Retrofitted Concrete Bridge Columns Under Shaketable Excitation

ACI Structural Journal, Jul/Aug 2005 by Laplace, Patrick N, Sanders, David H, Saiidi, M Saiid, Douglas, Bruce M, El-Azazy, Saad

Three 1/3-scale circular flexure-dominated, retrofitted, reinforced concrete bridge columns were studied using the University of Nevada, Reno, seismic shaketable system. Two steel jacket retrofit and one carbon fiber retrofit columns were tested and compared to a previously studied as-built column. The steel jacket retrofit columns showed increased capacity and ductility over the as-built but were adversely affected by low level motion and variable load path. A significant increase in capacity and ductility was observed for the carbon fiber jacket retrofit. Simple time-history analytical models can predict the performance of the columns adequately, although lap splice degradation is not well predicted.

Keywords: column; fibers; seismic.

(ProQuest Information and Learning: ... denotes formulae omitted.)

INTRODUCTION

After the San Fernando Earthquake in 1971, singlecolumn bridges were initially identified as the most vulnerable to earthquake damage. These columns were generally deficient in two respects; low transverse steel causing either premature shear failures or low ductile response, and lap spliced longitudinal steel at the base of the columns causing premature bond failure. Research led to retrofit procedures using steel jackets to improve these deficiencies. 1 Although many of the retrofitted steel-jacketed columns performed well in the 1994 Northridge Earthquake, new technologies such as carbon fiber retrofitting had yet to be implemented in a widespread way.2

The objective of the project was to investigate the performance of pre-1971 columns that have been retrofitted with steel and carbon fiber jackets,3,4 and to study the effect of load path. To meet these objectives, three flexuraldominated bridge columns were tested using the 1940 Imperial Valley Earthquake (El Centro) on the University of Nevada shaketable system: two steel-jacketed retrofit columns and one carbon fiber retrofit. The effect of load path was studied by varying the testing sequence between the two steel jacket retrofit columns. Simple analytical models for momentcurvature and nonlinear analysis were used to compare calculated performance versus measured.

RESEARCH SIGNIFICANCE

This study primarily focused on differences between steel jacket and carbon fiber retrofit schemes and the improvement of as-built columns, load path, and simple hysteretic models. Both steel jacket and carbon fiber retrofit schemes greatly improved the ductility of poorly detailed lap-spliced columns. The carbon fiber jacket performed better than the steel jacket retrofit. Simple hysteretic models provide good correlation with measured results, providing an alternative to finite element analysis. Load path had an initial effect on column capacity but the effects diminished after subsequent excitation.

SPECIMEN DESIGN

Prototype

The as-built prototype column used for the retrofits was the same design used in a previous study.3 The prototype had a lap splice length of 24d^sub b^, an aspect ratio of 4.5, and an axial load ratio of 10%. The prototype used normalweight concrete and 276 MPa (40 ksi) reinforcing bar, a 1.22 m (48 in.) diameter, and a length of 5.5 m (216 in.).

Three identical as-built column models were constructed for the implementation of steel shell and carbon fiber retrofits. For the specimens, a length scale of 1/3 relative to the fullscale prototype was chosen based on the test setup capacity (see Table 1). This scale ensured the column could be tested to complete failure.

Steel jacket retrofit columns 6FS1 and 6FS2

Two of the as-built columns were retrofitted with steel jackets to improve confinement in the plastic hinge zone and prevent premature lap splice failure. The steel jacket design was based on the Caltrans Memo to Designers.5 The design confinement stress σ^sub c^ and maximum design jacket strain ε^sub max^ were 2.07 MPa (300 psi) and 0.001, respectively. Equation (1) defines the required thickness of the steel jacket shell.

... (1)

Equation (2) defines the jacket volumetric ratio.

... (2)

where E is the steel jacket elastic modulus and D is the column diameter. The required jacket thickness based on a column diameter of 0.41 m (16 in.) was 1.12 mm (0.044 in.) (approximately 18 gauge steel). This small thickness requirement was impractical for fabrication and welding as a circular steel shell. The steel jacket thickness was increased to 3.175 mm (0.125 in.) to improve fabrication and weldability. A 25.4 mm (1 in.) gap was required at the base of the jacket between the jacket and footing surface to prevent concentrated stresses due to large column rotations and impact of the shell and footing. A 12.7 mm (0.5 in.) gap was formed between the shell and the column surface to allow for the application of the grout, using a metallic nonshrink grout.

Carbon fiber jacket retrofit Column 6FC

One as-built column was selected for a carbon fiber retrofit. The design of the carbon fiber retrofit was based on the Caltrans Memo to Designers.5 The design confinement and maximum strain for the carbon fiber jacket in the plastic hinge zone were the same as the steel jacket retrofit. The design requirements for the nonplastic hinge zone for the confinement stress and maximum jacket strain were 1.03 MPa (150 psi) and 0.004, respectively. The jacket modulus E is the carbon fiber elastic modulus in the primary fiber direction (the circumferential direction). The design equations for the carbon fiber wrap are the same as for the steel jacket design. After computing the required jacket thickness t^sub j^, the number of layers of carbon fiber was found from t^sub j^/t^sub f^ where t^sub f^ is the individual dry fiber sheet thickness of 0.165 mm (0.0065 in.), specified in the Caltrans Memo to Designers.5 The number of layers required in the plastic hinge zone is based on the elastic modulus E, given in the memo as 218.5 GPa (31,700 ksi). For the plastic hinge zone, 6.3 layers were required but seven layers were used. The nonplastic hinge zone region required 0.79 layers but one layer was used. The seven layers were applied to the column up to a height of 1.5D, where D is the column diameter. The remaining height was wrapped with one layer as per the nonplastic hinge zone requirement. A gap of 25.4 mm (1 in.) was created at the base of the column to prevent the carbon fiber wrap from contacting the footing surface under high rotations.