Seismic Retrofit of Lap Splices in Nonductile Square Columns Using Carbon Fiber-Reinforced Jackets

ACI Structural Journal, Nov/Dec 2006 by Harries, Kent A, Ricles, James R, Pessiki, Stephen, Sause, Richard

1. Increase τ^sub max^ such that the full tensile capacity of the lapped bars may be achieved;

2. Allow for the possibility of slip along the lap splice and increase the slip s^sub 2^, at which deterioration of the bond initiates, effectively increasing the Region 2 plateau; or

3. Increase both τ^sub max^ and s^sub 2^ to improve the behavior of the splice without eliminating the possibility of slip.

The provision of additional confinement in the form of an external FRP jacket may improve the Region 1 response, increasing τ^sub max^, as indicated in Table 1. This increase, however, has an upper limit defined by a pullout failure. In addition, the confining pressure required to develop the greatest value of τ^sub max^ (refer to Table 1) is quite high (7.5 MPa = 1088 psi) and is difficult to generate while limiting lateral strains. The requirement that lateral strains are controlled is implied by the defined behavior of Region 3. Once lateral strains are large enough to permit splitting cracks to develop, the bond stress-slip relationship begins to deteriorate. Therefore, although τ^sub max^ may be moderately increased, the primary objective of retrofitting a deficient tension lap splice is to maximize the value of lap splice slip at the onset of longitudinal splitting s^sub 2^, and thus the length of the Region 2 plateau.

It is likely that after splitting cracks begin to form, a retrofit jacket may serve to control the splitting, thus increasing the value of the residual bond stress τ^sub f^ (refer to Fig. 1 and Table 1) and reducing the slope of Region 3. This behavior is implied by the various design and behavior models already cited. In addition, this behavior has been noted in previous jacket retrofit studies of columns having initially poor lap splice configurations (Ballinger et al. 1992; Jin et al. 1994; Priestley et al. 1992).

A successful retrofit therefore must provide sufficient confinement to control lateral strains as well as mitigate the significant deterioration of bond due to cycling. The retrofit will postpone the onset of splitting and possibly reduce the severity of the subsequent deterioration. The final residual bond stress τ^sub f^ is therefore likely improved by the presence of a jacket. With a jacket, spalling is controlled and concrete integrity is maintained, enhancing residual bond capacity and safeguarding against significant loss of axial load carrying capacity. Regardless of retrofit strategy, however, the bond stress-slip relationship is ultimately governed by the slip corresponding to lug spacing.

COLUMN SPECIMENS

The test specimens used in this study were based on columns included in prototype nonductile reinforced concrete buildings designed by Kurama et al. (1996). The cross section dimensions, reinforcement, and service axial load of the test specimens are based on columns from ninestory and 12-story prototype structures designed according to the 1963 ACI 318 Building Code.

Full-scale column specimens having details and loading conditions based on the prototype columns were built and tested under combined axial and lateral loads as shown in Fig. 2. The column specimens were 458 mm (18 in.) square sections having eight No. 7 (22.2 mm diameter) longitudinal reinforcing bars and No. 3 (9.5 mm diameter) transverse ties located at 356 mm (14 in.) centers in the high moment test region extending 1780 mm (70 in.) from the base of the column. These ties were provided with only 90 bends for anchorage. The design was carried out assuming nominal unconfined concrete compressive strength of f'^sub c^ = 27.6 MPa (4000 psi) and a reinforcing bar yield strength of f^sub y^ = 414 MPa (60 ksi). Measured 28-day concrete compressive strength of the specimens and yield strength of the longitudinal and transverse reinforcing steel are given in Table 2.