Substructural damage detection using frequency response function based inverse dynamic substructuring

In this paper, a substructural damage detection approach is presented using frequency response function (FRF)-based inverse dynamic substructuring and FRF-based model updating. In practice, often only one subsystem of the global structure is critical and susceptible to damage and therefore needs monitoring. In the proposed method, one subsystem is the "main" subsystem and is susceptible to damage and the "residual" subsystem(s) are considered undamaged and the damage identification is only applied to the main subsystem. The FRF matrix of the main subsystem is obtained as a "standalone" component that is completely decoupled from the residual subsystem(s) without the need of interface virtual support or identifying interface forces. The FRF measurement is performed on the global damaged structure but the FRFs of the main substructure (damaged) is obtained by "decoupling" the known FRF(s) of the residual subsystem from the global system FRFs and the damage detection is performed only on the main substructure. An FRF-based model updating method using numerical sensitivities is then used for damage identification of the main subsystem. The FRF obtained from the finite element model of the main substructure is updated using the FRF of the damaged main substructure (obtained from decoupling) and the location and quantity of the damage is identified. Numerical and experimental examples are presented to illustrate the damage detection procedure and the damage in the main substructure is detected, located and quantified with good accuracy.