- Author
- Glaser, S. D. | Leeds, A. L.
- Title
- Estimation of System Damping at the Lotung Site by Application of System Identification.
- Coporate
- Colorado School of Mines, Golden, CO
- Sponsor
- National Institute of Standards and Technology, Gaithersburg, MD National Science Foundation, Washington, DC
- Report
- NIST GCR 96-700, August 1996, 195 p.
- Distribution
- Available from National Technical Information Service
- Contract
- NIST-GRANT-60NANB5D0074 NSF-AWARD-CMS94-57268
- Keywords
- system identification | earthquakes | system damping | seismic design | clay | sand
- Identifiers
- Lotung Site, Taiwan; procedures for system identification; parametric modeling of Lotung data; procedures for data processing; MATLAB procedure to integrate acceleration records to velocity and displacement; pore water pressure time histories for events 12, 16, 17; review of damping values measured in the field and laboratory; comparison of actual interval outputs to model interval outputs for all analyzed events
- Abstract
- Possibly the best set of data for earthquake excitation of soils exists for the test site operated by the Taiwan Power Company in conjunction with the Electric Power Research Institute (EPRI) at Lotung, Taiwan. At this site, two locations are instrumented with three-component accelerometers at depths of 47 m, 17 m, 11 m, 6 m, and at the surface. One array is in the free-field, while the other is adjacent to a one-quarter scale nuclear containment vessel. The site is also well instrumented with piezometers at various depths and locations. The simplified soil profile consists of 30 m to 35 m of silty sand and sandy silt with some gravel, overlaying a thick clay and silt deposit. The water table is within half a meter of the ground surface. This area is seismically active, and strong shaking generated by many earthquakes exhibiting a wide range of magnitudes have been recorded since 1986. For this study, the modal frequencies and damping ratios were calculated for events 3,4,7,8,9,10,12 and 16 with local magnitudes ranging from 4.5 to 7.0. The modal frequencies and damping ratios calculated are examined for the effect of local energy intensity and soil structure interaction. Modal frequencies are seen to decrease with increasing intensity once a certain threshold of acceleration/intensity is reached. This result is consistent with the data obtained by other authors using different techniques. For the 0-6 m interval the decrease in frequency with event energy is less pronounced under a model containment structure than in the free field. This soil-structure effect is increasingly diminished with depth and absent by the 17-47 m interval. Calculated damping values demonstrate an expected increase with input seismic energy. For the 0-6 m and 6-11 m intervals the damping values are higher under the model structure than in the free field. This distinction is completely missing in the 17-47 m results. The transition to non-linear behavior, while less pronounced with increasing depth, consistently occurs above a peak acceleration of 0.05 g or Arias Intensity of 100 m/sec. The results clearly indicate a degree of non-linear response over the intervals studied. Evidence of a decrease in specific interval fundamental frequency and an accompanying general trend of increased damping with higher seismic energy are clear. Comparison of the results of this study with previous work considered with the inherent superiority of parametric modeling for transient and/or non-stationary time series such as earthquakes indicate that system identification is a more robust method for identifying fundamental frequencies and damping values for layers of earth materials when borehole information is available.