Institute of Fundamental Technological Research

This paper describes an effective method of substructure isolation for local structural health monitoring (SHM). In practice, often only a small part of a larger structure is critical and needs monitoring. However, typical low-frequency SHM methods require modeling and analysis of the global structure, which can be costly, time-consuming and error-prone. The proposed approach is based on the virtual distortion method (VDM) and uses force distortions to model fixed supports in the boundary nodes to isolate the considered substructure from influences of the rest of the structure. Therefore, given an excitation of the substructure and the measured response of the global structure, the response of the substructure treated as fixed supported can be computed. Local-only monitoring is then possible using virtually any of the existing methods. However, consistently with the isolation methodology, strain distortions are used here for modeling of damages of the isolated substructure. The discrete adjoint variable method is used for the first time within the framework of the VDM in order to perform fast analytical sensitivity analysis and improve the computational effectiveness of the damage identification by one order of magnitude. A numerical example of a frame-truss with 5 and 10% noise level and an experiment of a cantilever beam are presented to validate the isolation methodology.

substructure isolation, damage identification, virtual distortion method (VDM), structural health monitoring, adjoint variable method