| description abstract | This paper analyzed damping ratios inferred from 1,335 seismic responses recorded in 154 instrumented buildings in California. These values were inferred using a parametric system identification technique in the time domain, and subjected to a series of reliability screening tests to retain only high-quality data. The resulting damping ratios conformed a data set of 1,037 high-quality values inferred exclusively from the seismic response of buildings, a database several times larger than previous studies of damping inferred from seismic response. The data set was analyzed using a linear mixed-effects statistical model to account for the fact that many of the data points were clustered, because they came from damping ratios in the same building shaken by various earthquakes. It was shown that damping decreases with increasing building height, which is the factor that best explained the relatively large variance observed in the data. Contrary to some previous recommendations, it was found that once the variation with height is taken into account, the primary structural building material is not statistically significant in the damping ratio of buildings subjected to earthquakes. However, when including the combined material and lateral resistant system as a factor in the statistical model, an additional 6% of the variance was explained. Results showed that steel buildings with moment-resistant frames have, on average, a slightly higher damping ratio than those with steel braced frames. The amplitude dependency of damping showed that there was no significant correlation between damping ratio and the overall lateral deformation demand in the building as measured by the peak roof drift ratio for amplitudes typically observed during moderate earthquake motions once a minimum level of amplitude is exceeded. | |
