Tomii, Yoshimura, and Morishita 1980
Bond tests were conducted on circular, square, and octagonal CFT columns and the methods to improve the bond strength between steel and concrete along the interface were investigated. In addition, the authors presented formulations to estimate the bond strength and slip.
Experimental Study, Results, and Discussions
The steel and concrete of the CFTs was supported simultaneously along the bottom. The axial load was applied at the top on to the steel tube alone, and the steel tubes were completely filled with concrete. Two methods were discussed to improve the bond strength between steel and concrete. In the first method, expansive concrete was used to fill the steel tubes. In the case of second method, steel tubes with checkered inside walls were provided. The checkered and smooth steel tubes were filled with either expansive or ordinary concrete. The D/t ratio was 46.9 and the L/D ratio was 4.9. The measured compressive strength of concrete ranged between 2.28 and 4.85 ksi. The measured yield strength of steel tube was 36.6 ksi and 36.3 ksi for the smooth and checkered tubes, respectively. All the tubes were annealed before testing.
The longitudinal strains at different locations of the steel tube were measured. When the strain distribution along the length of the column was plotted, it was observed that as the strain at the top of the steel tube increased, the strain distribution became nonparellel to the column axis. This was due to the transfer of load from the steel tube to the concrete core and the formation of bond stresses between the steel and concrete. When the strain value at the top reached 9x10-4, it was not possible to see the continuity of strains between concrete and steel along the length of the column except for the circular and octagonal specimens having both expansive concrete and a checkered steel tube inner surface. Thus, the bond strength for these specimens was the largest.
The circular specimens with ordinary concrete and a smooth internal tube surface had average bond strengths of 28.5 to 56.9 psi, which were lower than the bond strengths for round steel reinforcement in concrete. The bond strengths were found to decrease as the concrete strength got higher. These bond characteristics were attributed to the facts that shrinkage of high strength concrete was larger, and steel had a larger Poisson ratio in the elastic range than that of concrete.
The average bond strength for square and octagonal specimens with ordinary concrete and smooth internal tube surfaces ranged from 14.2 to 28.5 psi, which were half of the average bond strength values for the smooth circular specimens with ordinary concrete. The average bond strength for the square and octagonal columns remained constant as average slip increased.
When the columns were filled with expansive concrete, the same type of bond response was obtained for all of the specimens. The average bond strength decreased as the average slip increased. The bond strength of expansive concrete was found be greater than ordinary concrete initially, while the same strength was attained as the loading proceeded. For the square specimens, the concrete strength did not have any significant effect on bond strength. However, in the case of circular specimens, the bond strength was found to be larger when the concrete strength was high. The average bond strength of the square specimens was lower than circular ones when expansive concrete was utilized.
The columns with checkered surface had greater bond strength than the ones having a smooth surface. For some of the specimens, it was observed that average bond strength improved gradually as slip increased.
Analytical Study
The same equations discussed in Morishita et al. (1979a, b) for Fmb and Ds were proposed again in this work. In addition, the formulation to calculate the average bond stress when the strain compatibility between steel and concrete is not satisfied along the column length was given as:
In the above formulation, the longitudinal stress of steel at the bottom σsb was calculated approximately, as it was not measured. For this purpose, the longitudinal stress at the lower most measured point and the bond stress distribution between this point and bottom of the specimen were used.
For circular columns, the following equation was presented to calculate the required column length to transfer axial load at the beam-to-column connection from steel to concrete.
Reference
Tomii, M., Yoshimura, K., and Morishita, Y. (1980a). “A Method of Improving Bond Strength Between Steel Tube and Concrete Core Cast in Circular Steel Tubular Columns,” Transactions of the Japan Concrete Institute, Vol. 2, pp. 319-326.
Tomii, M., Yoshimura, K., and Morishita, Y. (1980b). “A Method of Improving Bond Strength Between Steel Tube and Concrete Core Cast in Square and Octagonal Steel Tubular Columns Transactions of the Japan Concrete Institute,” Vol. 2, 1980, pp. 327-334