Detection of Local Surface Defects in Metal Structures by the Method of Correlation Signal Processing
- PDF / 468,419 Bytes
- 7 Pages / 594 x 792 pts Page_size
- 59 Downloads / 227 Views
DETECTION OF LOCAL SURFACE DEFECTS IN METAL STRUCTURES BY THE METHOD OF CORRELATION SIGNAL PROCESSING Ya. P. Kulynych1, 2 and I. I. Tryhub1
UDC 537.874:621.317
To localize local subsurface defects in metal structures, we propose a method of correlation signal processing. Numerical and experimental results are presented to confirm the efficiency and reliability of the method. Keywords: eddy-current testing, defect, correlation method, inverse problem.
To detect defects in products made of conducting materials by using nondestructive eddy-current testing methods, it is necessary to interpret the parameters of the information signal. Since the information about the structure of the monitored product and its defects is contained in the measured characteristics of scattered electromagnetic fields, it is important to solve the inverse problem, i.e., to determine the location of defects and their geometric characteristics. The theory of nondestructive eddy-current testing contains numerous methods aimed at the solution of this inverse problem [1–3]. These methods can be split into two groups. In the first group, the solution of the inverse problem is reduced to the problem of image recognition [4, 5]. The measured signal is identified as a signal from one of the known types of signals obtained for defects either experimentally or by using theoretical models. A specific feature of these methods (and also their main drawback) is the necessity to have a set of sampling signals for a series of defects of various types, which is used to train an appropriate system of automatic recognition. If the set of signals is not sufficiently large, then the accuracy of classification of signals can be quite low. The other group of methods uses the repeatedly applied mathematical model of the direct problem, with variable geometric and electrophysical parameters of the defect until the norm of the difference between the signals obtained in the model and experimentally is minimized [6, 7]. The efficiency of this approach is determined by three factors: the theoretical model of the signal, the choice of the initial value for the approximation of the solution, and the algorithms of finding the global minimum. In this case, it is extremely important to choose an appropriate mathematical model for the description of the signal. Only models with minimal simplifications in the description of the physical processes guarantee acceptable results. However, this requires significant computation resources.
The second group also includes the correlation method of localization of the elongated subsurface defects [8, 9]. It is regarded as the first stage of the solution of the inverse problem of nondestructive eddy-current testing. The aim of the present work is to apply this method to the problem of determination of the location of a subsurface “local” defect, i.e., a defect whose electromagnetic field approximately coincides with the electromagnetic field of a point source (equivalent electric or magnetic dipole [10, 11]). The point of location of th
Data Loading...
