Seismic Resistance of Square Concrete Columns Retrofitted with Glass Fiber-Reinforced Polymer

ACI Structural Journal, Sep/Oct 2005 by Memon, Muhammad S, Sheikh, Shamim A

The specimen was aligned in the vertical plane using engineering levels. In the horizontal plane, plumb-bobs were used to match the centerline of the specimen with the line of action of the axial load. After the external instrumentation was installed, the specimen was loaded up to 50% of the predetermined axial load in 200 kN increments. The deformations at the four corners of the specimen were recorded using four LVDTs. If the difference between the average reading and the maximum or minimum displacement reading was more than 5%, the specimen was unloaded and necessary adjustments were made. The process was repeated until the specimen was properly aligned. The strain gauge readings were also used to confirm the alignment.

The test started with the application of pre-assigned axial load followed by the application of lateral displacement sequence (Fig. 5). The specimen was subjected to 75% of the yield or elastic displacement (Δ^sub 1^) in the first cycle; the displacement Δ^sub 1^ is also defined in Fig. 5. The displacement Δ^sub 1^ was determined from the theoretical section behavior of the column and integrating curvatures along its length. In the subsequent cycles, the lateral excursions were increased gradually (two cycles each to 2Δ^sub 1^, 3Δ^sub 1^, 4Δ^sub 1^ ....) until the specimen was unable to maintain the applied axial load resulting in the termination of the test.

Each of the two specimens chosen to undergo repair was subjected to axial load and lateral displacement excursions until, at minimum, the top and bottom concrete cover spalled off and the yielding of the longitudinal bars initiated. For Specimen ASGR-7NSS subjected to an axial load of 1490 kN, spalling of the top and bottom concrete cover occurred in the sixth cycle. The maximum displacement in the critical region was approximately 2.55Δ^sub 1^ and the average maximum compressive strain in the longitudinal bars was 0.00258. The amount of damage causing spalling of the concrete cover and yielding of the longitudinal steel was considered appropriate for most columns that are candidates for repair. The specimen was then returned to zero lateral displacement position for repair. For the purpose of safety during the repair process, the axial load was reduced to 66% of the original axial load value. The loose concrete was carefully removed from the damaged column as the repair took place under the applied axial load to simulate the field conditions. Nonshrink grout was used to repair the damaged specimen. After 3 days, when the grout had cured and dried, the specimen was wrapped with GFRP while it remained in the test frame. After the curing of GFRP sheets, the specimen was tested to failure with an axial load of 1490 kN.

The top and bottom concrete cover spalled off and the yielding of longitudinal bars initiated during the third cycle for Specimen ASGR-8NSS while maintaining an axial load of 2500 kN. The average maximum compressive strain in the longitudinal bars was 0.00251. With this damage, the repair of the specimen was carried out under an axial load of approximately 60% of its original value. The same procedure of repair, wrapping, and retesting was adopted as used for Specimen ASGR-7NSS.