Seismic Floor Acceleration Amplification Based on Instrumented Building Records

The estimation of seismic forces for acceleration-sensitive nonstructural components (NSCs) in buildings including architectural, mechanical, and electrical systems and building content requires a reliable prediction of the seismic floor accelerations along the building height. These accelerations can be estimated either by performing detailed time-history dynamic analyses on a finite-element model of the supporting building or by using simplified equations as proposed in the North American and European codes. In addition, a probabilistic estimate of floor acceleration demands is essential for risk-informed performance assessment and loss estimation of these NSCs. To this end, an analytical method for the prediction of peak floor acceleration demands was proposed in FEMA-P58 [ATC (Applied Technology Council). 2012. Seismic performance assessment of buildings, prepared for federal emergency management agency. FEMA-P58. Washington, DC: ATC] based on linear and nonlinear time-history analyses performed on idealized low-to-medium-rise building models. This paper presents a validation study of the FEMA-P58 equation using two databases of recorded horizontal rooftop accelerations in 41 instrumented buildings. The first database consists of 28 buildings that were subjected to different earthquake events in California, while the second database consists of 13 buildings that were subjected to the 2011 Tohoku earthquake in Japan. The ratios of the recorded rooftop-to-ground-level accelerations in both databases were computed and compared to the FEMA-P58 predictive equation. A comparison of predicted versus recorded median and dispersion of the rooftop acceleration amplification was made and discussed. The results showed that the equation underestimated the acceleration amplification for frame buildings.