Comprehensive Understanding of Near-field Acceleration Recordings by a New HHT Analysis

Colorado School Of Mines, Golden CO

Investigators

Abstract

CMS0085272 Title: "Comprehensive understanding of near-field acceleration recordings by a new HHT analysis" PI: R. Zhang Institution: Colorado School of Mines Abstract: Near-field ground acceleration recordings have revealed many vibration/wave characteristics that are uniquely different from the medium- to far-field ones. These include, among others, the low-frequency pulse-like wave signals and high-energy distribution in the beginning of the recordings. They are very useful primarily in the following two aspects: (1) As the output in modeling earthquake wave motions and/or explaining earthquake phenomena in the earth, they can help solve such seismological issues as source mechanism, directivity influence, and soil dynamic non-linearity. (2) As the input to geotechnical and structural engineering systems, they can be used to compute dynamic nonlinear responses and thus to evaluate seismic performance of systems. This analysis can quantify earthquake impact on various engineering systems designed on the basis of knowledge obtained form medium- to far-field ground motion records. However, before the 1999 Kocaeli (Turkey) and Chi-Chi (Taiwan) earthquakes, there existed only 8 ground motion recordings worldwide for earthquakes greater than a magnitude 7 and at a distance of less than 20 kilometers from the fault. The lack of data has restricted studies on the quantitative nature of near-field motion and its impact on engineering systems. In addition, conventional approaches (e.g., Fourier or Fourier-based wavelet analysis) supply either distorted or indirect information from the ground motion recordings. This might mislead, to a certain extent, the consequent use of the ground motion data for solving the aforementioned seismological and engineering issues. This study seeks to use a newly developed Hilbert-Huang Transform (HHT) to comprehensively analyze the near-field acceleration recordings of the Kocaeli and Chi-Chi earthquakes. In particular, the proposed study focuses on the following subjects: (1) singling out the undistorted low-frequency pulse-like wave signals from the original data and presenting a simple measure of them in terms of their peak and duration; (2) providing an accurate representation of time-frequency seismic energy distribution of the motion; and (3) revealing the commonality and difference in terms of vibration/wave characteristics of the motion among the near- and far-field recordings of each of the two earthquakes, among the near-field motions at observation sites with different orientations to the fault, and among the near-field motions of the two earthquakes compared with large-magnitude US earthquakes (e.g., 1992 Landers earthquake). A multi-disciplinary research team has been formed to perform the proposed project, consisting of the PI, and a senior structural engineering researcher (Erdal Safak) and a seismologist (Stephen Hartzell) at USGS at Denver, CO. The results from this study will be disseminated through the Internet at designated web pages related to these two earthquakes. Seismologists and engineers will be able to use the results for their further studies and applications in pertinent research frontiers. This study will not only help advance our knowledge of the earthquake source, but it will also improve the understanding of the damage potential of earthquake motion on various structures and thus enhance the contemporary seismic-resistance structural design.