A new modal-based damage location indicator G.R. Gillich 1 , Z.I. Praisach 1 , M. Abdel Wahab 2 , H. Furdui 1 1 Eftimie Murgu University of Resita, Department of Mechanical Engineering P-ta Traian Vuia 1-4, 320085, Resita, Romania e-mail: gr.gillich@uem.ro 2 Department of Mechanical Construction and Production, Faculty of Engineering and Architecture, Ghent University, Technologiepark Zwijnaarde 903, B-9052 Zwijnaarde, Belgium Abstract Vibration-based damage detection techniques use the change in modal data as an indicator to assess damages in the structure. Knowing the structural dynamic characteristics of the healthy and damaged structure, the estimation of the damage location and severity is possible by solving an inverse problem. This paper presents a mathematical expression relating damage location and depth to the frequency shifts of the bending vibration modes. This expression permits the extraction of a series of coefficients that characterize each damage location and are independent of the damage severity. The vector aggregating these coefficients for a given location constitutes a Damage Location Indicator (DLI) that unambiguously characterizes the position of a geometrical discontinuity in the beam. A set of vectors typifying all locations along the beam may be used as patters opposable to the damage signature found by measurements. The similarity between the signature and one of the patterns indicates the location of damage. 1 Introduction Monitoring of structures to assess their integrity has been one of the most important issues in civil and mechanical engineering in the last years. Numerous non-destructive techniques are now available, permitting to achieve one of the following damage assessment levels: a) detection, b) localization and c) severity. The vibration-based methods associate damage with changes in the dynamic response of structures, which occur due to the decreased beam capacity to store energy. The link between a damage and the changes in the structural response is described by mathematical expressions reflecting various models, from analytical beam models to statistical models predicting the beam dynamic behavior. Often the response changes refer to the natural frequencies, but other modal parameters can also be involved in damage detection algorithms. Extensive literature reviews on this topic are presented by Doebling et al. [1] and Sohn et al. [2]. They focused on methods and data required for damage identification by examining the changes in various types of measured structure responses. The use of inverse methods in damage detection using measured vibration data is presented by Friswell [3], who made a critical review on problems occurring with this approach, including modeling errors and environmental effects. Methods destined to predict the damage location and evaluate its severity are usually model-based. Various damage modeling approaches are presented in literature; see for instance Fritzen [4], Friswell [5], Dimarogonas [6] and Ostachowicz and Krawczuk [7] and Christides and Barr [8]. All these approaches can be classified into three main categories; nalely local stiffness reduction, discrete spring models, and complex models in two or three dimensions. The discontinuity is often considered as an open crack, in order to neglect the nonlinear effects due to a crack closure. Other authors like Yan et al. [9] considered a bilinear behavior of closing cracks. However, no models precisely predicting frequency changes due to discontinuities for a large number of vibration modes are available. Furthermore, the use of existing