Reliability of damping ratios inferred from the seismic response of buildings

Abstract This paper studies the reliability of damping ratios inferred from the seismic response of buildings employing a modal minimization parametric system identification technique. The reliability of inferred damping ratios is quantified through the use of three metrics: (1) dynamic amplification factors; (2) Arias modal contribution factors; and (3) reliability intervals. Dynamic amplification factors measure the amplification of 1% damped floor acceleration spectral ordinates of the recorded response at the roof in the vicinity of identified modal periods with respect to those of the motion recorded at the base of the building. It is shown that this metric allows to determine if the earthquake produced a minimum dynamic amplification of the first and second modes of the structure, to reliably identify their damping ratios. The Arias modal contribution provides a measure of the intensity of a computed modal response relative to the total predicted structural response. It is argued that only damping ratios from modes with a minimum contribution to the seismic response may be reliably estimated. The third metric corresponds to the reliability intervals of the identified damping ratios, which measure the sensitivity of the objective function to small variations in the damping ratio. It is shown that the objective function is typically significantly less sensitive to variations in damping ratios than to similar variations in modal periods, which leads to larger variability in the results for damping with respect to that obtained in modal periods. Reliability screening tests are developed based on establishing limiting values for these three metrics. It is shown that these screening tests are capable of discriminating inferred damping ratios that are reliable from those that are not, independently from the adopted identification technique, and also reducing the variability of the inferred results.