204 G. Busca et al. Fig. 22.8 Curvature estimation using Eq. (22.6): (a) beam1; (b) beam2; (c) beam3 22.9 Conclusions In this work the possible application of cameras to damage identification has been explored. The main advantages of cameras is that it is possible to estimate the structural dynamic behavior on a continuous field by one single device at a reasonable price and without any contact with the structure. The contactless characteristic is relevant, particularly for slender structures, where the mass and the damping added to the system by traditional accelerometers may dramatically change the structure behavior. In the actual paper, mode shapes are identified by analyzing the acquired images and then damage detection estimation is being performed relying on the difference between the curvature mode shapes of not damaged and damaged beams. Results show that the damage is correctly located (peak of 4curvature) for all the three beams only in the case of severe damage (i.e. slot deep equal to 30 % of the width). To further develop the proposed technique in order to obtain more reliable and effective damage detection, developments on both hardware for image acquisition (mainly in terms of resolution and acquisition frequency) and image processing are needed. As for the image processing, it is expected that the use of digital image correlation can reduce the motion detection uncertainty; however this requires a beam surface preparation in order to produce a very fine surface pattern.
RkJQdWJsaXNoZXIy MTMzNzEzMQ==