Journal of Seismology and Earthquake Engineering
Volume:7 Issue: 2, Summer 2005

During recent seismic events, non-ductile failure modes of many existing structures occurred. Retrofit of these structures before the earthquake provides a feasible cost-effective approach to reduce the hazard to occupants'' safety and owners'' investment. The response of two reinforced concrete frames was examined under seismic excitation. The 9-storey and 18-storey frames are part of the lateral load resisting system in two office buildings that were designed according to the 1960s code provisions. The frames were analyzed assuming flexible joint response by considering the joint shear deformation or assuming traditional rigid joints. Two rehabilitation techniques were proposed to improve the dynamic response of these frames. Fibre reinforced polymer (FRP) jackets were used as a local rehabilitation technique to enhance the joint shear strength and ductility. As another option, X-steel braces were installed in the middle bay of the frame along its height as an alternate lateral load resisting system. For each frame, failure sequence and interstorey drift were examined. It was found that FRP wrapping eliminated the brittle failure modes without significant change in the structural response. However, steel bracing significantly contributed to the structural stiffness and reduced the maximum interstorey drift of the frames.

The seismic performance of unrestrained objects is critically dependent on the displacement demand behaviour of the building floor. The risk of an object overturning can be estimated from the dual independent criteria of object width and height, as opposed to the usual single criterion of the object aspect ratio (or slenderness ratio) based on static analysis. An object is at risk from overturning if the displacement demand of the floor exceeds one-third of the width of the object. According to floor amplification clauses in earthquake codes of practice, the filtering effects of a building amplify ground motions up its height. However, the building may also behave as an isolation medium, which attenuates the transmitted motions. These two perceptions seem contradictory. This paper aims to resolve this significant dilemma and hence contribute to improving the fundamental understanding of the dynamical processes of damage to building contents. Floor spectra of buildings, as presented in the paper, demonstrate both amplification and isolation actions.

Attenuation relations are developed based on information in the Iranian acceleration data bank (IADB) containing 279 entries from about 30 seismogenic areas across the country. The peak ground horizontal (PGH) and peak vertical accelerations (PGV), varies from a few cm/s2 to over 1000 cm/s2. Moment magnitudes (Mw) vary from about 3.0 to 7.4; and earthquake depths vary from near surface to over 100 km; however, except a majority of depth that are kept at 33 km in the locating process, most depths are about 10 km. Epicentral distances (EPD) vary from 2 km to nearly 250 km. The data bank also includes four site conditions, S. The least squared multi-stage regression solutions for acceleration attenuation are calculated for three cases. The predicted PGH and PGV accelerations are compared with uncorrected high accelerations components near the source of Bam earthquake of 2003; the high PGH acceleration is reproduced, but the estimated PGV acceleration is lower by a factor of 2 to 3. In addition, estimations based in this work are compared with several other studies and discrepancies are discussed.