Blind identification of soil–structure systems

Surrounding soil can drastically influence the dynamic response of buildings during strong ground shaking. Soil’s flexibility decreases the natural frequencies of the system; and in most cases, soil provides additional damping due to material hysteresis and radiation. The additional damping forces, which are in non-classical form, render the mode shapes of the soil–structure system complex-valued. The response of a soil-foundation system can be compactly represented through impedance functions that have real and imaginary parts representing the stiffness and damping of the system, respectively. These impedance functions are frequency-dependent, and their determination for different configurations been the subject of a considerable number of analytical, numerical, and experimental studies. In this paper, we first develop a new identification technique that is capable of extracting complex mode shapes from the recorded free or ambient vibrations of a system. This technique is an extension of the second-order blind identification (SOBI) method, which is fairly well established in a number of other areas including sound separation, image processing, and mechanical system identification. The relative ease of implementation of this output-only identification technique has been the primary source of its appeal. We assess the accuracy and the utility of this extended SOBI technique by applying it to both synthetic and experimental data. We also present a secondary procedure, through which the frequency-dependent soil-foundation impedance functions can be easily extracted. The said procedure has a practical appeal as it uses only free or ambient responses of the structure to extract the foundation impedance functions, whereas current techniques require expensive and time-consuming forced-vibration tests.

► A new technique is presented for identification of non-classically damped systems.
► The proposed method is based on a second-order blind identification (SOBI).
► The veracity of the proposed technique is evaluated through both numerical and experimental examples.
► A new approach is presented to determine frequency-dependent soil-foundation impedance functions.