Selecting and converting of real eartquake records according to Tbec 2018

One of the most important elements in seismic analysis of structures is using acceleration records. The correct selection of acceleration records to be used in the analyses provides more realistic information about the reaction of the structures under the influence of earthquakes. Turkish Building Earthquake Code (TBEC 2018) requires to use of acceleration records compatible with the design spectrum of the structures in time history analysis. It is almost impossible to select the same acceleration records as the design spectrum. Therefore, acceleration records need to be scaled or converted. In this study, it is evaluated to convert earthquake records to provide spectral transformation instead of the simple scaling method widely preferred in the literature. The thesis aims to create acceleration recording sets that can be used in the analysis by finding records whose authenticity is preserved as much as possible and where the match with the target response spectrum is maximum among the actual earthquake records. Two hundred acceleration records were obtained from the Pacific Earthquake Engineering Research Center (PEER) database taking into account parameters such as fault type, earthquake magnitude, and distance to the fault for Diyarbakır city. For the DD-2 earthquake ground motion level, the design spectrum of each location with ZB, ZC, and ZD ground class was defined as the target response spectrum in the SeismoMatch, the acceleration records from the database were converted to be compatible with the spectrum. In combination with the average mismatch and maximum mismatch data of the spectrums belonging to the converted acceleration records, a different strategy was applied for the acceleration records to be selected by analyzing the number of iterations used in the converting in the SeismoMatch. In TBEC 2018, the most suitable acceleration recording sets were created for three locations that can be used in one- and two-dimensional analyses of the records in the catalog, taking into account the fact that there are no more than three records in the acceleration recording sets. Acceptable results were obtained when the number of iterations and mismatch percentages used in the acceleration recording sets were taken into account. In addition, the averages of the acceleration records spectrum converted for all three sets did not outage the design spectrum ordinates. In this way, the desired condition was observed in TBEC 2018.