LAUSR

Seismic uncertainties pose response estimation procedures with serious concerns about proper reflection of ground motion intensity measure (IM) and reasonable scaling of IM in a scenario-based study. The extent of response variability due to these uncertainties is quantified in this article for steel moment-resisting frames (SMRFs). For this purpose, three 3-, 7-, and 15-story SMRFs are designed and nonlinearly modeled in OpenSees Program. To establish an initial collection of building responses in a diverse range of intensities and for multiple ground motion records, incremental dynamic analysis is performed using a set of 44 records. At the initial phase, the efficiency of two well-known IMs, i.e., peak ground acceleration (PGA) and first-mode spectral acceleration Sa (T1), is re-evaluated for the studied structures. This is achieved by quantifying the dispersion of building response data at different levels of behavior nonlinearity. The dispersion of response data caused by PGA is shown to reach 1–13 times of the Sa’s where the 1 ratio is related to the linear response range of the shortest 3-story structure and the 13 ratios reflect the collapse response of the 15-story structure. Afterwards, scatter of building responses extracted using different number of acceleration records and various intensity scaling methods is evaluated. The dispersion of the response data shows clear correlation to the selected scaling method whereas using larger suites of records does not necessarily lead to a decrease of the data dispersion.