FDD modal identification from earthquake response data with evaluation of Soil-Structure Interaction effects

“Non-parametric” system identification via a refined FDD (rFDD) algorithm is developed to evaluate current modal properties of civil engineering buildings subjected to earthquake excitation. Innovatively, the algorithm operates in an OMAX (Operational Modal Analysis with eXogenous input) environment on base-excited buildings, to detect also potential Soil-Structure Interaction (SSI) effects. At first, the identification technique is explored analytically, dealing with a variable stiffness and damping foundation system. The OMAX condition comes from the use of base excitation records, which are employed as exogenous input. The standard fixed-base condition may simulate well the response to ambient loading or weak seismic excitations; rather, the flexible-base model shall deal with more realistic seismic responses of buildings, which may embed interactions between soil and structure. The developed analyses are performed on synthetic response signals, which are computed from a benchmark structure. Simulated signals are generated prior to dynamic identification, by adopting a given seismic input with variable foundation properties. The method demonstrates its full effectiveness, also for the detection of close modes and for the identification of high modal damping ratios. Issues pertaining to earthquake record processing are specifically addressed. The present work proves a necessary condition for the effectiveness of the present rFDD algorithm as a robust method to inspect current flexible-base strong ground motion modal parameters. This shall help in identifying possible variations of structural features and SSI effects along experienced seismic histories, and providing an effective tool towards Earthquake Engineering and Structural Health Monitoring purposes.