Experimental investigation of multi-layered strong wood-frame shear walls with nonstructural Type X gypsum wallboard layers under cyclic load

Abstract

A fundamental premise for the design of resilient light frame timber buildings in areas of high seismic activity consists of achieving strong shear walls while keeping nonstructural damage to a minimum, especially by preventing the brittle failure of Type X gypsum wallboard (GWB) finish layers. Under this premise, it has been commonly thought that the stronger the shear wall the lesser the structural contribution of the finishes, and therefore the effect of GWB finishes in strong timber shear walls has been typically ignored by design codes and mechanical models. In the research reported in this paper, the influence of nonstructural finishes on the structural characteristics of a strong light frame shear wall was experimentally and numerically investigated. The wall evaluated in this paper, denoted here Multi-Layered Strong Shear Wall (MLSSW), is representative of walls located at the 1st story of a 7-story building located in Chile, a country where the level of seismic activity is high. The MLSSW has a continuous rod system, double OSB sheathing, robust timber framing, closely spaced nails to attach the OSB sheathing layers to wood-frame members, and is covered at both sides by two layers of screwed Type X GWB. The characteristics of the MLSSW, particularly the quantity and type of nonstructural finishes, are different from those of typical light frame timber walls. Full-scale connection-level tests, both monotonic and cyclic, are reported first, followed by descriptions of the full-scale cyclic assembly-level tests. For comparison purposes, a bare strong wall (i.e., without the GWB layers) was tested as well. The structural response of the MLSSW was evaluated in terms of stiffness, strength, energy dissipation and plastic deformation characteristics. The applicability of the SDPWS analytical model was also evaluated, as well as the applicability of existing numerical models to predict the cyclic behavior of the MLSSW.