An S-Transform-based Technique for Shortening the Strong Motion Duration of Accelerograms for Rapid Time History Analysis

Document Type : Research Article


1 Ph.D., Independent Researcher, Tehran, Iran

2 Assistant Professor, Department of Civil Engineering, Najafabad Branch, Islamic Azad University, Najafabad, Iran

3 Associate Professor, Structure Research Center, International Institute of Earthquake Engineering and Seismology (IIEES), Tehran, Iran


There are some cases, such as irregular or complex structures, in which the simplified analysis methods recommended by prevalent seismic codes are not able to yield results with acceptable precision. In such circumstances, the use of Non-Linear Time-History Analysis as the most robust response analysis approach is mandatory. This method is usually very time-consuming (mainly due to the small time steps). Therefore, any technique to reduce the computational cost, while
keeping results in an acceptable range of precision, would be desired. There are some methods in the literature to increase the size of time step and simplify the recorded accelerograms by modifying time series. Although these methods may show negligible bias in results in terms of maximum response values of simple structures, they, definitely, impose some error due to the manipulation in the frequency content of the original signal. In this paper, an S-Transform based
sim-plification procedure is introduced to overcome the drawbacks of available methods. The proposed method enables users to trace and compare the timefrequency variation of original and simplified signal to keep the general timefrequency pattern of signal and prevent the undesired omissions of components. Eleven strong ground motion records were selected to be used in the efficiency evaluation of the proposed method. The results of analyses for a range of single and multi-degree of freedom non-linear dynamic systems confirm the ability of shortened records to represent their original counterparts in terms of response characteristics. The results show that by a reduction of 71% in the analysis cost, the maximum observed error in the estimation of collapse of MDOF structures will be lower than 10%.