Damage Detection of a Prestressed Concrete Beam Using Modal Strains

Abstract

Different methods are proposed in literature using experimental modal information to detect possible damage. In this paper a finite-element (FE) model updating technique is applied. The unknown properties of a FE model are adapted, such that the differences between experimental modal data (modal curvature in combination with eigenfrequencies and mode shapes) and the corresponding analytical predictions are minimized. An iterative sensitivity based algorithm is used for solving this optimization problem. The method is applied to the damage assessment of a gradually damaged prestressed concrete beam. It is assumed, that damage can be characterized by reducing the bending stiffness. The main focus of this paper is to analyze the influence of using modal curvatures. In contrast to eigenfrequencies and mode shapes, modal curvatures are very sensitive to local changes of the bending stiffness nearby the sensor location, but insensitive to local changes far from the measurement location.