Output-only identification for linear structures has known a growing interest, mainly due to the advantage of employing the straight available ambient excitations (wind, traffic) and accurately measuring the corresponding responses: in this way the typical drawbacks of input-output identification tests associated with the application of a controlled excitation on large structures, are avoided; meanwhile, the monitoring of structures in their operative conditions is guaranteed. Relying only on the output, however, implies important limitations, first of all the assumption of a Gaussian white noise input. Among the various output-only identification algorithms, the Stochastic Subspace Identification (SSI) is of particular interest as it involves a reduced computational effort compared to the competing ones. The objective of this paper is to evaluate the effectiveness of the SSI with the use of seismic responses, which are shorter, more intense and have a significantly different frequency content than the commonly employed ambient vibrations. The applicability of the technique to bridges isolated with High Damping Rubber Bearings (HDRB) is also investigated. Modal parameters (frequencies, damping ratios) of a physical model tested on shaking table are identified by comparing four different weighting matrices defined in literature for SSI.
Per scaricare il contenuto integrale è necessario essere Socio Anidis, scopri come associarti
News Vedi tutte