18 Feasibility of Using Fringe Projection System for Corrosion Monitoring in Metals of Interest in Cultural Heritage 113 Fig. 18.1 Picture of the optical fringe projection system Moreover, at the end of each immersion time, even the specimen images are acquired by a FP system. The optical measuring system is shown in Fig. 18.1. The set-up consists of a standard CCD camera with a resolution of 1392 1038 pixels, a frame rate of 17 fps, an ADC of 10 bit and an LCD projector with a resolution of 1024 768 pixels. Camera and projector are set in a classical triangulation scheme. The camera is equipped with optics with a focal ratio of 2.8 and a focal length of 50.2 mm. A frame allows to place the object so that it can be fully illuminated by the projector light. The fixed geometric parameters are the origin of the camera, the distance between the camera and the projector, equal to 530 mm, the angle between the camera and the projector, equal to about 53ı, and the distance between the camera and the object. The software tool used to analyze the pattern is implemented in Matlab, while a LabVIEW routine is used to drive the projector. The pattern adopted in this experiment is simply made of vertical stripes with a sinusoidal distribution of gray levels, according Eq. (18.2). I .x; y/ DA.x; y/ Csin.2 xf C'.x; y// (18.2) Where, I is the light intensity distribution, Arepresents the background illumination, f is projection frequency, ® is the phase and x in plane horizontal coordinate. The pattern was initially projected on the sample before starting the corrosion process. Successively at different stages during the corrosion test the sample was replaced in the holder and the fringe pattern was projected again on it. For each of the acquired image the 2D FFT was calculated to analyze the frequency content inside the image [41, 42]. Additionally, to check the presence of the corrosion feederate, the surface was examined by optical microscopy and its weight reduction was analyzed. 18.3 Results and Discussion The FFT of the recorded images respectively at the beginning of the test and after 20 and 110 h, are shown in Fig. 18.2. A detail of the frequency content is displayed as recorded after 20 h (a) and 110 h (b) in Fig. 18.3. The reduction of the contrastt can be seen through the decrease in the amplitude spectrum. This is associated with a reduction of the contrast due to the matting of the material, since the advancement of corrosion induces the formation of a patina. Still in Fig. 18.3, in the circles in red, it can be observed the presence of off-axis orders frequencies which were not initially present. This can be associated with the modulation of the projected fringes connected with the formation of the patina, in other words they can be view as indicators of occurring corrosive phenomena. In Fig. 18.4a the trend of the contrast in correspondence of different frequency orders is plotted. It can be observed that the carrier frequency corresponding to the projected grating decreases as a consequence of the matting of the surface; at the same time the contrast of off-axis orders frequency increase, as shown in Fig. 18.4b. Also by analyzing (Fig. 18.5) the difference of the FFT spectrum before and after the corrosion process it is possible to observe to presence of new spectral components. In order to support and confirm the study performed by FP at the end of the tests the surface morphology of sample was analyzed by OM. The surface morphology of bronze sample after 20 and 110 h of exposure to synthetic rain are shown in
RkJQdWJsaXNoZXIy MTMzNzEzMQ==