UC San Diego

The cost of earthquake-induced damage to nonstructural components and systems (NCSs) has greatly exceeded the cost of structural damage. Although, numerous types of NCSs exist and contribute to this damage, one of the most prevalent is the light-weight (metal or wood stud) interior partition wall. The interior partition wall (PW) is a complex system with multiple attachments between floors, structural walls, or columns. Although these systems are not anticipated to interact with the building significantly; to date, few studies have been conducted to comprehensively evaluate if they influence the response of the building. What is known is that PWs are subjected to the dynamic environment of the building and continue to observe damage at much lower amplitude seismic demands than that of the structural system. As a result, building designers would benefit from numerical tools for predicting the impact of PW subsystems on the overall building response and evaluating the PW behavior itself. To this end, a numerical model is developed that captures the in-plane seismic response of full-height cold-formed steel framed PWs. This behavior is lumped into a nonlinear zero-length spring for ease of implementation in beam- column type finite element analyses. Experimental data is used to calibrate the model by representing the force- displacement and the dissipated energy relationships. Two different error metrics are used to illustrate the model's robustness with particular attention to the prediction of experimental behavior. To evaluate the effects of the PW on the building response, a suite of nine building models with a wide fundamental period distribution is selected. Eigenvalue, nonlinear pushover, and nonlinear time history analyses are used to examine the effects on the interstory drift and floor acceleration. Analyses indicate that the fundamental period can decrease up to 15%, due to the added stiffness of the partition wall. If the PW is neglected in the building analysis, the peak interstory drift and floor acceleration can be underestimated by approximately 50%. By tracking the period of the coupled (building-partition wall) system, the PW is shown to soften before the structural system. This softening confirms the vulnerability of the PW system at lower spectral demands, which is consistent with earthquake reconnaissance and experimental observations