Soil–structure interaction effects estimated with recursive subspace estimation methods

Abstract

Subspace identification methods are used to monitor the modal properties of buildings on soft soils where soil–structure interaction effects are present. Two buildings located in Mexico are analyzed: the PC building in Mexico City and the SIS building in Acapulco City. Despite differences in signal quality, both instrumentations enable the analysis of soil–structure interaction (SSI) effects due to the accelerometer’s disposition along the soil, base, and stories of the buildings. A methodology that combines the decomposition of the recorded motions and the recursive 4SID estimation method to estimate the fundamental frequencies of the soil–structure and structure systems and the frequencies related to horizontal translation and rocking movements is proposed. These frequencies are compared with those obtained from spectral and modal minimization analyses. The parametric estimation uses seismic and ambient vibration records, the last available only in the SIS building. The results consistently estimate the fundamental frequencies of the soil–structure, structure, and rocking systems, with less variation when employing the R4SID method. The translation frequency presents the highest variation but the lowest participation in the system response. The soil–structure and structure frequencies in ambient vibration records show similar values; the interaction is not observable. However, the structure frequency can be recovered using the soil–structure, rocking, and translation frequency estimates, reporting values similar to those based on seismic events. With reliable estimation results and minimal computational processing cost, the proposed recursive 4SID-based method is suitable for its incorporation in structural health monitoring systems, considering SSI effects.