Poster #225, Seismology

Time-Varying Shear-Wave Velocities in a High-Rise During the 2019 M6.4 and M7.1 Ridgecrest Earthquakes from Wavefield Interferometry

Monica D. Kohler, & German A. Prieto
Poster Image: 

Poster Presentation

2021 SCEC Annual Meeting, Poster #225, SCEC Contribution #11507 VIEW PDF
Seismic interferometric techniques for structural health monitoring are applied to data from Community Seismic Network MEMS accelerometers permanently installed on nearly every floor of a 52-story steel moment-and-braced frame building in downtown Los Angeles. Wavefield data from the instrumented high-rise from before and during the 2019 M6.4 and M7.1 Ridgecrest, California earthquakes are processed for impulse response functions. The concept of impulse response functions is generalized here to show that a building’s nonlinear response can be monitored and quantified through time-varying measurements of representative pseudo-linear systems in the time domain. The building was not damaged this earthquake, but temporary nonlinear behaver observed during the strong motions provides a unique opportunity to test this method’s ability to map time-varying properties. Shear-wave velocities in the building’s east-west direction are found from linear regression to the impulse response function’s first, direct arrival of energy and are compared with a consistently-observed average shear-wave velocity of 225 m/s from ambient vibration time periods. The broadband velocities are reduced by as much as 10% during building shaking, and their restoration to pre-earthquake levels is found to be a function of the decrease in shaking amplitude levels over a broad frequency band. These observations are all made over time scales of seconds, throughout the entire duration of building shaking. After cross correlation or deconvolution, the response of the building that has been extracted is independent from the ground coupling, soil-structure interaction, and complex earthquake source and wave path signatures. The benefits of this approach to structural health monitoring are that: i) it does not depend on modal identification, structure type or geometry, structural materials, or similar a priori assumptions, ii) it can map nonlinear response characterization on small spatial scales, iii) it comprises an automated near-real-time system identification and damage detection method that can potentially be applied to the rapidly growing field of permanent, continuously-recording, sensor networks.