Quantification of hidden corrosion in aircraft structures using enhanced D-Sight NDT technique

Automated Identification of Hidden Corrosion Based on the D-Sight Technique: A Case Study on a Military Helicopter

“…As discussed above, because of the small magnitude of deformation resulting from the pillowing effect, hidden corrosion is difficult to detect with numerous NDT techniques, especially in the early stages of its development. The results of previous studies [ 28 ] have shown that the lowest detectable magnitude is at the level of 30 μm. This result was obtained using the D-Sight NDT technique, which demonstrated a high sensitivity to such deformations.…”

“…Inspections using D-Sight techniques have been implemented with hardware created by Diffracto Ltd. in a system known as DAIS. The principle of operation for this testing device is based on the above overview of the D-Sight testing approach and can be found in numerous previous publications; see, e.g., [ 27 , 28 ]. However, during inspections, numerous factors influencing the quality of the resulting D-Sight images need to be taken into consideration, such as the reflectivity of the tested surface and the position of the testing device, which has a direct influence on the angle of observation and illumination.…”

“…The current study focuses on the improvement of the D-Sight technique for the purpose of quantitatively evaluating D-Sight images and automating the evaluation process based on real aircraft structures after successfully testing the developed approach on specimens with artificially introduced deformations that simulate hidden corrosion [ 28 ]. For this purpose, historical data from periodic inspections of the Mi family of military helicopters (see Figure 3 ), which are a part of the arsenal of the Polish Armed Forces, were considered as a case study.…”

“…The calculations were performed on a Windows 10 laptop equipped with an Intel ® Core™ i7 quad-core processor and 16 GB of RAM. The preprocessing algorithm, developed previously and in [ 28 ], consists of three main steps: image alignment, orthonormalization, and illumination equalization. The parameters of the algorithm can be found in [ 28 ].…”

“…In [ 27 ], the authors proposed a method of image processing that includes procedures to reduce the influence of the angle of observation and non-uniform illumination and detect corroded areas, and it included the first attempts at their quantification. Furthermore, laboratory tests were performed on specimens with simulated hidden corrosion to evaluate the sensitivity of the D-Sight technique and find a correlation between the true dimensions of the corrosion spots obtained using reference methods, both planar and in the direction normal to the surface of the tested structures, as well as those estimated based on image processing of D-Sight images [ 28 ].…”