Seismic Retrofit of Bridge Joints in Central U.S. with Carbon Fiber-Reinforced Polymer Composites

ACI Structural Journal, Mar/Apr 2007 by Silva, Pedro F, Ereckson, Nicholas J, Chen, Genda D

RESEARCH SIGNIFICANCE

A fundamental basis of capacity design relies on carefully selecting and detailing regions to develop inelastic actions, while all other regions are designed to remain essentially elastic under seismic loads. Research derived from this program proposes a CFRP retrofit scheme that can be used to mitigate design deficiencies that do not meet current capacity design requirements. Furthermore, a simple yet efficient analytical model is proposed to account for the flexibility of unreinforced joints retrofitted with CFRP composites.

EXPERIMENTAL PROGRAM

Two 4/5-scale units, designated as Units 1 and 2, were constructed and tested under simulated seismic loads. Design of the test units was based on a prototype structure, which was selected from 13 bridges built within the NMSZ. These bridges were evaluated within the RC beam-column subassemblies using well-established bridge seismic analysis/ assessment tools. The prototype bridge bent was selected from one of the bents that was constructed with inadequate capacity design considerations within the beam-column subassembly and is depicted in Fig. 1(a) with the corresponding setup shown in Fig. 1(b).

Test setup

The overall test setup is depicted in Fig. 1, which is very similar to the setup used in the research program by Naito et al. (2002). The columns were 0.610 m (24 in.) in diameter with a clear height from the bent cap interface to the height of load application of 2.02 m (79.5 in.), leading to an aspect ratio of nearly 3.3. The bent caps were 0.74 mm (21.1 in.) wide by 0.88 m (34.6 in.) deep by 5.18 m (17 ft) in length with a clear span between supports of 4.57 m (15 ft), or an aspect ratio to the center of the bent cap of 2.6. Both the column and the bent cap lengths corresponded approximately to the distance from the centerline of the members to the seismic moment inflection point obtained from the prototype bridge bent (Chen et al. 2005).

As shown in Fig. 1, on top of the load stub, a hydraulic jack was used to apply the simulated gravity load. This axial load was then transferred to the bent cap through I-sections, which were positioned on the underside of the bent cap at a distance of 1.83 m (6.0 ft) to simulate the position of the bridge girders in the prototype. The total axial load applied on the column was 710 kN (160 kips), which corresponds to an estimated axial load ratio of 8%.

Loading protocol

The loading protocol for the two test units is presented in Fig. 2, where it is shown that Unit 1 was tested in three phases. Unit 2 was strengthened according to nearly the same retrofit scheme as Unit 1, but strengthening was fully completed prior to testing and a few modifications were implemented to improve its seismic performance.

Phase I-Unit 1 was initially tested in its virgin condition up to onset of the first failure mode, which was characterized by shear failure of the column. This failure mode was corroborated by preliminary investigations, which indicate that shear failure was likely to occur between the displacement ductility levels of 2 and 3. Testing was discontinued and the column was strengthened for shear and confinement using CFRP sheets.