Combining Discrete Wavelet Transform and Strong Ground Motion Duration to Reduce Computational Costs in Dynamic Analysis of Structures

Abstract

The dynamic time history analysis is the most accurate method for calculating the structure response relative to the vibrating load. The use of this method is time-consuming due to the long-duration earthquakes and is generally less commonly used. Wavelet trans-forms are one of the new methods for filtering waves. The duration of a strong ground motion is actually a concept that can be used to turn an earthquake with a long time into an earthquake with a shorter time. In this paper, two concepts of filtration using wavelet transform and the use of a duration of strong ground motion are used for the time history analysis of structures. For this purpose, the acceleration of the earthquake in 5 states was filtered by wavelet transform. In each stage, the high and low frequencies are separated, and the number of earthquake acceleration points is half-past. Then the beginning and the end of the main earthquake and the approximate waves obtained from the wavelet transform are removed using the concept of the duration of the strong ground motion. Finally, two two-dimensional and three-dimensional structures are analyzed using the main earthquake and the waves obtained by combining the two concepts of wavelet transform and the duration of strong ground motion. The results show that by reducing the number of acceleration points of the main earthquake record to more than 95%, the dynamic response error is less than 8%.