Shakir-Khalil and Mouli 1990
Following up on the study reported in 1989, the authors conducted nine more tests on full-scale rectangular CFT columns, incorporating two column sizes. The majority of the columns were subjected to biaxial bending. Also tested were a series of short CFT specimens in axial compression to establish the squash load of stub columns having this type of section. The results were discussed and compared with the British standard BS5400. The main parameters of study were the concrete strength, the tube strength, and the effective length.
Experimental Study, Discussion, and Results
Short Column Tests. Tests were performed to determine the effect of different parameters on the carrying capacity of various sections. The capacity increased with an increase in the size of the steel section and the with the use of higher strength concrete. However, the capacity decreased with the use of higher strength steel and/or an increase in the column length. In both of the latter cases, the steel carries more of the load, decreasing the concrete's contribution. The benefits of the concrete's compressive strength begin to diminish and the CFT behaves more like a hollow tube.
Long Column Tests. The columns were oriented in a horizontal position with the larger cross-sectional dimension in the vertical direction to minimize the effect of the column self-weight. The loading process was very similar to that described in Shakir-Khalil and Zegiche (1989). The failure of the columns was typified by yielding in the compression corner of the tube at midheight at about 90% of the failure load, followed by tension yielding on the opposite corner, and leading to an overall buckling failure. No local buckling was noticed in any of the sections. As expected, filling the tubes with concrete produced a considerable gain in strength -- 12% to as much as 65%, depending on the column dimensions and material properties. The percent gain was augmented by an increase in concrete strength and an increase in the tube size. Increasing the yield strength of the steel and increasing the length had detrimental effects.
Bond. Bond tests paralleling those conducted in Shakir-Khalil and Zegiche (1989) were repeated here with similar results. It was noted, however, that the higher strength concrete produced the higher value of bond strength.
A computer program developed by the authors was briefly described. They used the program to generate several interaction diagrams, plotting P/Pu versus M/Mu. The values of both quantities causing full plasticity of the cross section were obtained by the program. The calculation of the ultimate moment Mu was based on the use of rectangular stress blocks assuming strengths of 0.6fcu for concrete in compression and no strength in tension, and fy for steel in both tension and compression. The curve was obtained by varying the position of the neutral axis, and considering the equilibrium between the resultant stress blocks and the applied eccentric load. These could be generated for various sizes and properties of the columns.
Shakir-Khalil, H. and Zeghiche, Z. (1989). “Experimental Behavior of Concrete-Filled Rolled Rectangular Hollow-Section Columns,” The Structural Engineer, Vol. 67, No. 19, pp. 345-353.
Shakir-Khalil, H. and Mouli, M. (1990). “Further Tests on Concrete-Filled Rectangular Hollow-Section Columns,” The Structural Engineer, Vol. 68, No. 20, pp. 405-413.