Abstract：The rotational components of strong ground motion near faults have differences in the response to the horizontal and vertical pendulums of seismograph, and according to this principle, a method for obtaining the rotational components of ground motion is established using the wavelet anlysis. The local time-frequency characteristics of uncorrected horizontal and vertical earthquake accelerograms using the discrete wavelet transform are discussed. Then a new threshold method, which processes the wavelet coefficients for horizontal earthquake accelerograms, is established to obtain the rotational components of ground motion. Finally, the correctness and precision of the method for obtaining the rotational components using the wavelet analysis is verified through shaking table tests. The results show that the calculated residual tilt displacement of rotational components of ground motion is consistent with the actual one. And the low-frequency part of the horizontal Fourier spectra of ground motion with the rotational component being filtered out is similar to the vertical ones. The correctness of the proposed method is verified through shaking table tests. Compared with the results processed by the Fourier analysis, the wavelet analysis can achieve more accurate results, and it is consistent with the actual situation.
[1] GRAIZER V M, KALKAN E. Prediction of spectral acceleration response ordinates based on PGA attenuation[J]. Earthquake Spectra, 2009, 25(1): 39-69. [2] GRAIZER V M. Tilts in strong ground motion[J]. Bulletin of the Seismological Society of American, 2006, 96: 2090-2102. [3] CHE Wei, LUO Qi-feng. Time-frequency response spectrum of rotational ground motion and its application[J]. Earthquake Science, 2010, 23(1): 71-77. [4] LEE V W, TRIFUNAC M D. Torsional accelerograms[J]. Soil Dynamics and Earthquake Engineering, 1985, 4: 132-142. [5] LI Hong-nan, SUN Li-ye, WANG Su-yan. Improved approach for obtaining rotational components of seismic motion[J]. Nuclear Engineering and Design, 2004, 232(2): 131-137. [6] GRAIZER V M. Effect of tilt on strong motion data processing[J]. Soil Dynamics and Earthquake Engineering, 2005, 25: 197-204. [7] TRIFUNAC M D, TODOROVSKA M I. Duration of strong motion during Northridge, California, earthquake of January 17,1994[J]. Soil Dynamics and Earthquake Engineering, 2012, 38: 119-127. [8] DONGELO D, SIMEONE V. Geomorphometric analysis based on discrete wavelet transform[J]. Environmental Earth Science, 2014, 71(7): 3095-3108. [9] NAGA P, EATHERTON M R. Analyzing the effect of moving response on seismic response of structures using wavelet taransform[J]. Earthquake Engineering & Structual Dynamics, 2014, 43(5): 759-768. [10] ANSARI A, NOORZAD A, ZAFARANI H, et al. Correction of highly noisy strong motion records using a modified wavelet de-nosing method[J]. Soil Dynamics and Earthquake Engineering, 2011, 30(11): 1168-1181. [11] 张郁山, 张凤仙. 基于小波函数地震动反应谱拟合法[J].土木工程学报, 2014, 47(1): 70-81. (ZHANG Yu-shan, ZHANG Feng-xian. Matching method of ground-motion response spectrum based on the wavelet function[J]. China Civil Engineering Journal, 2014, 47(1): 70-81. (in Chinese)) [12] 郭 杨, 倪煌俊, 柯宅邦. 应用小波分析法检测预应力管桩裂缝的研究与实践[J]. 岩土工程学报, 2013, 35: 1224-1227. (GUO Yang, NI Huang-jun, KE Zhai-bang. Practical research on detecting cracks in pre-stressed pipe piles by wavelet analysis[J]. Chinese Journal of Geotechnical Engineering, 2013, 35(1): 1224-1227. (in Chinese)) [13] BASU D, WHITTAKER A S, CONSTANTINOU M C. Extracting rotational components of earthquake ground motion using data recorded at multiple stations[J]. Earthquake Engineering & Structual Dynamics, 2013, 42(3): 451-468.
