Quantification of seismic performance factors for cold-formed steel structures

Abstract

Shear wall panel (SWP) made of cold-formed steel (CFS) is one of the lateral load resisting systems adopted in light gauge steel constructions. It is composed of CFS C-shaped framing members (studs and tracks) attached to steel or wood sheathing using screw fasteners. The objective of the research study addressed in this thesis is to define a seismic design and verification procedure for CFS framed buildings that can integrate the current seismic design framework of Eurocode 8. The approach comprises the definition of a set of design criteria, the selection and design of a set of archetype buildings, the development of nonlinear building models in the OpenSees finite element (FE) software followed by the conduction of nonlinear static (pushover) and Incremental Dynamic Analyses (IDA) of the archetype buildings following the Federal Emergency Management Agency (FEMA) P695 methodology. Two hysteresis models that take into account strength and stiffness deterioration as well as pinching, have been developed and implemented in the official OpenSees release (version 2.4.5 and above) as uniaxialMaterials entitled “CFSSSWP and CFSWSWP” for steel- and woodsheathed CFS-SWP, respectively. The proposed deteriorating models are validated using the experimental test results obtained from the literature, where a good agreement has been achieved. A seismic design procedure for CFS framed structures employing sheathed SWPs, compatible with the framework of the Eurocodes, is proposed. In order to assess the structural behaviour and generate the required data for the appraisal of the seismic design procedure, the OpenSees FE environment was used to simulate the nonlinear behaviour of CFS-SWP adopting a novel deteriorating hysteresis model. Pushover analyses and IDA have been carried out on 54 CFSSWP frames having 2-, 4- and 5-storey designed with varying seismic intensity levels. Fragility curves based on buildings collapse probability have been developed following the FEMA P695 methodology. Based on the defined design requirements, the CFS structural system evaluated in this study is shown to meet the acceptance criteria for a behaviour factor (q) equal to 2 for low- and moderate-seismicity. Further, the probabilistic seismic performance and risk assessment of CFS-SWP structures is presented adopting conventional steel moment-resisting frame systems as a benchmark with the aim of exploring the viability of using CFS-SWP as a new structural solution in seismic prone regions. Based on probabilistic seismic hazard analyses (PSHA), a site-specific selection of ground motion records for IDA has been carried out adopting the Conditional Spectrum (CS) as a more realistic target response spectrum. Subsequently, the seismic risk was evaluated over the structure lifetime (i.e., 50 years) in terms of the annual probability of exceeding the Damage Limitation, No-Local Collapse and Near Collapse limit states. The importance and usefulness of the risk metrics are highlighted and adopted as an indicator to explore the behavioural features of both structural systems. Overall, the assessment procedure showed that both systems present an acceptable seismic performance and therefore the CFS-SWP can be seen as a reliable structural solution to achieve performancebased objectives in seismic regions. Subsequently, a FE modelling protocol for screw connected, back-to-back built-up CFS columns is developed and validated using results from experiments conducted at Johns Hopkins University as part of a collaborative project. The motivations for the effort are (1) to provide modelling results to augment experiments directed at improving design guidance for built-up CFS columns, and (2) to provide a simulation path for modelling built-up CFS columns in shear wall chords that commonly experience cyclic demands. Shell FE-based models were created in ABAQUS and include monotonic loading, nonlinear geometric and material behaviour, geometric imperfections based on laser scanned measurements of tested specimens, and a contact model that includes friction. Additionally, the screw fasteners were integrated into the modelling protocol using user-defined element (UEL) subroutines capable of reproducing strength and stiffness deterioration under monotonic load as well as the pinching that occurs when screw fasteners are subjected to cyclic loads. Monotonic, concentric compression tests on 17 back-to-back CFS columns using two cross section sizes and varying fastener layouts with sheathing conditions, were simulated. Buckling deformations, strength and collapse mechanisms obtained by the models were in close agreement with the experimental results. An assessment of the loading demand on screw fasteners reveals the conservatism in built-up column fastener layout and design as required by the North American Specification for the Design of Cold-Formed Steel Structural Members (AISI S100-16 section I1.2). Also, under the tested semi-rigid column end conditions, there is little boost in axial capacity with the addition of member end fastener groups (EFGs) at the top and bottom of the columns. The developed modelling protocol will also be used, in future work, to characterise the monotonic and cyclic behaviour of axially-loaded columns so that chord stud buckling limit states could be captured in seismic simulations of CFS framed shear walls.