加振実験による地震計記録の長周期成分の信頼度の照査

Abstract

In this paper, the performance of three different kinds of seismographs, that of, aSMAC-B2 accelerograph, a SAMTAC-17E which is a digital servo-type accelerograph, and aJMA low-magnification strong-motion seismograph, were examined from a shaking table testwith the differential transformer-type and winding-type (DP-500 B) displacement meters, as a means to disclose inadequancy of long-period (2-20 sec) component of conventionalacclerograms. A skaking table was driven by input of JMA records observed during theNihonkai-chubu earthquake of 1983. The analyzable ranges in terms of amplitude and periodof the records given by these instruments were discussed by using performance criteria suchas waveform, cross correlation, coherence function, Fourier spectrum and energy, etc. Theresults obtained indicate that: (1) JMA seismograms without saturation were largelysatisfactory in the period over about 5 sec, (2) The recording accuracy of SAMTAC-17Einstrument was found insufficient over the period range from 5 to 6 sec, resulting in quitedifferent displacement waveform, and (3) The observation capacity in view of signal to noiseratio of long-period component of SMAC records is not very high, resulting in the largervalue than record amplitude obtained by strong motion seismometer of displacement-type.