Experimental Investigation of System-Identification-Based Damage Assessment on Structures

In this paper, damage assessment algorithms based on the system identification (SI) method are examined in lab experiments. The SI method identifies structural parameters by finite-element method by minimizing the error between measured and analytically computed responses. A regularization scheme is applied to alleviate the ill-posedness of an inverse problem by adding a regularization function to the primary error function. Two algorithms were developed, that are dependent on the type of measured response, to assess damage. Static displacements from static loading and modal data from impact vibration were measured in lab experiments on a grid-type model bridge. The baseline structural model and baseline properties are determined with experimental data obtained from the tests on the undamaged model structure. Damage is simulated by saw cutting the cross section at various depths and is identified as the reduction in the structural stiffness of the elements around the crack. Through the experimental work, the applicability of the SI-based damage assessment algorithms were rigorously investigated.