ABSTRACT
The authors have recently completed a series of full-scale fire tests on rotationally restrained concrete filled tubular column assemblies. These columns were pinned at one end and rotationally restrained at the other. Test parameters include two tube sizes, two connection types, different levels of column axial loading and initial bending moments. The test fire exposure followed the standard fire curve. According to the current design recommendation in ENV 1994-1-2, the effective length of these columns should be 0.7 times their length. However, the results of these tests suggest that local buckling of the steel tube played a vital role in determining the column effective length. For the thicker steel tubes (RHS 100 × 200 × 12.5 mm) where there was no local buckling, the ENV recommendation is applicable. For the thinner tubes (RHS 100 × 200 × 5 mm) where local buckling occurred, the results of design calculations suggest that the column effective length should be taken as the distance from the largest local buckle near the connection to the pinned column end. For many columns, this means that the column effective length should be the column full height. The main factor to affect the position of the largest local buckle appears to be the applied axial load level whilst the influences of connections and initial bending moments do not seem to show any clear trend.
