Vibration based stiffness reconstruction of beams and frames by observing their rotations under harmonic excitations - Numerical analysis

Recent progress in data acquisition technology makes it possible to measure not only translational but also angular vibrations of beam and frame structures. This raises a question whether these new methods can effectively be applied in damage structural detection and localization. For this purpose a damage detection method aiming at reconstructing the stiffness distributions of a structure from its vibration measurements is tested when rotational degrees of freedom are added to the dynamic model of the structure. The stiffness reconstruction is formulated as a minimization problem in terms of harmonic vibrations of the structure and its finite element model. Two examples of a beam and frame structure are analyzed in detail. To better test the damage detection algorithm, measuring noise with various levels has been added to the analyzed signal. The results of the numerical analyses show that the application of genetic algorithms and a Levenberg-Marquadt local search appeared to be even more effective in damage detection when the angular amplitudes of harmonic vibrations are acquired. This creates good prospects for the future applications of angular accelerometers in structural health monitoring of civil engineering structures. (C) 2009 Elsevier Ltd. All rights reserved.