Virdi and Dowling 1980
A study of concrete-filled steel tubes was undertaken to investigate the ultimate bond strength between the concrete core and steel tube in the absence of shear connectors. A number of parameters were studied including the concrete compressive strength, the length of the steel-concrete interface, the D/t ratio, the surface roughness, and the effect of manufacturing imperfections, among others. Several push-out tests were conducted, each varying one parameter at a time. Three specimens were tested for each given parameter. The authors gave each parameter a detailed and comprehensive treatment, resulting in an very thorough examination of bond.
Experimental Study, Results, and Discussion
The specimens were tested by loading the concrete with a steel plate 0.5 in. smaller than the inside diameter of the steel tube. The concrete was loaded at a rate of 3.0 kips per minute and allowed 1.5 in. of travel. All specimens showed a uniform load-longitudinal deflection response, with a high initial bond stiffness followed by a marked reduction in stiffness to a relatively flat slope. Only the tubes with smooth inner surfaces did not exhibit some residual strength (i.e., have a positive load-deflection slope) after the concrete core ran out of travel.
The critical bond strength in the tests was taken as the bond strength at a local critical strain at the interface of 0.0035 (in correlation with the ultimate crushing strain of concrete). It was required to establish a value that ignored the frictional effects encountered in the latter stages of the load-deflection curve. This could have been accomplished using the 0.2% offset stress, but the former method was preferred since the offset stress method would require an initial stiffness value, which is difficult to obtain due to initial settlement of the concrete. The bond strength value at 0.0035 was termed the “ultimate bond strength.”
The tests showed that the ultimate bond strength was not affected to a large extent by the concrete strength, interface length, tube thickness, or tube diameter. The main factor in the amount of bond is the mechanical keying of the concrete with the irregularities of the steel tube. This keying may be manifested in two ways. First, the concrete is bonded by the interaction with surface irregularities due to the roughness of the tube. This type of interlocking, termed microlocking, must be overcome if the concrete core is to move as a whole. It is microlocking which defines the ultimate bond strength. At the ultimate bond strength, the concrete at the interface crushes and the stiffness decreases substantially. The second type of interlocking, macrolocking, occurs due to the nonuniformity of the tube, i.e., out-of-straightness or out-of-roundness. Macrolocking accounts for the frictional resistance provided beyond the ultimate bond strength and was illustrated by the remarkably parallel load-deformation slopes of all the tests in the later stages of the curve. It was further confirmed by tests of tubes that had smooth inner surfaces and were manufactured to more exact tolerances. These tests showed a complete absence of macrolocking. Both microlocking and macrolocking may be improved by better compaction resulting in an increase in bond strength.
By applying a statistical analysis to the test values, the characteristic ultimate bond strength recommended for design was determined as 150-160 psi.
Virdi, K. S. and Dowling, P. J. (1980). “Bond Strength In Concrete Filled Steel Tubes,” IABSE Periodica, August, pp. 125-139.