Numerical Investigation of Thin Gauge, High Strength Cold-Formed Steel Strap Braced Walls Under Combined Lateral and Vertical Loads

Abstract

World-wide, cold-formed steel (CFS) construction is becoming increasingly popular for residential buildings. The advantages of CFS are high dimensional accuracy, lightweight, resistance to warping, shrinking and creeping, excellent durability, low embodied carbon and excellent CNC manufacturing abilities. In Australia and New Zealand, the CFS members are made of higher strength, thinner gauge material than Northern Hemisphere. Traditional applications have been in one and two-storey houses, and in New Zealand, these have shown excellent performance in recent severe earthquakes from 2010 to 2016. There is a growing demand to extend CFS construction into mid-rise buildings up to 6 storeys. For a 6-storey CFS structure in high seismic areas, higher demands are imposed than those on two-storey buildings, which necessitates a different approach. Diagonal straps (strap braces) acting as tension members are developed overseas for such CFS lateral force resisting systems. A significant amount of lateral stiffness and energy dissipation through plastic deformation must be developed for a wall panel in these mid-rise buildings to be seismically resilient. In addition, some of these lateral load resisting walls are vertical load-bearing walls for the structure. At lower levels, they are required to resist significant constant vertical loading during the earthquake, thus being subjected to combined lateral and vertical load. Extensive experimental testing and numerical modelling has been undertaken on strap braced walls under lateral loading made from relatively thick, medium-strength steel members in Northern Hemisphere CFS construction and some under combined lateral and vertical loading. There has been limited lateral load testing of the Australasian high strength, thin gauge CFS walls. These walls have in Australia have shown very different behaviour to the Northern Hemisphere walls. There has been no experimental testing or numerical modelling of these thin gauge and high strength strap braced CFS walls under combined lateral and vertical loading. Currently, the research gap is being addressed in parallel research projects; stage one is numerically based and runs alongside stage one of an experimentally based project using a purpose-built test rig that can deliver constant vertical and cyclic lateral load. The first stage of this work aims to determine the performance of the individual strap braced wall panels under combined loads. The panels are designed using capacity design principles to accommodate seismic overloading by yielding the strap braces in tension without failure of the connections, studs, chords or tracks. The present numerical study investigates the behaviour of a proposed typical CFS strap braced panel constructed from 0.95mm thick, grade 550 steel members under lateral and vertical loading. The numerical models have been developed to the most realistic extent using the commercial line element software SAP2000. Various boundary conditions have been implemented to observe the response of the wall with different strap configurations and connections under lateral and vertical loading. Since SAP2000 is not a finite element analysis software, several probable boundary conditions have been developed, and the influence of these determined. The study includes observing the inelastic behaviour of the tension strap braces through the load-displacement characteristics. Various configurations of strap braces have been implemented to determine the inter-storey drift and compare that with the requirements of the Seismic Loadings Standard, NZS 1170.5 (Standards New Zealand 2004). These models have shown that the one-sided and both-sided strap braced configurations easily meet the strength and stiffness requirements of the standard. The nonlinear buckling analysis carried out to determine the out-of-plane bucking generated from the compression loads from the upper storeys showed similar behaviour to the sidesway buckling modes of the experimental wall tested in the Structures Testing Laboratory at the University of Auckland. The numerical models have enabled an understanding of the tension, compression and shear transfer mechanisms within the wall panel and from the wall to the foundation system, which will be required to achieve good wall/floor/wall performance when these walls form vertically stacked shear walls in mid-rise buildings. The numerical modelling and experimental testing have shown the need for a stiff load path in the horizontal plane between adjacent vertical walls and between the bottom wall and the foundation to prevent failure of the wall tracks, which are weaker than the chords and studs. It is expected that if a wall-to-floor-to-wall connection can be developed, which allows rocking-induced uplift of the tension side of the wall while suppressing sliding shear between walls leading to track failure, this will greatly benefit the overall seismic response. A potential detail for this has been developed, and the numerical modelling has determined the influence of a range of realistic horizontal and vertical stiffnesses of the wall/floor/wall connections on the overall response of the wall. This will be important to inform the next stage of the experimental testing, focussing on the wall/floor/wall and wall/foundation systems. The different analyses of the presented numerical models show potential for such walls to be implemented in mid-rise buildings in New Zealand. Nevertheless, before a potential execution, a detailed study should be made to observe the performance of such CFS shear walls under the shear uplift in mid-rise buildings. In addition, for the numerical steps to be implemented in the industry with SAP2000, options for the connection between the strap braces and gusset plates should be further studied to make the connection precise and realistic.