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

K^sub e^ = elastic stiffness,

δ^sub y^ = yield deflection,

K^sub 1^ = post yield stiffness,

δ^sub s^ = deflection at the start of the strength degradation,

K^sub 2^ = strength degradation stiffness,

K^sub ul^ = K^sub e^(δ^sub y^/δ^sub m^)^sup β^ = unloading stiffness

δ^sub m^ = maximum specimen deflection during analysis, and

β = unloading coefficient.

These variables were calculated from the moment-curvature analysis converted to a force-displacement plot by calculating a plastic hinge length using Priestley's9 equation. The elastic stiffness is calculated using the slope through the point of the first bar yield. The post yield stiffness can be calculated using the failure point or a perfectly plastic slope, and the equal-area method,7 which provides the yield displacement. The strength degradation slope is more difficult to determine because moment-curvature analysis does not provide well-defined failure envelopes. The unloading coefficient is typically taken as 0.25 for concrete columns. Priestley's9 method for calculating plastic hinge length correlates well with the measured displacements for the columns.

Retrofitted columns

Figure 9 shows the predicted versus measured time-history response for Column 6FS2. The results of the analysis compare well for lateral forces and lower ductility, but the effect of the lap splice slippage for the larger measured displacements can be seen by the large residual offset that occurs in this figure. The hysteretic model can capture lower ductility but the lap splice slippage occurring during the response cannot be captured accurately. Figure 10 shows the predicted versus measured time-history response for Column 6FC.

CONCLUSIONS

Although the scope of testing was limited to three 1/3-scale circular columns with two types of restraining jackets, the following general conclusions may be drawn from the results:

1. The study showed that Caltrans retrofitting procedures improve both capacity and ductility of as-built columns using either steel jacket shells or carbon fiber sheets;

2. The column retrofitted with carbon fiber had performance levels significantly greater than those with steel jacket retrofits;

3. The lap splice in the steel-jacketed column was sensitive to multiple cycles and degraded rapidly before achieving high displacement ductility. The lap splice in the column retrofitted with carbon fiber was not as susceptible to the low level cycles;

4. Load path has a measurable effect on initial performance of reinforced concrete bridge columns subjected to dynamic excitation;

5. Undamaged columns subjected to a high amplitude motion exhibited a slightly higher measured capacity than columns subjected to incrementally increasing ground motions typical in shaketable testing; and

6. Moment-curvature analysis and simple hysteretic models were able to predict the response of columns with acceptable accuracy, although the incorporation of a refined lap-splice model would improve the displacement predictions.

DISCLAIMER

Conclusions of this report are those of the authors only, and should not be construed to be endorsed by Caltrans.