective is to acquire data for the calibration and development of assessment models. The ultimate limit state is of special interest, due to limited data is available as only few full-scale failure tests have been performed. The overall aim is to assess reinforced concrete bridges more accurately.
The bridge was a 121.5 m long five-span, prestressed concrete girder bridge, see Figure 1. The superstructure was 15.60 m wide with three parallel girders (height: 1923 mm, width: 410 mm) connected with a slab in the top. In the destructive tests two of the girders were strengthened using two systems of carbon fiber reinforcing polymer (CFRP): (1) one composed of near surface mounted (NSM) rods and (2) one using prestressed laminates. View of the Kiruna Bridge from south (2014-02-19). https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/fig307_1.jpg"/>
In the service load tests, the dynamic response was studied (Enochsson et al., 2011). The measured dynamic response in summer conditions were in good agreement to calculated dynamic properties of the bridge. Tests in winter conditions resulted in stiffer structural behavior, due to freezing effects on pavement, concrete and ground. This was not reflected in the FE analysis.
For investigation of the bridge at the ultimate-load level, an extensive experimental program was designed (Bagge et al., 2014). The focus was on the overall bridge behavior and the ultimate load-carrying capacity. Both the girders and the deck slab were studied. According to the tests and the FE analysis the two CFRP strengthening systems only had a limited influence on the structural behavior. The system (1) with NSM rods functioned well with no debonding failure. However, due to poor bond properties system (2) with prestressed laminates debonded at a low load level.
Predictions of the bridge behavior, using nonlinear FE analysis, showed good agreement with the girder failure test, see Figure 2. Existing differences can be explained by simplifications and lack of information about the test procedure, at time of performing the analyses. Greater consistency is expected using model updating based on the actual circumstances. Load-deflection curve of the bridge girder failure tests according to test and FE analysis. https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/fig307_2.tif"/>
The study confirms that even with a limited amount of information, a good predictions of the behavior could be obtained using this method.
