Page 1 of 10 Peak Horizontal Floor Accelerations due to Near-Fault Ground Motion Benyamin Monavari M.Sc., Department of Civil Engineering, Faculty of Engineering, Kharazmi University, Tehran, 15719-14911, Iran. std_monavari@khu.ac.ir Ali Massumi Associate Professor in Structural Engineering, Department of Civil Engineering, Faculty of Engineering, Kharazmi University, Tehran, 15719-14911, Iran. massumi@khu.ac.ir ABSTRACT: This study uses representative numerical models of four reinforced concrete frame buildings and several types of ground motion for incremental dynamic analysis and investigates the distribution of peak horizontal floor acceleration along the height of buildings in response to slight and major damage. Ground motions in a near-fault region have special characteristics, particularly when these motions contain forward directivity effects such as intense velocity pulses. The seismic design of structures located close to active faults must account for these characteristics. This study compares peak horizontal floor acceleration from two near-fault data sets; one containing 30 pulse-type motions with forward directivity effects and the other containing 20 motions without forward directivity effects. The results show that current seismic code provisions do not provide an adequate characterization of peak component accelerations. This study also confirms that peak horizontal floor acceleration distribution changes in response to changes in peak ground acceleration. 1. Introduction The failure of nonstructural components after the 1971 San Fernando earthquake was recognized as critical for two reasons: (1) the damage of nonstructural components resulted in major economic loss and; (2) it posed a threat to life (Ray-Chaudhuri and Hutchinson 2004). Special design requirements should be considered where the partial or total collapse of structural and non-structural components (NCs) must meet life safety and collapse prevention performance levels. It is also very important to prevent damage to structural and non-structural components at the immediate occupancy and operational performance levels. Hirakawa and Kanda (1997) reported that, as a result of the 1995 Hyogo-ken Nanbu earthquake, 40% of 210 reinforced concrete (RC) buildings and 40% of structural components had sustained damage. Consequently, mitigation of nonstructural damage will reduce economic loss. In ground motions like those produced by the 1989 Loma Prieta, 1994 Northridge, and 2001 Nisqually earthquakes (Shephard et al. 1990; Hall 1995; Filiatrault et al. 2001), economic loss from nonstructural components generally exceeded that from structural components. Several studies have reported that economic loss from nonstructural components are substantially greater than those resulting from structural damage (Ayers et al. 1973; Reitherman and Sabol 1995). To estimate the design force for acceleration-sensitive NCs, several recent building codes recommend a linear variation of peak horizontal floor acceleration (PHFA) along the height of a building (Singh et al. 2006). Uniform building code recommendations (UBC 1997) and the building seismic safety council (NEHRP 94) suggest that PHFA varies from one at ground level to four times the peak ground acceleration (PGA) at roof level (trapezoidal distribution). In contrast, NEHRP 2003 and the International Building Code (IBC) (ICC 2006) assume a linear variation where the PGA at roof level is three times that of the PGA at ground level. The provisions used in these codes were developed empirically on the basis of floor acceleration data recorded in buildings during California earthquakes (Kehoe and Freeman 1998).