JP4591850B2 - Ultrasonic inspection method and apparatus - Google Patents

Description

  The present invention relates to an ultrasonic inspection method and apparatus for inspecting an internal defect present in a structure.

  The soundness evaluation of structures such as nuclear pressure vessels and pipes includes (1) detection of defects in materials by nondestructive inspection, (2) measurement of defect size by nondestructive inspection (sizing), and (3) structure by fracture mechanics. Usually performed in the order of soundness evaluation. Among these, the TOFD method (Time Of Flight Diffraction Technique) is known as a defect size measuring means.

  The TOFD method is an inspection method using a diffracted wave. A defect (for example, a flaw) is placed at the center position with the transmitting probe and the receiving probe facing each other, and a longitudinal wave is excited from the transmitting element. Is a means for detecting a lateral wave, a diffracted wave from a defect, and a bottom surface echo, and measuring the depth (height) and size (length) of the defect from the arrival time difference and sound speed information. The TOFD method is disclosed in Patent Document 1, for example.

The TOFD method has a feature that can accurately measure the depth (height) and size (length) of a defect, but it is necessary to receive weak diffracted waves. There is a problem that cannot be applied.
For example, when the TOFD method is applied to stainless steel pipes, the structure is limited to pipes with a wall thickness of about 10 mm or less, such as structures reaching a wall thickness of 40 mm or more, such as pressure vessels and recirculation pipes for nuclear power generation facilities. Etc., the S / N ratio was too low due to the influence of scattering noise scattered at the crystal grain boundaries and the like, and accurate measurement could not be performed.

  In order to solve this problem, means using wavelet transform has already been proposed (for example, Patent Documents 2 and 3).

  The “ultrasonic signal processing device” of Patent Document 2 uses a directional multidimensional base 61 as shown in FIG. 11 in order to quantify directional components in a multidimensional space with respect to an inputted ultrasonic inspection signal. Using the transform means 62 for performing wavelet transform, the feature extraction means 63 for performing feature extraction from the wavelet transform coefficients generated by the transform means and calculating the feature points, and using the feature points generated by the feature extraction means, Discriminating means 66 for discriminating echoes based on the difference in surface distribution.

  The “ultrasonic echo signal processing apparatus” of Patent Document 3 is an ultrasonic echo signal processing apparatus used when inspecting a defective portion of the inspection object according to an ultrasonic echo signal obtained from the inspection object. As shown in FIG. 12, the waveform conversion unit 72 that generates a conversion coefficient by rotational wavelet conversion of the ultrasonic echo signal, and corresponds to the defective portion from the conversion coefficient according to the direction of the equiphase plane A significant component selecting unit 73 that selects a transform coefficient related to the echo as a significant component, and a coefficient synthesizing unit 74 that synthesizes the significant component based on a predetermined weighting factor and outputs the resultant as a filtered signal. .

JP 2002-162390 A, “Ultrasonic Inspection Apparatus and Method” JP 2001-165912 A, “Ultrasonic Signal Processing Device” JP 2003-66017 A, “Ultrasonic Echo Signal Processing Device”

The wavelet transform is a convolution of a mother wavelet with a received waveform. This transformation includes “discrete wavelet transformation” and “continuous wavelet transformation”.
The “discrete wavelet transform” is a discrete wavelet coefficient a and b, which has a short processing time but has a drawback of reduced resolution. Therefore, it could not be applied to ultrasonic inspection that requires highly accurate time measurement.

On the other hand, “continuous wavelet transform” is obtained by continuously taking coefficients a and b, and is suitable for extracting a specific frequency component of a signal.
However, in order to calculate the continuous wavelet transform, first, it is necessary to determine the scaling coefficient a and to perform convolution of the received waveform and the mother wavelet for each b while continuously changing b. The amount becomes enormous. Therefore, although time measurement can be performed with high accuracy, there is a problem that processing time is long and measurement in real time cannot be performed.

  However, in the field of inspecting structures such as nuclear pressure vessels and piping, in order to apply ultrasonic inspection, it is possible to perform continuous wavelet transform in almost real time and inspect internal defects in the structure in almost real time. It was requested.

  The present invention has been developed to solve such a demand. That is, an object of the present invention is to provide an ultrasonic inspection method and apparatus capable of inspecting an internal defect existing in a structure almost in real time even when the wall thickness is large and the S / N ratio is low due to the influence of noise or the like. Is to provide.

According to the present invention, an ultrasonic inspection method for inspecting a defect by propagating an ultrasonic wave into a structure, receiving a received waveform including a diffracted wave and a reflected wave from an internal defect,
A received waveform FFT step of obtaining a received waveform FFT by performing a fast Fourier transform on the received waveform;
A wavelet FFT step for obtaining a wavelet FFT by performing a fast Fourier transform on the mother wavelet;
An FFT integration step of multiplying the received waveform FFT and the wavelet FFT to obtain an integrated FFT;
There is provided an ultrasonic inspection method comprising: a coefficient analysis step of obtaining wavelet coefficients by performing inverse fast Fourier transform on the integrated FFT .

  Further, the method has a contour display step for obtaining a wavelet coefficient in a predetermined frequency range while changing the scaling coefficient a of the mother wavelet and displaying a contour map of the wavelet coefficient, and from the contour map, an optimum frequency (scaling coefficient a ).

  In addition, an ultrasonic transmission step of transmitting an ultrasonic pulse from the surface of the structure to an internal defect existing in the structure, and the received waveform obtained by diffracting the ultrasonic pulse by the internal defect is received by the surface of the structure. An ultrasonic receiving step.

Moreover, according to the present invention, there is provided an ultrasonic inspection apparatus for inspecting an internal defect existing in a structure from a received waveform received from the structure,
For a given frequency,
A received waveform FFT apparatus that obtains a received waveform FFT by performing a fast Fourier transform on the received waveform;
A wavelet FFT device that obtains a wavelet FFT by performing a fast Fourier transform on the mother wavelet;
An FFT integrator for multiplying the received waveform FFT by a wavelet FFT to obtain an integrated FFT;
There is provided an ultrasonic inspection apparatus comprising: a coefficient analysis apparatus that obtains a wavelet coefficient by performing a fast inverse Fourier transform on the integrated FFT.

According to a preferred embodiment of the present invention, there is provided a contour display device that displays a contour map of a wavelet coefficient by obtaining a wavelet coefficient in a predetermined frequency range while changing the scaling coefficient a of the mother wavelet.
An optimum frequency (scaling coefficient a) used for inspection is determined from the contour map.

Also, an ultrasonic transmission device that transmits ultrasonic pulses from the surface of the structure to internal defects present in the structure;
An ultrasonic receiving device that receives the received waveform obtained by diffracting the ultrasonic pulse due to the internal defect on the surface of the structure.

According to the method and apparatus of the present invention, all calculations are completed with only two FFTs (Fast Fourier Transform), one multiplication, and one IFFT (Inverse Fast Fourier Transform). Compared with conversion, the calculation time is significantly reduced.
Therefore, even when the wall thickness is large and the S / N ratio is low due to the influence of noise or the like, it is possible to inspect internal defects present in the structure almost in real time.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings. In each figure, common portions are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 is an overall configuration diagram of an ultrasonic inspection apparatus according to the present invention. As shown in this figure, the ultrasonic inspection apparatus of the present invention includes an ultrasonic transmission apparatus 12, an ultrasonic reception apparatus 14, and an ultrasonic analysis apparatus 16.

The ultrasonic transmission device 12 includes an ultrasonic transmitter 12a and a transmission probe 12b, and transmits an ultrasonic pulse 3 from the structure surface to an internal defect 2 existing inside the structure 1.
The ultrasonic receiving device 14 includes an ultrasonic receiver 14a and a receiving probe 14b, and includes a diffracted wave 3b in which an ultrasonic pulse is diffracted by the internal defect 2, a lateral wave 3a propagating along the flaw detection surface, and a bottom echo. Received waveform 4 including 3c is received on the surface of the structure.
The ultrasonic analysis device 16 includes an analysis computer 16a and an image display device 16b, instructs the ultrasonic transmission device 12 to transmit the ultrasonic pulse 3, receives the reception waveform 4 from the ultrasonic reception device 14, and performs wavelet transform. The signal processing such as the above is performed to detect and size the defect.

A case where the above-described ultrasonic inspection apparatus is applied to the TOFD method will be described below. The present invention is not limited to the TOFD method, and can be applied to other means.
When applied to the TOFD method, the transmission probe 12b and the reception probe 14b are positioned in close contact with the surface of the structure 1 with an interval L, and the defect 2 (for example, a flaw) is disposed at the center position. The means arranged at the center position makes the time from the reception waveform 4 to the transmission of the ultrasonic pulse 3 to the reception of the diffracted wave 3b described later, for example, the shortest.
Next, the ultrasonic pulse 3 is excited from the transmission probe 12 b, and the depth of the defect 2 is measured from the received waveform 4 received by the ultrasonic receiver 14.
Further, the size (length) of the defect 2 can be measured by scanning the transmission probe 12b and the reception probe 14b in the depth direction of FIG. 1 and performing the analysis.

FIG. 2 is a diagram illustrating an example of a received waveform received by the ultrasonic receiver 14.
As shown in this figure, the received waveform 4 includes a lateral wave 3a propagating along the flaw detection surface, a diffracted wave 3b from the defect 2, and a bottom echo 3c. Therefore, these can be detected, and the depth and size of the defect 2 can be measured from the arrival time difference and sound speed information.

FIG. 3 is a schematic view showing an ultrasonic inspection means according to the present invention.
In this figure, the analysis computer 16 a has a received waveform FFT device 21, a wavelet FFT device 22, an FFT integration device 23, and a coefficient analysis device 24.
The reception waveform FFT device 21 has a function of obtaining a reception waveform FFT 5 a by performing Fourier transform on the reception waveform 4.
The wavelet FFT device 22 has a function of obtaining a wavelet FFT 5b by Fourier transforming the mother wavelet 6 having a predetermined scaling coefficient a (predetermined frequency). Alternatively, the wavelet FFT 5b obtained by substituting the scaling coefficient “a” into the equation (1) of the equation 1 may be directly calculated.

The FFT integrating device 23 has a function of obtaining an integrated FFT 5c by multiplying the received waveform FFT 5a and the wavelet FFT 5b.
The coefficient analyzer 24 has a function of obtaining the wavelet coefficient 8 of the received waveform 4 by performing inverse Fourier transform on the integrated FFT 5c.
These devices 21, 22, 23, and 24 are preferably stored as software programs in a storage device (HD, RAM, external memory, etc.) in the analysis computer 16a.

  In FIG. 3, the image display device 16 b has a function of displaying the contour map 9 of the wavelet coefficient from the wavelet transform result 7 with respect to the scaling coefficient within a predetermined range.

Next, the method of the present invention will be described. The ultrasonic inspection method of the present invention receives the received waveform 4 in the ultrasonic transmission step S1 and the ultrasonic reception step S2 using the apparatus of FIG. 1 described above.
In the ultrasonic transmission step S1, an ultrasonic pulse 3 is transmitted from the structure surface to the internal defect 2 existing in the structure 1.
In the ultrasonic wave reception step S2, the reception waveform 4 including the diffracted wave 3b in which the ultrasonic pulse is diffracted by the internal defect 2, the lateral wave 3a propagating along the flaw detection surface, and the bottom surface echo 3c is received on the surface of the object.

The ultrasonic inspection method of the present invention further includes a received waveform FFT step S4, a wavelet FFT step S5, an FFT integration step S6, and a coefficient analysis step S7, as shown in FIG.
In step S3, the scaling coefficient a is continuously changed within a predetermined range.
In the reception waveform FFT step S4, the reception waveform 4 is Fourier transformed to obtain a reception waveform FFT5a.
In the wavelet FFT step S5, the mother wavelet 6 is Fourier-transformed to obtain a wavelet FFT 5b. Alternatively, the wavelet FFT 5b is obtained by substituting the scaling coefficient “a” into the equation (1) of Equation 1.
The received waveform FFT step S4 and the wavelet FFT 5b may be performed simultaneously or in the reverse order.
In the FFT integration step S6, the reception waveform FFT5a and the wavelet FFT5b are multiplied to obtain an integration FFT5c.
In the coefficient analysis step S7, the integrated FFT 5c is subjected to inverse Fourier transform to obtain the wavelet transform result 7 of the received waveform 4.

Furthermore, the ultrasonic inspection method of the present invention includes a contour line display step S8, a scaling coefficient selection step S9, and a wavelet coefficient determination step 10.
In the contour line display step S8, the wavelet coefficient 8 is obtained while changing the scaling coefficient a for the predetermined range of the scaling coefficient a in conjunction with step S3, and the contour map 9 of the wavelet coefficient is displayed.
In the scaling coefficient selection step S9, the scaling coefficient a from which the strong wavelet coefficient 8 is obtained by the lateral wave 3a, the diffracted wave 3b, and the bottom echo 3c is selected from the contour map 9 as an optimum value or an appropriate value.
In the wavelet coefficient determination step 10, the wavelet coefficient 8 based on the selected scaling coefficient a is determined as an optimum value or an appropriate value. This wavelet coefficient 8 is a relationship diagram between time and wavelet coefficients, and substantially corresponds to a received waveform whose S / N ratio is substantially improved. Therefore, the depth and size of the defect 2 can be measured from the arrival time difference and the sound speed information of the lateral wave 3a, the diffracted wave 3b, and the bottom surface echo 3c in this figure.

FIG. 4 is a schematic diagram of the above-described method of the present invention, and FIG. 5 is a schematic diagram showing a conventional method.
The continuous wavelet transform is expressed in the form of a convolution of the received waveform x (t) and the mother wavelet ψ (t) as shown in Equation (2) of Formula 2.

  The conventional method of FIG. 5 shows a convolution procedure in the time domain for a certain frequency (for a certain scaling factor: a). In the conventional calculation method using time domain convolution, the wavelet coefficient 57a at each position 56a is calculated while moving the position 56a of the mother wavelet. Therefore, it is necessary to calculate the convolution by the number of points of the reception waveform 54, and it takes time to calculate the wavelet coefficient 58 for the entire reception waveform.

On the other hand, the method used in the present invention shown in FIG. 4 shows a convolution procedure in the frequency domain for a certain frequency (for a certain scaling factor: a). In the calculation method using convolution in the frequency domain, first, FFTs 5a and 5b of the received waveform 4 and the mother wavelet 6 are respectively obtained. Next, the obtained FFTs 5a and 5b are multiplied, and the result 5c is subjected to inverse FFT to obtain the wavelet coefficient 8.
Therefore, in the method of the present invention, all calculations are completed with only two FFTs, one multiplication, and one IFFT, so that the calculation time is significantly reduced as compared with the conventional method.

The present invention described above has the following features.
(1) Since the continuous wavelet is applied to the TOFD method, the S / N ratio is greatly improved.
(2) The peak arrival time difference can be uniquely determined by using the continuous wavelet.
(3) Since a continuous wavelet of the type that performs convolution in the frequency domain is used, real-time analysis can be realized using a computer.
(4) Since the frequency shared by both the lateral wave and the diffracted wave is extracted and used for analysis, the frequency used for the TOFD method can be easily determined, and the peak arrival time difference can be accurately determined.

Examples of the present invention will be described below.
FIG. 6 is an analysis screen on the developed image display device 16b according to the present invention. In this figure, 1 is a received waveform (vertical axis: output, horizontal axis: time), 2 is an image (vertical axis: frequency, horizontal axis: time, 3 is a signal (vertical axis: wavelet coefficient, horizontal axis: time) obtained by extracting a specific frequency from the contour map of 2.

(Specimen shape and flaw detection method)
The ultrasonic inspection apparatus according to the present invention shown in FIG. 1 is manufactured, and a 46.2 mm thick austenitic stainless steel (SUS304, 45.3 mmW × 200 mmL × 46.2 mmT) is prepared as a test piece corresponding to the structure 1 Then, a through slit having a width of 0.8 mm was prepared as an internal defect 2 at the center of the test piece.
Using the ultrasonic inspection apparatus of FIG. 1, the flaw detection surface is the opposite surface of the slit (internal defect 2), and the probes 12b and 14b are for TOFD with a center frequency of 5 MHz, a transducer shape of 5 × 5 mm, and a refraction angle of 60 °. One set of broadband probes (manufactured by Japan Probe) was used.
The detected signal (received waveform 4) is taken into an analysis computer 16a via an A / D converter (not shown), and the above-described signal processing such as wavelet transform processing is performed in real time using originally developed software. The received waveform 4 is displayed on the screen of the image display device 16b as shown in FIG.
In this test, three types of signals were received with different S / N ratios in order to show the effectiveness of the continuous wavelet transform. The S / N ratio was changed in a pseudo manner depending on the presence or absence of 1024 averaging and the presence or absence of an attenuator (40 dB).

(Flaw detection result)
FIG. 7 is a diagram showing a received waveform according to the embodiment of the present invention. This figure shows a received waveform when the height from the flaw detection surface to the defect tip is 36.4 mm. It can be seen that the waveforms of (a) when not averaged and (c) when attenuated by 40 dB have a low S / N ratio and it is difficult to identify the diffracted wave 3b at the tip of the defect. Further, even when averaged (b), identification was difficult.

FIG. 8 is a contour diagram of wavelet coefficients according to the present invention. This figure shows the result of continuous wavelet transform of the received waveform of FIG. 7 according to the present invention. In this figure, the horizontal axis represents the same time as the received waveform, the vertical axis represents the frequency corresponding to the scaling coefficient a, and the luminance in the figure represents the strength of the wavelet coefficient. The intensity of the wavelet coefficient in this figure means that the dark part is low and the bright part is high.
For the mother wavelet 6, a Gabor function was used. Wavelet transform was performed from 1 MHz to 7 MHz in 200 kHz steps.

FIG. 9 is an explanatory diagram of the analysis frequency determination method according to the present invention. In this figure, (A) is a contour map of wavelet coefficients, and (B) is a wavelet coefficient for a certain scaling coefficient a.
The frequency components of the lateral wave 3a and the diffracted wave 3b are different from each other. As shown in FIG. 9A, the lateral wave 3a includes only a low-frequency ultrasonic component, and the diffracted wave 3b includes only a high-frequency ultrasonic component. Therefore, the frequency (scaling coefficient a) that the two waves have in common is selected and used for the analysis.
For example, in the example of FIG. 9A, both the lateral wave 3a and the diffracted wave 3b have frequency components in the vicinity of 3 MHz. Therefore, by extracting the frequency component of 3 MHz, both the lateral wave 3a and the diffracted wave 3b are obtained. Clearly shown wavelet coefficients (FIG. 9B) can be obtained.
From FIG. 9B, arrival time information of the lateral wave 3a and the diffracted wave 3b can be obtained and used to calculate the crack height.

  Similarly, in FIG. 8, when the wavelet transform result of the signal (C) attenuated by 40 dB is observed, the existence of the diffracted wave 3b at the tip of the defect can be clearly confirmed. The original signals have different amplitudes and S / N ratios, but almost no difference is recognized in the wavelet transform results. Further, it was found that the lateral wave 3a mainly includes a low frequency component, the diffracted wave 3b at the upper end of the defect mainly includes a high frequency component, and the bottom echo 3c includes a low frequency component to a high frequency component. These pieces of information can be used to identify each signal.

  Next, the defect 2 was sized using the difference in arrival time between the lateral wave 3a and the diffracted wave 3b at the tip of the defect. Since the frequency band of each wave is different as described above, sizing is performed using a 3 MHz wavelet coefficient having both components from FIG.

FIG. 10 is a diagram showing a wavelet transform result according to the present invention. This figure shows a 3 MHz wavelet coefficient. Sizing was performed using the peak arrival time difference between the lateral wave and the diffracted wave at the tip of the defect, but the measured value of the distance from the inspection surface to the tip of the defect was not averaged against the actual measurement value of 36.4 mm. The received waveform (A) was 36.5 mm, the averaged (B) was 36.5 mm, and the further attenuated by 40 dB (C) was 36.6 mm.
All the results were found to agree with the measured values with good accuracy. In addition, since the peak arrival time difference is uniquely determined, there is a possibility that variation in measurement results among inspectors can be suppressed, and automation of measurement can be easily realized.

  As described above, according to the present invention, the S / N ratio is improved by using continuous wavelet transform even with a thick material having a large ultrasonic attenuation and a bad S / N ratio of the received signal. For example, sizing by the TOFD method It was confirmed that can be done.

  In addition, this invention is not limited to embodiment mentioned above, Of course, it can change variously in the range which does not deviate from the summary of this invention.

1 is an overall configuration diagram of an ultrasonic inspection apparatus according to the present invention. It is a figure which shows an example of a received waveform. It is a schematic diagram which shows the ultrasonic inspection means by this invention. FIG. 2 is a schematic diagram of the method of the present invention. It is a schematic diagram which shows the conventional method. It is a figure which shows the analysis screen of the apparatus by this invention. It is a figure which shows the received waveform by the Example of this invention. FIG. 4 is a contour map of wavelet coefficients according to the present invention. It is explanatory drawing of the determination method of the analysis frequency by this invention. It is a figure which shows the wavelet transformation result by this invention. 10 is a schematic diagram of an “ultrasonic signal processing device” in Patent Document 2. FIG. 10 is a schematic diagram of an “ultrasonic echo signal processing device” in Patent Document 3. FIG.

Explanation of symbols

1 structure, 2 internal defects
3 Ultrasonic pulse, 3a lateral wave, 3b diffracted wave, 3c bottom echo,
4 Receive waveform, 5a Receive waveform FFT,
5b Wavelet FFT, 5c Integration FFT,
6 Mother wavelet, 7 Wavelet transform result,
8 wavelet coefficients, 9 contour plots,
12 ultrasonic transmitter, 12a ultrasonic transmitter, 12b transmission probe,
14 ultrasonic receiver, 14a ultrasonic receiver, 14b receiving probe,
16 ultrasonic analyzer, 16a computer for analysis,
16b image display device (contour line display device),
21 reception waveform FFT device, 22 wavelet FFT device,
23 FFT integration device, 24 coefficient analysis device

Claims (6)

An ultrasonic inspection method for inspecting a defect by propagating ultrasonic waves into a structure, receiving a received waveform including a diffracted wave and a reflected wave from an internal defect,
A received waveform FFT step of obtaining a received waveform FFT by performing a fast Fourier transform on the received waveform;
A wavelet FFT step for obtaining a wavelet FFT by performing a fast Fourier transform on the mother wavelet;
An FFT integration step of multiplying the received waveform FFT and the wavelet FFT to obtain an integrated FFT;
A coefficient analysis step of obtaining a wavelet coefficient by performing inverse fast Fourier transform on the integrated FFT, and an ultrasonic inspection method. A contour display step for obtaining a wavelet coefficient in a predetermined frequency range while changing the scaling coefficient a of the mother wavelet and displaying a contour map of the wavelet coefficient;
The ultrasonic inspection method according to claim 1 , wherein an optimum frequency (scaling coefficient a) used for inspection is determined from the contour map. An ultrasonic transmission step of transmitting ultrasonic pulses from the surface of the structure to internal defects present in the structure;
Ultrasonic inspection method according to claim 1 or 2, wherein the internal defects with ultrasound pulses having an ultrasonic reception step of receiving by diffracted the received waveform structure surface, characterized in that. An ultrasonic inspection apparatus for inspecting internal defects existing in a structure from a received waveform received from the structure,
For a given frequency,
A received waveform FFT apparatus that obtains a received waveform FFT by performing a fast Fourier transform on the received waveform;
A wavelet FFT device that obtains a wavelet FFT by performing a fast Fourier transform on the mother wavelet;
An FFT integrator for multiplying the received waveform FFT by a wavelet FFT to obtain an integrated FFT;
An ultrasonic inspection apparatus comprising: a coefficient analysis device that obtains a wavelet coefficient by performing a fast inverse Fourier transform on the integrated FFT. A contour display device that displays a contour map of a wavelet coefficient by obtaining a wavelet coefficient in a predetermined frequency range while changing the scaling coefficient a of the mother wavelet;
The ultrasonic inspection apparatus according to claim 4 , wherein an optimum frequency (scaling coefficient a) used for inspection is determined from the contour map. An ultrasonic transmission device for transmitting ultrasonic pulses from the surface of the structure to internal defects present in the structure;
The ultrasonic inspection apparatus according to claim 4 , further comprising: an ultrasonic reception device that receives the reception waveform obtained by diffracting an ultrasonic pulse due to the internal defect on a structure surface.

Images