Hollow section
members are often employed as bracing elements, for both structural and
aesthetic reasons. This paper describes an investigation of the response of
such members to cyclic axial loading. The influences of concrete or mortar
infill and of member slenderness are addressed.
Steel and composite
members employing three different section sizes were subjected to monotonic and
cyclic axial displacements in the inelastic range. The monotonic test specimens
had an aspect ratio of three to promote local buckling, while preventing
overall lateral buckling. On the other hand, the cyclic tests considered
members of two different overall lengths: 1100mm and 3300mm. In the case of the
monotonic test specimens and the shorter cyclic test specimens, both hollow and
filled specimens were tested. The normalised slendernesses of the cyclic test
specimens varied from 0.4 to 3.2, to cover the possible practical range, and
both elastic and inelastic buckling was observed. Failure of the less slender
specimens was initiated by local buckling in compression, followed by rupture
in tension. The presence of concrete infill was observed to influence this mode
of failure. Local tensile end-failures were observed with some specimens of
relatively small cross-section. The ductility capacities and energy dissipation
of the individual specimens are compared, and the effect of slenderness and
infill are quantified.
To
aid interpretation of the experimental results, the general purpose structural
analysis software LUSAS was employed to study local inelastic behaviour. This
showed that under tension loading, the presence of the concrete infill can lead
to non-negligible hoop stresses in the steel section, which in turn affect the
ductility capacity of the member. Design predictions of the buckling strength
and post-buckling resistance of the specimens are compared with the
experimental results. The implications of the results on seismic design
procedures for structures incorporating these types of members are also
presented and discussed.