This and the companion paper present an analysis of the amplitude and time-dependent changes of the apparent frequency of a seven-story reinforced-concrete hotel building in Van Nuys, Calif. Data of recorded response to 12 earthquakes are used, representing very small, intermediate, and large excitations (peak ground velocity, vmax = 0.6 − 11, 23, and 57 cm/s, causing no minor and major damage). This paper presents a description of the building structure, foundation, and surrounding soil, the strong motion data used in the analysis, the soil-structure interaction model assumed, and results of Fourier analysis of the recorded response. The results show that the apparent frequency changes form one earthquake to another. The general trend is a reduction with increasing amplitudes of motion. The smallest values (measured during the damaging motions) are 0.4 and 0.5 Hz for the longitudinal and transverse directions. The largest values are 1.1 and 1.4 Hz, respectively, determined from response to ambient noise ...

In this, the second of a two-part paper, the analysis of the apparent frequency of a seven-story reinforced-concrete hotel building in Van Nuys, Calif., is extended to consider its time-dependent changes, both short and long term. The instantaneous apparent frequency is measured by two methods: windowed Fourier analysis and zero-crossings analysis. The results show that it changes from earthquake to earthquake and during a particular earthquake. The results also suggest “self healing” believed to result from settlement of the soil with time and dynamic compaction from aftershock shaking. Implications of such high variability of the system frequency on structural health monitoring, control of response, as well as on the design codes are discussed. Nonlinear response of the foundation soil acts as a sink of the incident seismic wave energy. It is suggested that it could be exploited in future designs to serve as a powerful and inexpensive energy-dissipation mechanism.

Wave travel times of vertically propagating waves are measured in the Van Nuys 7-storey hotel, located in the Los Angeles metropolitan area, from impulse response functions computed by deconvolution of the recorded earthquake response. East-West response data from 11 earthquakes over a period of 24 years are analysed. Changes in wave travel times are used to infer about local (between sensors) and global changes of structural stiffness, from one event to another, and with time during the earthquakes that damaged the building (San Fernando and Northridge). Also, wave travel times are used to estimate the fundamental fixed-base frequency of the building, f₁, which is compared with independent estimates of the soil–structure system frequency f_sys during the same earthquakes and during five ambient vibration tests. The analysis shows that monitoring only the changes of f_sys can be misleading for structural health monitoring and can lead to erroneous alarms, while monitoring changes of f₁ over suitably chosen time windows (before, during, and after excitation by strong earthquake motions) can be a powerful and robust tool for structural health monitoring. It is concluded that, under favourable conditions, this method can be used as a tool for global and local structural health monitoring. Copyright © 2007 John Wiley & Sons, Ltd.