An experimental investigation of axially loaded CFTs using spiral welded tubes was described and the results were compared with the ACI-NBC design formulas and the values obtained from the tangent modulus method.
Experimental Study, Results, and Discussion
The procedure involved testing specimens under pure axial load to failure. The spiral welded tubes used in the test were manufactured by bending a narrow steel strip in a helix and welding up the tube along the junction of the steel edges. At the paper's publication, this type of tubing cost 40% to 50% less than equivalent cold drawn seamless tubing.
Residual Stresses. Care was taken to measure the residual stresses in the tubes. Large manufacturing stresses were realized. The results were graphed in the paper and the author suggested that the values serve as a pessimistic bound since the ratio of wall thickness to diameter (a factor in the amount of residual stresses present) of the tested tubes was greater than typical tubes in practice. He also recommended taking the stress-strain properties from compression tests of complete cross sections which gave a much better indication of residual stresses than coupon tests.
Axially Loaded Columns. The long columns failed by buckling and exhibited no abrupt loss of capacity at large strains, i.e., the columns behaved plastically much like the steel stress-strain curve. Overall, the spiral welded tubes behaved much like seamless tubes.
The author assumed the following as a lower bound for the strength of a stub column under axial compression (see also Gardner and Jacobson, 1967) :
This value was conservative for all experimental values. For long columns, as in the 1967 paper, the author used the tangent modulus formula:
Est and Ect are the respective steel and concrete tangent moduli. These values were obtained from the uniaxial compression test for the steel and by calculation from the triaxially loaded stub column for the concrete. Estimating the tangent modulus for steel in such a fashion was incorrect if the steel was loaded biaxially. The author found, as Furlong (1968) did, that at higher strains, Poisson's ratio for concrete exceeded that of steel and a hoop tension was induced in the steel tube. The yield criterion of the steel tube was expressed as the sum of the tensile hoop stress and the compressive longitudinal stress. Therefore, the longitudinal stress at which yield occurred under biaxial loading could be significantly less than that under uniaxial loading. Since the tangent modulus of the steel decreased rapidly near yield, using uniaxial properties for the steel was wrong since it was probably yielding much sooner.
The ACI-NBC method as described in the summary of Gardner and Jacobson (1967), may be used with confidence for design in both seamless and spiral welded tubes (which behaved much like seamless).
Gardner, N. J. (1968). “Use of Spiral Welded Steel Tubes in Pipe Columns,” Journal of the American Concrete Institute, Vol. 65, No. 11, pp. 937-942.