RU2246724C1 - Method of ultrasonic testing of material quality - Google Patents

Abstract

FIELD: nondestructive testing; ultrasonic testing.

SUBSTANCE: to detect signal reflected from a defect and to estimate location of defect, information properties of signal are analyzed at output of receiver of flaw detector by means of large-scale wavelet testing methodology which allows to find small in amplitude local changes of signal including "dead zone" of flaw detector. Wavelet-spectrograms of reference signal are formed at output of receiving path of flaw detector which corresponds to case when defect is absent in tested sample. When defect of unknown quality appears in tested material, wavelet-spectrogram of controlled signal at output of receiving path of flaw detector Wf (a, t) and its a_i cross-section at different scales of a_i would differ from wavelet-spectrogram of reference signal which is adequate to detection of signal reflected from defect. To find the location of defect the wavelet-spectrograms of shortened realizations of reference and controlled signals are formed sequentially for time of τ_y. Shortening of signals is performed from the end of sequence for the same value till wavelet-spectrograms of shortened realizations of reference and controlled signals as well as their cross-sections become practically equal at different scales of a_i. Length of those shortened realizations of reference and controlled signals has to be evaluation of defect location.

EFFECT: improved precision.

2 cl, 11 dwg

Description

The invention relates to the field of non-destructive testing of materials and products by ultrasonic methods and can be used to detect defects in welds and in the main material, including cracks, sinks, lack of fusion, non-fusion, slag inclusions, etc., and can also be used in radar to detect targets in the near field and assess their location.

The analogues of the present invention are various variants of methods and devices for ultrasonic quality control of materials described, for example, in [1] and implemented in patents:

- RU 97105446, “Method of ultrasonic control of the thickness of products”, publ. 1999.03.27;

- RU 99103394, “Ultrasonic flaw detector”, publ. 12/12/20;

- RU 99115325, “Ultrasonic method for monitoring products and materials”, publ. 05/05/20.

It is known that when ultrasonic testing of materials by an echo-pulse method, the control range is limited by the isolation between the transmitting transducer and the receiving path of the flaw detector and the reverberation-noise characteristics of the transducer, which determine the dead zone of the flaw detector or that minimum thickness of the product, on which the amplitude of the echo pulse reflected from its internal surface decreases so much that it becomes commensurate with the level of interference and is not amenable to clear registration.

The magnitude of the dead zone l _min depends on the acoustic characteristics of the piezoelectric element, protector, damper and prism included in the transducer:

where c is the propagation velocity of ultrasonic vibrations in the material,

τ _u - the duration of the exciting pulse, which determines the forced oscillations of the piezoelectric element,

τ _n - the duration of the transition process, depending on the duration of the free oscillations of the piezoelectric element.

The reverberation-noise characteristics of the converter depend primarily on the attenuation coefficient of the waves in the damper and prism [2].

The classical method of shortening the dead zone of a flaw detector by reducing the mechanical quality factor by damping is not effective enough, since it reduces the sensitivity of an ultrasonic flaw detector. The so-called electrical methods of damping the radiating transducer [3] have more universal capabilities, when implemented, a steep voltage drop A ₁ is applied to the half-wave resonant piezoelectric element, exciting sinusoidal mechanical vibrations in it, which decay exponentially. After a time τ ₃ equal to half the period of the natural resonant frequency of the piezoelectric element, another voltage drop of the same steep and almost equal in amplitude amplitude A ₂ is applied to the latter, exciting mechanical vibrations in it, but 180 ° phase-shifted. A superposition of damped oscillations leads to their mutual compensation over the entire duration of these radio pulses, with the exception of the first half-wave of oscillations resulting from the first drop of the exciting electric voltage. By changing the parameters of the excitation mode of the piezoelectric elements τ _s , A ₁ and A ₂ , it is possible to realize the resulting oscillations with different levels of vibration compensation. This allows you to provide the necessary sensitivity when detecting defects both in the near monitoring zone and in the far. However, these methods are not implemented in the inverse transformation of acoustic vibrations in the transducer in the receiving mode.

In order to reduce the reverberation-noise characteristics, separately combined direct and inclined converters are used, which, in comparison with combined direct converters, have a more complex constructive solution, and their application complicates the control procedure [2].

The closest analogue of the proposed invention is a method for determining the size of a defect during ultrasonic inspection of products, which consists in the fact that the ultrasound transducer scans the product, simultaneously excites ultrasonic vibrations in it and the amplitude of the received signals and the distance between the positions of the ultrasonic transducer judge the size of the defect, and before control products by ultrasonic transducer scan a tuning sample having a calibration defect, the size of which is at an order or more exceeds the wavelength of the excited ultrasonic vibrations, fix the amplitudes of the received and the corresponding displacement value of the ultrasonic transducer, determine the threshold level of the amplitude of the received signal and compare the amplitudes and distances between the positions of the ultrasonic transducer of the product and the sample, and when the signal amplitude reaches the maximum value, the defect classified as extended and its dimensions are defined as the distance between the positions of the ultrasound transducer at which the amplitude of the signal is equal to the threshold level, and if the amplitude of the signal from the defect does not reach the maximum value, then the defect is classified as local and its value is defined as the distance between the positions of the ultrasonic transducer, at which the amplitude of the signal changed from zero to the obtained amplitude (RU 2000104686, 2001.12.20).

A disadvantage of the known methods of ultrasonic testing is the impossibility of detecting a defect and assessing its location when the signal reflected from the defect is in the dead zone of the flaw detector.

The essence of the alleged invention lies in the fact that in the known method of ultrasonic testing during the procedure for detecting a signal reflected from a defect and evaluating its location, the information properties of the signal at the output of the detector are analyzed using the methodology of multi-scale wavelet analysis, which allows us to identify sufficiently small wavelets amplitude local changes in the signal, including in the dead zone of the flaw detector. This is achieved due to the fact that “wavelet spectrograms” of the reference signal are generated at the output of the flaw detector receiving path, which corresponds to the case when there is no defect in the test sample. If a defect appears in the test material of unknown quality, the wavelet spectrogram of the controlled signal at the output of the flaw detector receiving path W f (a, t) and its a _i section at various scales a _i will differ from the wavelet spectrogram of the reference signal, which adequate to detect the signal reflected from the defect. To determine the location of the defect, “wavelet spectrograms” of shortened implementations of the reference and controlled signals are generated for a time τ _y , and the shortening of these signals is performed sequentially by the same amount until the “wavelet spectrograms” of shortened implementations of the reference and controlled signals and their cross sections at various scales a _i become almost the same: the length of such shortened implementations of a reference or controlled signal is an estimate of the location location of the defect.

Figure 1 shows a functional diagram of one of the variants of the device of ultrasonic quality control of the material, which implements the proposed method. Here you can see the controlled material 1, sample 2 of the material without defect, receiving piezoelectric transducer 3, transmitting piezoelectric transducer 4, generator 5 of probing pulses, receiving device 6, storage device 7, “I” circuit 8, program device 9, wavelet spectrogram generator 10 , calculator 11 a _i - sections of the “wavelet spectrogram”, the device 12 comparison.

First, the transmitting piezoelectric transducer 4 scans a material sample 2 in which there are no defects, and then the controlled material 1, while the scanning modes of the transmitting piezoelectric transducer 4 are set by a modulating signal U _m (t), which is generated at the output of the probe generator 5. The ultrasonic vibrations received by the receiving piezoelectric transducer 3 are sent to the receiving device 6, from the output of which, when working with the controlled material 1, and when working with the reference material 2, the corresponding control signals U _to (t) and reference U _e (t) enter the storage device 7 . Through the circuit “And” 8 at the command U _omc (t) from the software device 9 produce a sample of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t), and then these signals are sent to the shaper 10 “wavelet spectrograms”.

It is known [4] that if the function Ψ (t) satisfies two conditions:

- its average value is zero,

- the function Ψ (t) rapidly decreases as t → ± ∞,

then it is a wavelet and for any function f (x) we can find the function Wf (x, a), which depends on two variables: time x and scale a _i .

. В отличие от преобразования Фурье, которое не позволяет локализовать частотные компоненты сигнала во времени и, вследствие этого, строго применимо только для анализа стационарных сигналов, вейвлет-преобразование не имеет такого ограничения.For a fixed value of the scale a _{i, the} function Wf (x, a) is a convolution of the original function f (x) with a wavelet stretched a _i - times. Since the convolution of functions is equivalent to their multiplication in the frequency domain, the cross section for a _i = const in the image of the wavelet transform (1) shows changes in the studied function at frequencies close to

. In contrast to the Fourier transform, which does not allow to localize the frequency components of the signal in time and, therefore, is strictly applicable only to the analysis of stationary signals, the wavelet transform does not have such a limitation.

Next, the first program output device 9 is formed a number, which is a scale value _i, according to which in the calculator 11 and the calculated -spektrogramm _i a _i -section "Wavelet Image" Wf _e (x, a) the reference signal U _EM (t ) and a _{i is the} section of the “wavelet spectrogram” Wf _k (x, a) of the control signal.

In the next step, the a _i- section of the “wavelet spectrogram” Wf (x, a) of the reference signal U _em (t) and the control signal U _к (t), calculated for the same scale values a _i , goes to the comparison device 12 : if a _i -sections of the “wavelet spectrogram” of the reference U _em (t) and the controlled U _to (t) signals are different, then there is a defect in the “dead zone”. Otherwise, there is no defect. Information about the detection of a defect is formed at the first output of the comparison device 12.

The procedure for detecting a signal reflected from a defect is explained in the examples shown in FIGS. 2-5. The signal implementations in these examples were obtained with a PICUS 10C flaw detector with a direct combined piezoelectric transducer, which provides excitation of acoustic waves and their acoustoelectronic conversion at a frequency of 2.5 MHz. In the wavelet transforms of these implementations, the Morlet complex wavelet was applied [4].

Figure 2 presents the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) (graph 2a), which determines the dead zone of the flaw detector, and the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) (graph 2b), formed when there is a defect in the material with a diameter of 2 mm, the minimum distance from which to the piezoelectric transducer is 2 mm, and also the wavelet spectrogram of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) (graph 2 s) and pilot signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) (graph 2 d). Signals are received at the following flaw detector modes:

- the amplitude of the modulating signal | U _m (t) | = 100 V,

- gain of the receiving device K _y = 20 dB.

Matching reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) the signals shown in FIG. 2 do not reveal significant differences in them, however, their wavelet spectrograms differ significantly. Similar significant differences are also present in a _i -sections of the “wavelet spectrograms” of these signals, which are presented in FIG. 3:

- on the graph 3 a - a _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 27,

- on the graph 3 b - and _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 27,

- on the graph 3 s - and _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 45,

- on the graph 3 d - a _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 45.

From the graphs presented in FIG. 2 and FIG. 3, it follows that the appearance in the dead zone of a flaw detector of any heterogeneity causes the formation of significant differences in the wavelet spectrograms of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) signals, as well as a _i- sections of the “wavelet spectrograms” of these signals, which allows us to consider the differences of wavelet spectrograms and their sections at different scales a _{i of the} reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) signals as a criterion for detecting a signal reflected from any heterogeneity that is a defect in the material being controlled.

The above detection procedure allows you to detect a defect when it is outside the dead zone of the flaw detector. Figure 4 and figure 5 presents graphs illustrating this procedure when the minimum distance of the defect from the piezoelectric transducer is 108 mm, and the signals are obtained under the following flaw detector modes:

- the amplitude of the modulating signal | U _m (t) | = 200 V,

- gain of the receiving device K _y = 20 dB.

Here:

- on the graph 4 a - reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t)

- on the graph 4 b- control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t)

- on the graph 4 s - wavelet spectrograms of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t)

- on the graph 4 d - wavelet spectrograms of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t)

- on the graph 5 a - a _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 27,

- on the graph 5 b - a _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 27,

- on the graph, 5 s - and _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 45,

- on the graph 5 d - a _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 45.

After detecting a defect in accordance with the above procedure, the signal from the second output of the comparison device 12 is fed to the input of the software device 9, in which, with a step, the time parameters of which are consistent with the amplitude-frequency characteristics of the acoustic and receiving paths of the flaw detector, the action time of the command during the “I” circuit is open: in this case, the implementation length of the reference U is reduced $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) signals, and shortening is performed from the end of the implementation of these signals. The a _i sections of the “wavelet spectrogram” Wf (x, a) of shortened realizations of the reference U are calculated $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) the signals calculated for the same scale values a _i and compare them in the device 12 comparison. Reducing the length of the implementation of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) the signals are produced until their “wavelet spectrograms” differ. In this case, the length of the implementations of the reference U _em (t) and the controlled U _to (t) signals is an estimate of the distance from the piezoelectric transducer to the defect.

The procedure for assessing the location of a defect located in the dead zone of the flaw detector, explain the graphs in Fig.6-8.

Figure 6 presents a _i- section of the “wavelet spectrograms” of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) signals whose implementation length is shortened from their end to 6 mm:

- on the graph 6 a - a _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 27,

- on the graph 6 b - and _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 27,

- on the graph 6 s - a _i -section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 45,

- on the graph 6 d - and _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 45.

Figure 7 presents a _i- section of the “wavelet spectrograms” of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) signals whose implementation length is shortened from their end to 4 mm:

- on the graph 7 a - a _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 27,

- on the graph 7 b - and _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 27,

- on the graph 7 s - a _i -section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 45,

- on the graph 7 d - a _i -section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 45.

On Fig presents a _i- section “wavelet spectrograms” of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) signals whose implementation length is shortened from their end to 2 mm:

- on the graph 8 a - a _{i is the} cross section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 27,

- on the graph 8 b - and _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 27,

- on the graph, 8 s - and _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 45,

- on the graph 8 d - and _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 45.

Analysis of the graphs shown in Fig.6-8, shows that a _i -sections of “wavelet spectrograms” of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) of the signals at the same scales a _i practically do not differ only when these signals are shortened to 2 mm, therefore the duration of these signals, corresponding to a distance of 2 mm, is an estimate of the location of the defect relative to the piezoelectric transducer.

The location estimation procedure described above allows the location of a defect to be estimated when it is outside the flaw detector dead zone. The procedure for assessing the location of the defect in this case is explained by the graphs in Figs. 9-11.

Figure 9 presents a _i- section of the “wavelet spectrograms” of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) signals whose implementation length is shortened from their end to 115 mm:

- on the graph 9 a - a _i is the cross section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 27,

- on the graph 9 b - a _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 27,

- on the graph 9 s - a _i -section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 45,

- on the graph 9 d - a _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 45.

Figure 10 presents a _i- section of the “wavelet spectrograms” of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) signals whose implementation length is shortened from their end to 110 mm:

- on the graph 10 a - a _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 27,

- on the graph 10 b - and _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 27,

- on the graph 10 s - and _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 45,

- on the graph 10 d - and _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 45.

11 shows a _i -section of “wavelet spectrograms” of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) signals whose implementation length is shortened from their end to 108 mm:

- on the graph 11a - and _{i is the} cross section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 27,

- on the graph 11b - and _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 27,

- on the graph 11c - and _{i is the} section of “wavelet spectrograms” of the reference signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) for scale a _i = 45,

- on the graph 11d - a _{i is the} section of “wavelet spectrograms” of the control signal U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) for scale a _i = 45.

The analysis of the graphs shown in Fig.9-11, shows that a _i -section “wavelet spectrograms” of the reference U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{uh}}$ (t) and control U $\genfrac{}{}{0ex}{}{\text{zu}}{\text{to}}$ (t) of signals at the same scales a _i practically do not differ only when these signals are shortened to 108 mm, therefore, the duration of these signals, corresponding to a distance of 108 mm, is an estimate of the location of the defect relative to the piezoelectric transducer.

The above results indicate the high efficiency of the proposed method, which ensures the detection of a defect and the assessment of its location in the "dead zone" of the flaw detector and outside it. This is especially true when controlling the quality of welds, pipes, wheelsets, etc.

Literature

1. Krautkremer J., Krautkremer G. “Ultrasonic control of materials. Handbook ”, Moscow: Metallurgy, 1991, 752 p.

2. V.G. Scherbinsky, N.P. Aleshin. “Ultrasonic control of welded joints”, M.: Publishing house of MGTU im. N.E.Bauman, 2000, 496 p.

3. M.V. Korolev. “Echo-pulse thickness gauges”, Moscow: Mashinostroenie, 1980, 111 p.

4. I.Daubechies “Ten Lectures on Wavelets. CBMS-NSF Regional Conference Series in Applied Mathematics, v. 61 - Philadelphia: SIAM, 1992.

Claims (2)

1. The method of ultrasonic quality control of materials, characterized in that, in contrast to the known method of ultrasonic control, the detection of the signal reflected from the defect and its location are made in the "dead zone" of the flaw detector, for this, to detect a defect in the controlled material at the output of the receiving path the ultrasonic flaw detector record the reference signal, which corresponds to the case when the defect in the "dead zone" is absent, and the control signal during ultrasonic inspection of the sample is unknown of quality, then successively produce said signals for calculating the "wavelet spectrograms" Wf (a, t) and a _i - sections "wavelet spectrograms" at the same scale a _i, are inputted to calculator -sections a _i a _i -sections " wavelet spectrograms "from a generator of scale a _i , compare a _i -sections a _i -sections of" wavelet spectrograms "with the same scales a _i over the entire time interval: if the corresponding sections of the reference and control signals at an arbitrary scale a _i do not differ, then there is no defect in the controlled material, but the EU whether at any scale a _i -sections of the “wavelet spectrograms” of the reference and control signals are different, then there is a defect in the “dead zone”, to determine the location of the defect, the reference and control signals are identical in time of implementation, shortened by τ _i from end, which in the programming device a predetermined time interval τ _i, for which the reference signals are not controlled and are input to the shaper "wavelet spectrograms" Wf (a, t) output from the aND circuit, wherein said signal producing shortening pos edovatelno at the same value as long as the "wavelet spectrogram" truncated reference implementations and the control signals and their cross sections at different scales a _i are almost the same, then the length of shortened realization reference or control signal is an estimate of the defect's location. 2. The method of ultrasonic quality control of materials according to claim 1, wherein the detection of a defect in the controlled material and the assessment of its location is carried out when the defect is outside the "dead zone" of the flaw detector.

Images