RU2292544C2 - Method of improving noise immunity of ultrasonic inspection - Google Patents

Abstract

FIELD: ultrasonic inspection.

SUBSTANCE: signal, received by ultrasonic converter, is divided to several time zones. Signals are divided physically by means of, for example, commutating unit. Components of main signal are selected from the main signal, which components are disposed in areas of waiting of signals, reflected from defects. Maximal signal amplitude within zone of waiting of signal, reflected from defects, is determined. Value of the amplitude is compared with admissible one and decision is made on presence or absence of defect on base of results of comparison. Results of the estimation are later used for making decision on necessity of repeat inspection of whole article or of only suspicious part of the article. Amplitude of frequency components are determined are determined in signal spectra, being located in areas of signals waiting, which amplitudes have to be character or non-character for probing pulse spectrum or for received useful signals spectrum. Amplitudes of useful components are compared with amplitudes of frequency components, being non-characteristic for useful signal spectrum. Results of the comparison are used for judging on truth of decision on discontinuity detection. To clean signal off regular component, not only basic cycles of coherent or non-coherent storage of received signals is carried out, but additional cycles as well. Realization, which was cleaned off regular component, is received by subtracting total values of realization, achieved as a result of relatively short and relatively long cycles of storages.

EFFECT: improved noise immunity at ultrasonic inspection.

5 dwg

Description

The invention relates to the field of non-destructive testing and can be used in ultrasonic testing systems, mainly installations for automated ultrasonic testing of sheet and long products and pipes.

The transducer of an ultrasonic flaw detector emits pulses of high-frequency elastic vibrations into the product. The radiation of the probe pulses occurs mainly along the normal to the surface of the controlled products, hundreds or even thousands of times per second. Accordingly, the receiving transducer (often receiving and emitting using the same transducer) hundreds or thousands of times per second receives responses from the product or, as they are also called, “implementations”.

Each implementation contains signals reflected from the rental surfaces and from defects, as well as pulses due to interference, which in most cases is acoustic or electromagnetic in nature [1].

In the absence of interference, each implementation contains information that makes it possible to judge the thickness of the controlled product at the location of the ultrasonic transducer, the presence or absence of discontinuities, etc. Interference can distort the information received, cause the wrong decision to be made about the physical and consumer properties of the controlled product.

Various methods are used to reduce the level of interference, in particular, coherent (or incoherent) accumulation of the implementation. Useful signals that are “tied” to the probe pulse and therefore have relatively slowly varying amplitudes and phases from realization to realization are effectively added up. The interference, arrival time, the amplitude and phase of which are mostly random in nature, is not so efficiently summed in the drive. Therefore, coherent (as well as incoherent) accumulation in most cases can effectively increase the ratio of "useful signal / interference".

However, there are specific disturbances that may turn out to be partially or completely synchronized with the probe pulses and, therefore, with useful signals. Such interference is very dangerous because it can accumulate in a coherent (or incoherent) drive. An example of a source of synchronous interference is a control computer complex, which is part of any modern multichannel system of automated ultrasonic monitoring. In many cases, synchronous interference is also oscillations not attenuated in the product caused by previous probing pulses, or stray vibrations of scale particles, or air bubbles that have fallen into the working area of an ultrasonic transducer (EMAT or probe) [2], [3].

For random, but quite powerful and long-term interference, there is also the possibility of “breaking through" the drive. The following circumstance contributes to this.

Known methods for recording signals from pulsed ultrasonic flaw detectors require the use of broadband filters, which is due to the need for selection of defects located fairly close to natural reflectors, which, as a rule, form powerful "bottom" signals.

A broadband filter provides reception of signals without “falling over” their fronts, which favorably affects the resolution. However, the noise immunity of such a filter in some practical cases may be unsatisfactory.

When trying to match the filter characteristics with the spectrum of the received signals, the shape of the latter is significantly distorted, the signal fronts “fall over”, and this adversely affects the possibility of detecting defects, especially when controlling sheet metal of relatively small thickness.

In the absence of interference, broadband reception provides significant advantages.

The method is illustrated by illustrations.

Figure 1. Broadband filter receiver. Type of sales typical for sheet and long products.

Figure 2. Consistent filtering receiver. The influence of the passband of the receiver of a pulsed ultrasonic flaw detector on the detection of defects located near natural reflectors, where

1 - probe pulse;

2, 3 - the first and second "bottom" pulses;

4 - signals from defects.

Figure 3. The confluence of pulses 2 and 3 for the case of a decrease in the thickness of the rolled four times.

Figure 4. Reception of signals by a broadband receiver.

Figure 5. An example of dividing the received signal into zones where

A - zone of action of the probe pulse;

B, D - waiting zones for signals reflected from defects;

C, E - waiting zones of "bottom" signals;

1, 2, 3, 4 - as in figure 1.

The advantages of using broadband receivers are as follows.

1. The signals at the output of the broadband receiver have a significantly (sometimes several times) shorter "dead zone". In Fig. 1, the probe pulse is much shorter than that in Fig. 2. Because of this, one of the pulses due to the discontinuity, visible in figure 1, "disappeared" in the "dead zone" in figure 2 (the first after the probe pulse).

2. The ability to identify defects located in close proximity to natural reflectors. In FIG. 2, we cannot confidently note the fact of the presence or absence of some signals reflected from defects, which, nevertheless, are clearly visible in FIG.

3. Broadband reception provides a wider range of rolled thicknesses that can be controlled. For example, when the thickness is reduced by four times, control using a matched receiver is impossible: pulses 2 and 3 in figure 3 will practically merge. At the same time, for the broadband receiver (Figure 4), monitoring will be possible, since the gap between pulses 2 and 3 will still be sufficient to register pulses reflected from defects in it.

Another advantage of a broadband receiver is its ability to provide significantly higher accuracy in determining the distance between the "bottom" pulses. This is extremely important, for example, when performing ultrasonic thickness measurement.

However, all these advantages of broadband reception appear only in conditions when there is no interference.

In the presence of interference, for example, industrial pulsed electromagnetic disturbances, broadband reception leads to false positives of flaw detection equipment. These false positives lead to unreasonable rejection and in most cases are completely unacceptable to the manufacturer of metal products.

This is especially true for cases of the use of high-performance multichannel systems of automated ultrasonic testing, standing directly in the technological stream of rolled metal production. An important property of such systems, unlike manual flaw detectors, is the need to carry out monitoring on a “one pass” basis. It is very difficult, and often physically impossible, to return to the controlled area.

On the other hand, the optimization of the reception channel bandwidth is always a compromise and in some cases does not bring the desired result. As a rule, such systems have neither the required noise immunity nor satisfactory resolution.

The proposed method can effectively eliminate this contradiction. The present invention is based on the task of creating an ultrasonic monitoring method with increased noise immunity, eliminating the distortion of received signals and providing high reliability of the monitoring results.

This task is achieved by the fact that in the method of increasing the noise immunity of ultrasonic testing, including broadband reception using ultrasonic transducers, registration and processing of ultrasonic signals, according to the invention, the signal received by the ultrasonic transducer is conventionally divided into several time zones, using, for example, switching devices to perform physical the separation of signals located in different time zones, at least from the main signal, its composition is separated Signals located in the waiting zones of signals reflected from defects are subjected to independent processing, for example, signals from the zones of operation of the probing pulse and waiting for "bottom" signals are fed to sufficiently broadband amplifiers, which ensure the preservation of the shape of the pulses and the steepness of their edges, signals from the waiting areas of signals reflected from defects, they are filtered, providing the maximum signal / noise ratio at the output of the corresponding device, the maximum signal amplitude is determined In the waiting areas of signals reflected from the defect, the value of this amplitude is compared with the maximum permissible value and, based on the results of this comparison, a decision is made on the presence or absence of a defect, and an additional decision is made on the results of this assessment about the need to re-check the entire product or only its “suspicious” "Of the plot, in the spectrum of signals located in the waiting areas of signals reflected from the defect, determine the amplitudes of the frequency components, characteristic and uncharacteristic of the spectrum probe pulses or the spectrum of the received useful signals, compare the values of the amplitude of the “useful” components with the amplitude of the frequency components that are not characteristic of the spectrum of the useful signal, according to the results of this comparison, the reliability of the decision to detect a discontinuity is judged, not only the main cycles are performed to clear the signals from the regular component coherent or incoherent accumulation of received signals, but also auxiliary cycles, and the implementation, cleared of the regular component, gets t by subtracting the total realizations derived from relatively short and relatively long accumulation cycles.

For example, if the frequency components that are not characteristic of the useful signal prevail in the spectrum of signals received in zone B and D, they conclude that the conclusion about the detection of discontinuity is unreliable. It is very likely that exceeding the allowable threshold occurred as a result of interference.

The inventive method additionally also assumes the possibility of cleaning the signal in the waiting areas of pulses reflected from defects from the regular component, which at the box office can be caused, for example, by undamped oscillations or by the presence of spurious waves excited by piezoelectric or electromagnetic-acoustic transducers.

Exceeding the permissible threshold may occur as a result of interference.

The regular component in the rental may be due, for example, undamped oscillations or the presence of spurious waves excited by piezoelectric or electromagnetic-acoustic transducers.

Information sources

1. J. Krautkremer, G. Krautkremer. Ultrasonic control of materials. Reference manual. Per. with him. - M.: Metallurgy, 1991, 752 p.

2. RF patent 2123401.

3. A.V. Kirikov, A.N. Zabrodin. Features of the use of EMAT with ultrasound testing. The magazine "In the world of non-destructive testing", N1, March 2002, S. 5-8.

Claims (1)

A method for increasing the noise immunity of ultrasonic monitoring, including broadband reception with ultrasonic transducers, recording and processing of ultrasonic signals, characterized in that the signal received by the ultrasonic transducer is conventionally divided into several time zones, using, for example, switching devices, physically separate the signals located in different time zones, at least from the main signal, its components are located in the waiting zones signals reflected from defects, the signals from different zones are subjected to independent processing, for example, signals from the zones of operation of the probe pulse and standby bottom signals are fed to sufficiently broadband amplifiers that ensure the preservation of the shape of the pulses and the steepness of their edges, signals from the standby zones of signals reflected from defects, subjected to filtering, providing the maximum signal to noise ratio at the output of the corresponding device, determine the maximum amplitude of the signal in the waiting areas of the signals reflected from a defect, the value of this amplitude is compared with the maximum permissible value, and according to the results of this comparison, a decision is made on the presence or absence of a defect, based on the results of this assessment, an additional decision is made on whether it is necessary to re-check the entire product or only its “suspicious” section, in the signal spectrum, the signals reflected in the waiting zones reflected from the defect determine the amplitudes of the frequency components that are characteristic and uncharacteristic of the spectrum of the probe pulses or the spectrum of the received useful signals, compare the values of the amplitude of the “useful” components with the amplitude of the frequency components that are not characteristic of the spectrum of the useful signal, according to the results of this comparison, the reliability of the decision to detect discontinuities is judged, not only the main cycles of coherent or incoherent accumulation are performed to clear the signals from the regular component received signals, but also auxiliary cycles, and the implementation, cleared of the regular component, is obtained by subtracting the total implementation, p obtained as a result of a relatively short and relatively long accumulation cycles.

Images