The paper is concentrated on a non-traditional application of the Kalman Filter algorithm and outlines a procedure that can be used to assess structural damage. The technique is essentially based on the estimate of unknown stiffness parameters of the structure, whose changes over time can be related to the structure’s deterioration and local damage. In this way, suspected areas can be detected and further investigated. In our approach, it is assumed that experimental records are taken at regular intervals by means of shaking tests and/or during micro-seismic activity, so that optimal values of convenient parameters can be determined by solving an inverse problem. The proposed procedure has been applied to the estimate of bulk and shear modules, whose changes denote possible damage due to isotropic and deviatoric stress components, in accordance with the theoretical framework which several damage models are based on. More specifically, we simulated experimental tests on dam structures by assuming some damage at certain locations. Finally, an error indicator was introduced, with the aim of finding an objective measure of the accuracy of the results obtained by solving the inverse problem.
