Enhanced phase method of signal detection for ultrasonic magnetostriction defectoscopy of power equipment

The magnetostrictive method of ultrasonic flaw detection has certain advantages, in particular, the ability to control objects with complex geometry, at their high temperature, the ability to control dry contact between the transducer and the object, difficult access to the controlled area, etc. The peculiarities of the generation of ultrasonic waves by magnetostrictive transducers and their distribution in control objects determine the low level of the signal/noise ratio, which limits the possibilities of practical application of this method. The aim of the paper is to improve the phase method for detecting signals of magnetostrictive receivers with a low signal/noise ratio and to check the effectiveness of the proposed technical solution for solving problems of non-destructive testing of energy equipment elements using magnetostrictive defectoscopes. The paper discusses the phase method of detecting radio pulse signals of magnetostrictive converters against the background of additive noise, which is based on a combination of the capabilities of the discrete Hilbert transformation, which makes it possible to determine the envelope and phase of signals, and methods of statistical processing of the results of phase measurements. The proposed signal processing algorithm was studied both in a model experiment and when processing real magnetostrictive defectoscope signals. The proposed method makes it possible to detect radio pulse signals with a signal/noise ratio close to 1. The reliability of the obtained data is confirmed by the results of computer simulation. The considered method of detecting signals can be used in ultrasonic magnetostrictive defectoscopes and other diagnostic systems operating in conditions of reduced signal/noise ratio. Keywords: magnetostrictive defectoscope, ultrasonic defectoscopy, phase methods of signal processing, Hilbert transform, envelope, phase, sample resulting length of the vector.