Frequency Domain Corrections of Earthquake Accelerograms with Experimental Verifications

Strong-motion data engineering studies. It constitute is important the foundation that this data of earthquake be accurately recorded, processed and analyzed. In this report, an attempt is made to improve the current data processing method used to correct and filter the uncorrected accelerograms.

The work presented in this report was submitted by Omar Khemici as a dissertation to the Department of Civil Engineering of Stanford University in partial fulfillment of the requirements for the degree of Doctor of Philosophy. This report has three parts. The first one presents a comparative study between the time-domain correction method suggested in the Standard Data Processing of Accelerograms developed at Caltech (California Institute of Technology) about ten years ago, and its frequency domain equivalent. This equivalent method uses the frequency response functions obtained digitally from the Caltech routine. A new and efficient procedure called the Stanford Accelerogram Correction Procedure (SACP) is formulated in the second part. This method uses the fast convolution technique to perform the digital filtering of earthquake accelerograms. There are several advantages to this method. It is simple since all of the usual operations such as the instrument correction, the baseline correction, the low-pass filtering, the integration, etc., are executed with the use of a single filter for each of the three outputs, namely the corrected acceleration, velocity and displacement. The processing of the four earthquake records with the SACP permitted a saving of 37% of the CPU time with respect to the Caltech method. The Fourier amplitude spectra are directly calculated with the SACP. The corrected acceleration obtained with the SACP is higher than that obtained with the Caltech program. A discrepancy of 10% in the corrected acceleration was observed in one of the four earthquake records that has higher frequency content. However, the corrected velocity and displacement obtained with both methods are very close to each other. Finally, the last part consists of experimental testing using the shake table. Two different motions were simulated on the table a standard accelerograph and a beam interferometer. The integrated displacement showed excellent agreement with the measured displacement when an appropriate low cut-off frequency is and recorded simultaneously with high-precision displacement laser displacement from both methods applied to the filtering.