PSI - Issue 32

O.V. Bocharova et al. / Procedia Structural Integrity 32 (2021) 299–305 Bocharova O.V., Andzjikovich I.E., Sedov A.V., Kalinchuk V.V. / Str ctural Integrity Procedia 00 (2021) 000 – 000

304 6

Fig. 5. Location of images in the space of recognition for samples with various types of defects (accelerometer 2).

Fig. 4, 5 show the results of construction of the specific diagnostic space and location of the images for samples with various types of defects obtained on the basis of measurements of accelerometer 1 and accelerometer 2 respectively. In Fig. 4, 5 spectral feature 1 and spectral feature 2 are the coefficients in decomposition of vectors  R N i f in the selected basis. Inside the clusters there are images of the response functions for various samples: cluster 1 corresponds to the sample without defects, cluster 2 corresponds to the sample with slot; cluster 3 corresponds to the sample with embedded steel plate; cluster 4 corresponds to the sample with two embedded steel plate. Three measurements were made for the reliability of the measurements for each type of the defect. A series of measurements demonstrates a sufficiently high degree of repeatability. The results of the conducted experiments demonstrate that there is a precise spatial distribution of images in the diagnostic space depending on the type of defect. The results of the experiments showed that both the transmitted and the reflected wave field are informative in the recognition of plane defects. The best results of defect recognition have been achieved when measuring the transmitted wave field. Use the proposed method provides a clear recognition of the presence of a defect, its type and linear size in the diagnostic space of images. It should be noted that when measuring transmitted wave, the first spectral feature reacts to the presence of a defect and its type (inclusion or slot). The first spectral feature reacts weakly to the linear size of the defect. The second spectral feature reacts to the presence of a defect, its type and linear size. When measuring the reflected wave, the first and second spectral features respond to all characteristics of the defect, but to a lesser extent to its type. Spectral features react rather strongly to the presence of a defect and its linear size. Obtained results confirm the effectiveness of the proposed method for processing signals in diagnostics of inhomogeneities. 4. Conclusion An approach for detecting hidden inhomogeneities and determining the type of inhomogeneity in materials, based on monitoring changes in the parameters of surface wave fields, is proposed and implemented. To increase the informativeness of the wave field, an original method has been developed based on the use of optimal orthogonal decompositions of signals in an adaptively tunable basis. The results of the experiments showed that when recognizing the presence of a plane inhomogeneity and determining its type, the transmitted wave is more informative, that must be taken into account when choosing the locations of the sensors.