مهندسی عمران شریف (Aug 2015)
GROUND MOTION PREDICTION (ATTENUATION) RELATIONS FOR SPECTRAL ACCELERATION IN IRAN
Abstract
The design spectrum is an important input data for seismic design. Currently, elastic design spectra are used in seismic codes. The most usual methods for providing design spectra are probabilistic and deterministic, in which ground motion prediction equations (attenuation relations) are employed. In scientific societies, research into ground motion prediction equations is being undertaken to increase their accuracy. Iran also needs such accurate relations, due to its high seismicity. Considering the promotion of the national seismic acceleration network and the improved past earthquake acceleration records in Iran, it is necessary to develop new ground motion prediction relations based on new records with more accuracy.A ground motion prediction equation (GMPE) for spectral acceleration is proposed for Iran. The proposed relations are supported by analyses of 806 acceleration records from 330 arthquakes that occurred in the past. Each of the selected records has been modified separately. The modifications on acceleration records include filtering and base-line correction. The moment magnitudes of the selected arthquakes range from 4 to 7.3 and almost 98\% of them have epicenter distances less than 200 Km. Based on tectonic conditions, 3 relations are introduced for Iran including the Zagros region and the Alborz central Iran region. These relations relate spectral acceleration to parameters like earthquake magnitude, distance between sites and seismic sources and site onditions. The sites are classified as rock, and hard and soft soils, based on average shear wave velocities up to 30 meters in depth. The data used in this research includes records until 2010, and an effort is made to increase the accuracy of the proposed relations by disregarding low accurate data. Based on all these data, 3 simplified and practical relations are presented for acceleration spectra in Iran. In addition, the results of this research are compared with similar results by other researchers.
