AUTOMATION OF BEARING PARTS’ GRINDING SURFACE DEFECT ETECTION AT VORTEX-CURRENT TESTING WITH SUBSTANTIATION OF WAVELET TYPE CHOICE 

Shumarova Ol'ga Sergeevna, Postgraduate student, Saratov State Technical University named after Y. A. Gagarin (77 Politekhnicheskaya street, Saratov, Russia), shumarova.olga@mail.ru
Ignat'ev Stanislav Aleksandrovich, Doctor of engineering sciences, professor, sub-department of automation, control, mechatronics, Saratov State Technical University named after Y.A. Gagarin (77 Politekhnicheskaya street, Saratov, Russia), atp@sstu.ru
Samoylova Elena Mikhaylovna, Candidate of engineering sciences, associate professor, sub-department of automation, control, mechatronics, Saratov State Technical University named after Y. A. Gagarin (77 Politekhnicheskaya street, Saratov, Russia), atp@sstu.ru

681.5

Background. The purpose of this work is to improve a method of information signal processing at vortex-current quality control of bearing parts’ grinding surfaces on the basis of automatic recognition of local defects by means of an optimal choice of a wavelet type. According to the goal the work solves the following tasks: development of a method of automatic recognition of a local defect type and creation of a program module for automatic search and intellectual algorithms of recognition of the most widespread local defects of bearing parts’ grinding surface according to vortex-current testing with substantiation of a wavelet choice criterion, consisting in defining a ratio of wavelet coefficients of defect and noise, criteria-estimated by the size of the mean square deviation increased by a coefficient of limiting sensitivity. Herewith, at the first stage it is necessary to recognize information signals by an amplitude component. If there is a matchof the set level of a corridor of a mean square deviation from defects in one feature space, at the second stage the recognition will be carried out by the level of a corridor of a mean square deviation of a phase component of information signals.
Materials and methods. The conducted research was based on the methods of discrete signals processing, the device of wavelet-transformations, the theory of recognition of images. Processing of results of vortex-current testing was conducted using the specially developed software written and realized by means of the Delphi programming language. Pilot studies were conducted using the automated system of vortex-current testing (ASVT) on the basis of automatic devices of bearing parts testing (ABK-P2 and PVK-K2M) under production conditions of JSC EPK-Saratov.
Results. The authors achieved an increase of efficiency of the vortex-current method of quality control of bearing parts’ grinding surface through the automatic analysis of the degree of heterogeneity of an upper layer and the detection of typical defects by means of special methods of the wavelet-analysis of information signals, and also substantiation of the wavelet type choice allowing to increase the quality of testing results.
Conclusions. The authors developed the special software product allowing to analyze the discrete data obtained from a vortex-current sensor for recognition of defects. Introduction of the software into the automatic system of quality monitoring of precision products manufacturing allows to efficiently control the technological process. The developed software product is under investigation and after some adjustment to concrete technological processes is subject to practical implementation.

vortex-current testing, local defects of an upper layer of bearing parts, wavelet-transformation, monitoring.

1. Gludkin O. P., Gorbunov N. M., Gurin A. I., Zorin Yu. V. Vseobshchee upravlenie kachestvom [General quality management]. Moscow: Radio i svyaz', 1999, 600 p.
2. Ignat'ev S. A., Gorbunov V. V., Ignat'ev A. A. Monitoring tekhnologicheskogo protsesa kak element sistemy upravleniya kachestvom produktsii [Technological process monitoring as an element of product quality management system]. Saratov: SGTU, 2009, 160 p.
3. Kulakov Yu. M., Khrumkov V. A., Dunin-Barkovskiy I. V. Predotvrashchenie defektov pri shlifovanii [Grinding defects prevention]. Moscow: Mashinostroenie, 1975, 144 p.
4. Shumarova O. S., Ignat'ev S. A., Samoylova E. M. Vestnik Saratovskogo gosudar-stvennogo tekhnicheskogo universiteta [Bulletin of Saratov State Technical University]. 2014, no. 2, pp. 192–195.
5. Yaroshek A. D., Bystrushkin G. S., Pavlov B. M. Tokovikhrevoy kontrol' detaley mashin [Vortex-current testing of machine parts]. Kiev: Naukova dumka, 1976, 124 p.
6. Dorofeev A. L., Kazamanov Yu. G. Elektromagnitnaya defektoskopiya [Electromagnetic testing]. Moscow: Mashinostroenie, 1980, 280 p.
7. Goswami J. C., Andrew K. Chan A Wiley-Interscience Publication. 1999, 306 p.
8. Smolyantsev N. K. Osnovy teorii veyvletov. Veyvlety v MATLAB [Wavelets in MATLAB]. Moscow: DMK press,2005, 304 p.
9. Shumarova O. S., Ignat'ev S. A. Vestnik Saratovskogo gosudarstvennogo tekhnicheskogo universiteta [Bulletin of Saratov State Technical University]. 2013, no. 4 (73), pp. 128–131.