JP2005331439A - Analysis method and analyzer of flaw detection signal - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an analysis method and an analyzer of a flaw detection signal for enhancing a S/N ratio of a defect echo and reducing a time-consuming work for analyzing data. <P>SOLUTION: The analyzer 10 has a spatial filter apparatus 11 for implementing a filtering process of the inputted flaw detection signal 18, a sample feature quantity database 15 provided with the sample feature quantity data of the defect echo and pseudo echo, an automatic analyzer 14 for matching the sample feature quantity database 15 with the feature quantity found from the flaw detection signal or the flaw detection signal after the filtering process and analyzing it, an optimal display 16 for optimally displaying the flaw detection signal or the flaw detection signal after the filtering process on a B scope, and a relative position display 17 for displaying an image of a specimen model on a screen on which the flaw detection signal or the flaw detection signal after the filtering process is displayed. The analyzer 10 extracts and analyzes the defect echo. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a flaw detection signal analysis method and analysis device in a flaw detection apparatus using ultrasonic waves or radar.

FIG. 6 is a schematic configuration diagram showing a configuration of a general ultrasonic flaw detector. As shown in the figure, from the ultrasonic transmission / reception unit 1a of the flaw detector 1 to a probe 3 which is a sensor attached to a subject 2 such as a steel material directly or via a delay material such as water, a fixed period A pulse signal having T ₀ is transmitted. The probe 3 converts the received pulse signal into an ultrasonic pulse and applies it to the subject 2.

  The ultrasonic pulse applied into the subject 2 is reflected by the bottom surface of the subject 2, the defect 2a, etc., and is received by the probe 3 again. The probe 3 converts the reflected wave into an electric signal and transmits it to the ultrasonic transmission / reception unit 1a. The ultrasonic transmission / reception unit 1a amplifies the electric signal, sends it to the display unit 1b as an echo signal as a flaw detection signal, and records it in the recording unit 1c.

  The flaw detection signal recorded in the recording unit 1c is sent to the analyzer 4 via a network or a recording medium. The analysis device 4 performs various signal processing and analysis on the echo signal, and outputs a signal processing result, presence / absence of a defect (analysis result), and the like.

  In the ultrasonic flaw detector having such a configuration, the echo signal output from the ultrasonic transmission / reception unit 1a includes a lot of noise in addition to the defect echo. If the noise contained in the echo signal is large, the reliability of the flaw detection result is greatly impaired. This noise is roughly classified into three types: electrical noise, material noise, and shape noise.

  The electrical noise includes external noise caused by electromagnetic waves mixed into the probe 3, the ultrasonic transmission / reception unit 1a and the connection cable, internal noise generated by an amplifier incorporated in the ultrasonic transmission / reception unit 1a, and the like. Consists of.

  On the other hand, as material noise, for example, an ultrasonic wave is scattered at the crystal grain boundary of the material of the subject 2 and a reflected wave returns to the probe 3 from other than the defect, and is included in the echo signal as a scattered echo. There is noise. Such an echo signal caused by scattering from the crystal grain boundary of the specimen material is called a forest echo and is called a pseudo echo instead of a defect echo caused by a defect of the specimen. The shape noise is an echo generated depending on the shape of the subject 2.

  Reducing these noises contained in the echo signal and improving the S / N ratio of the defect echo is extremely important in carrying out ultrasonic flaw detection with high accuracy. As a countermeasure for reducing this pseudo echo, it has been proposed to remove the pseudo echo using a frequency filter from an echo signal obtained by making an ultrasonic pulse having a wide bandwidth enter the subject. This is used on the assumption that the frequency of the defect echo is lower than the frequency of the forest echo (refer to Patent Document 1 below).

JP-A-2-186261 JP-A-6-207928

  However, the analysis apparatus that employs the signal processing method for the echo signal described above still has the following problems to be solved. That is, in the conventional method, in order to improve the S / N ratio of defect echoes, one-dimensional and time-series bandpass filter processing and addition averaging processing (signal processing in the depth direction in which ultrasonic waves are incident) are performed. .

  Therefore, although the method described above is effective for electrical noise, it is less effective for other noises. In addition, since the frequency of the defect echo is not necessarily lower than the frequency of the forest echo, the time-series one-dimensional bandpass filter cannot completely separate the forest echo from the defect echo, and the extracted defect The echo still has room for improving the S / N ratio.

  In addition, as a method for extracting the defect echo, threshold processing of the signal amplitude (absolute threshold, relative threshold or relative threshold according to distance such as DAC) is performed, but the defect echo that is the extraction target is overlooked. If the threshold value is lowered in order to reduce the number of noise echoes, the number of detected noise echoes from factors other than defects increases and excessive extraction increases. Therefore, a more accurate defect echo extraction method is desired.

  Further, as a flaw detection method, there are a two-dimensional scan using X and Y components which are scan directions on the surface of the subject, and a flaw detection method using a phased array sensor to which a component of an ultrasonic incident angle is further added. However, with these methods, a large number of time-series data can be obtained. Therefore, when analyzing the data by displaying it in the B scope, the optimal display position was manually selected and analyzed for all the data. So take a very long time.

  Furthermore, in the flaw detection method using a two-dimensional scan or a phased array sensor, as described above, a large amount of time-series data is obtained, so that the amount of data to be processed becomes enormous. Therefore, the operation of the analyzer took time and the data transfer took time.

  The present invention has been made in view of the above-described situation, and improves the S / N ratio of a defect echo to be extracted, automatically optimizes an echo signal to be analyzed, and reduces the labor involved in data analysis. It is an object of the present invention to provide a method and an apparatus for analyzing a flaw detection signal that can perform inspection.

A method for analyzing a flaw detection signal according to the present invention for solving the above-mentioned problems is as follows.
In an analysis method for analyzing a flaw detection signal obtained by moving a sensor or changing an incident angle of an ultrasonic wave or a radar to make the ultrasonic wave or radar incident on a subject,
The flaw detection signal analysis method is characterized by filtering the flaw detection signal with a multidimensional spatial filter and extracting a defect echo.

  A multidimensional spatial filter is a spatial filter of two or more dimensions, and is a filter that removes a specific spatial shape. By filtering the flaw detection signal using a multidimensional spatial filter, pseudo echoes such as electrical noise, forest echoes, and shape echoes are reduced, and the S / N ratio of the defect echoes to be extracted is improved.

Further, in the above-described flaw detection signal analysis method,
The multidimensional spatial filter has a function of filtering a pseudo echo derived from the shape or tissue of the subject, and is a method for analyzing a flaw detection signal.

  The shape echo signal from the subject is derived from a pseudo echo derived from the shape of the subject, for example, a change in the pipe diameter or pipe shape of the pipe-shaped subject, or a shape called seaning existing in the vicinity of welding. It is known that there is a pseudo echo. The unique pseudo echo derived from the shape of the subject is a defect echo by removing the pseudo echo in the direction in which the shape distribution is known to be uniform, for example, in the case of seaming, along the weld. To improve the S / N ratio.

Further, in the above-described flaw detection signal analysis method,
The multidimensional spatial filter is a morphological type filter or a median type filter.

  By using a morphological filter or a median filter as the spatial filter, changes in the signal waveform of the defect echo before and after filtering are suppressed.

Further, in the above-described flaw detection signal analysis method,
Further, the defect echo sample signal and / or the pseudo echo sample signal or its feature quantity database is matched with the flaw detection signal, its feature quantity, or the filtered flaw detection signal or its feature quantity. This is a method for analyzing a flaw detection signal.

  A sample data of defect echoes and pseudo echoes known in advance by empirical rules or simulations or their feature values are compiled into a database, and the obtained flaw detection signals or their feature values are compared with database information. Improve accuracy.

Further, in the above-described flaw detection signal analysis method,
The flaw detection signal comprises an X and Y component that is a scan direction of the subject surface and a time component that is an ultrasonic wave or radar depth direction incident on the subject. Is the method.

  An example in which the present invention is applied to a so-called two-dimensional scanning flaw detection method (a flaw detection signal obtained is a three-dimensional signal to which a time component is added) in which a probe is scanned two-dimensionally on the surface of a subject. is there.

Further, in the above-described flaw detection signal analysis method,
The flaw detection signal includes an X and Y component that is a scanning direction of the subject surface, a time component that is an ultrasonic wave or radar depth direction incident on the subject, and an ultrasonic wave or radar incident direction. It is a flaw detection signal analysis method characterized by comprising an incident angle component.

  In addition to the above-described two-dimensional scanning, a flaw detection method using a so-called phased array sensor that scans by changing the incident direction of ultrasonic waves or radar (the obtained flaw detection signal is a four-dimensional signal to which an incident angle component is added. ) Is an example in which the present invention is applied.

Further, in the above-described flaw detection signal analysis method,
When the flaw detection signal or the flaw detection signal after the filtering process is displayed on a B scope, the optimum cross section in the space composed of the X, Y component and the time component is automatically selected so as to display the defect echo. This is a method for analyzing a flaw detection signal.

Further, in the above-described flaw detection signal analysis method,
Analyzing flaw detection signals, wherein when the flaw detection signals or flaw detection signals after the filtering process are displayed on a B scope, the incident angle component is automatically selected so as to be in the incident direction toward the defect echoes. Is the method.

  As a method of automatically selecting and processing the optimum cross section in the space composed of the X, Y component and the time component, and further the incident angle component, it can be realized by repeating convergence calculation or performing principal component analysis, for example. This reduces the labor of analysis and reduces the amount of data to be processed.

Further, in the above-described flaw detection signal analysis method,
In the flaw detection signal analysis method, the flaw detection signal or the flaw detection signal after the filtering process is displayed in correspondence with the model of the subject.

  Since it is difficult to grasp the generation position and reflection direction of the echo to be analyzed using only the flaw detection signal, the image of the object model is superimposed on the screen displaying the flaw detection signal, or the flaw detection signal is displayed on the object model. It can be displayed by sticking or displayed in a separate window. This makes it possible to easily determine the possibility of a pseudo echo source or a defect echo source, such as estimating a pseudo echo based on the specific shape of the subject, thereby reducing the labor of analysis.

The flaw detection signal analyzer according to the present invention that solves the above problems is as follows.
In an analyzer for analyzing a flaw detection signal obtained by moving a sensor or changing an incident angle of an ultrasonic wave or a radar to make the ultrasonic wave or radar incident on a subject.
The flaw detection signal analyzer has filtering means for filtering the flaw detection signal with a multidimensional spatial filter, and extracts the defect echo.

In the above-described flaw detection signal analyzer,
The multidimensional spatial filter is a flaw detection signal analyzer having a function of filtering a pseudo echo derived from the shape or tissue of the subject.

In the above-described flaw detection signal analyzer,
The multidimensional spatial filter is a flaw detection signal analyzing apparatus characterized by being a morphological filter or a median filter.

In the above-described flaw detection signal analyzer,
Furthermore, it has a database composed of the sample signal of the defect echo and / or the sample signal of the pseudo echo or the feature amount thereof, the database, the flaw detection signal or the feature amount thereof, or the flaw detection signal after the filtering process or the It is a flaw detection signal analyzer characterized by having an automatic analysis means for matching and analyzing a feature quantity.

In the above-described flaw detection signal analyzer,
The flaw detection signal comprises an X and Y component that is a scan direction of the subject surface and a time component that is an ultrasonic wave or radar depth direction incident on the subject. Device.

In the above-described flaw detection signal analyzer,
The flaw detection signal includes an X and Y component that is a scanning direction of the subject surface, a time component that is an ultrasonic wave or radar depth direction incident on the subject, and an ultrasonic wave or radar incident direction. It is a flaw detection signal analyzer characterized by comprising an incident angle component.

In the above-described flaw detection signal analyzer,
When the flaw detection signal or the flaw detection signal after the filtering process is displayed on a B scope, the optimum cross section in the space composed of the X, Y component and the time component is automatically selected so as to display the defect echo. An inspection device for flaw detection signals, characterized by having an optimum cross-section selection means.

In the above-described flaw detection signal analyzer,
When the flaw detection signal or the flaw detection signal after the filtering process is displayed on a B scope, it has optimum incident angle selection means for automatically selecting the incident angle component so as to be in the incident direction toward the defect echo. This is a characteristic flaw detection signal analyzer.

In the above-described flaw detection signal analyzer,
An apparatus for analyzing a flaw detection signal, comprising: relative position display means for displaying the flaw detection signal or the flaw detection signal after the filtering process in association with the model of the subject.

According to the flaw detection signal analysis method and analysis apparatus according to the present invention,
In an analysis method and an analysis apparatus for analyzing a flaw detection signal obtained by moving a sensor or changing an incident angle of an ultrasonic wave or a radar to make the ultrasonic wave or radar incident on a subject.
Since the flaw detection signal is filtered by a multi-dimensional spatial filter and a defect echo is extracted,
By reducing pseudo echoes such as forest echoes and shape echoes, the S / N ratio of defect echoes to be extracted can be improved.

In the above-described flaw detection signal analysis method and analysis apparatus,
Since the multidimensional spatial filter has a function of filtering a pseudo echo derived from the shape or tissue of the subject,
It is possible to improve the S / N ratio of the defect echo by filtering and removing the pseudo echo in the direction in which the shape and the tissue distribution are known to be uniform.

In the above-described flaw detection signal analysis method and analysis apparatus,
Since the multidimensional spatial filter is a morphological filter or a median filter,
Changes in the signal waveform of the defect echo before and after filtering can be suppressed.

In the above-described flaw detection signal analysis method and analysis apparatus,
Further, the defect echo sample signal and / or the pseudo echo sample signal or its feature quantity database is matched with the flaw detection signal or its feature quantity or the flaw detection signal or its feature quantity after being filtered. I decided to analyze
The accuracy of analysis can be improved.

In the above-described flaw detection signal analysis method and analysis apparatus,
Since the flaw detection signal is composed of the X and Y components that are the scan direction of the subject surface and the time component that is the ultrasonic wave or radar depth direction incident on the subject.
The present invention can be applied to a so-called two-dimensional scanning flaw detection method (a flaw detection signal obtained is a three-dimensional signal to which a time component is added) in which the probe is scanned two-dimensionally on the surface of the subject. it can.

In the above-described flaw detection signal analysis method and analysis apparatus,
The flaw detection signal includes an X and Y component that is a scanning direction of the subject surface, a time component that is an ultrasonic wave or radar depth direction incident on the subject, and an ultrasonic wave or radar incident direction. Because it is made up of incident angle components,
In addition to the above-described two-dimensional scanning, a flaw detection method using a so-called phased array sensor that scans by changing the incident direction of ultrasonic waves or radar (the obtained flaw detection signal is a four-dimensional signal to which an incident angle component is added. The present invention can be applied to.

In the above-described flaw detection signal analysis method and analysis apparatus,
When the flaw detection signal or the flaw detection signal after the filtering process is displayed on a B scope, the optimum cross section in the space composed of the X, Y component and the time component is automatically selected so as to display the defect echo. I decided to do so
The labor of analysis can be reduced and the amount of data to be processed can be reduced.

In the above-described flaw detection signal analysis method and analysis apparatus,
When the B-scope display of the flaw detection signal or the flaw detection signal after the filtering process, the incident angle component is automatically selected and processed so as to be in the incident direction toward the defect echo.
The labor of analysis can be reduced and the amount of data to be processed can be reduced.

In the above-described flaw detection signal analysis method and analysis apparatus,
When displaying the flaw detection signal or the flaw detection signal after the filtering process, the model of the subject is displayed correspondingly,
It is possible to easily determine the possibility of a pseudo echo source or a defect echo source, such as estimating a pseudo echo based on a specific shape of the subject, thereby reducing the labor of analysis and improving the analysis accuracy.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments according to the present invention will be described in detail with reference to the drawings, but the present invention is not limited thereto. FIG. 1 is a schematic configuration diagram of an analysis apparatus that performs a flaw detection signal analysis method according to an embodiment of the present invention.

<Configuration of analyzer according to this embodiment>
As shown in the figure, the analysis device 10 according to the present embodiment includes a spatial filter device 11 (filtering means) for filtering the flaw detection signal 18 input to the analysis device 10, sample feature quantity data of pseudo echoes and pseudo-faults. A sample feature database 15 including sample feature data of echoes, and an automatic analyzer 14 (automatic analysis device 14) that matches and analyzes the sample feature database 15 and a feature value obtained from a flaw detection signal or a flaw detection signal after filtering processing. Analysis means), an optimal display device 16 that performs optimal display when the flaw detection signal is displayed on the B scope, and a relative position display that displays the image of the subject model on the flaw detection signal or the flaw detection signal display screen after the filtering process. And a device 17 (relative position display means).

  The spatial filter device 11 is a multi-dimensional (two or more dimensional) spatial filter, and examples of the spatial filter device 11 include a morphological filter device 12a and a median filter device 12b. In addition, it also has a shape noise filter device 13 for filtering a pseudo echo derived from the shape of the subject, and the shape noise filter device 13 includes a morphology type and a median type 13a.

  The optimum display device 16 includes an optimum section selection device 16a (optimum section selection means) that optimizes the X, Y components and time components of the flaw detection signal, and an object that optimizes the ultrasonic incident angle component of the flaw detection signal. There is an optimum incident angle selecting device 16b (optimum incident angle selecting means). The analysis device 10 receives the flaw detection signal 18 and performs various signal processing in the device, and then outputs an analysis result 19.

<Types of flaw detection signals>
Examples of the flaw detection signal 18 include a flaw detection signal obtained by one-dimensional scanning, a flaw detection signal obtained by two-dimensional scanning, a flaw detection signal obtained by a phased array ultrasonic flaw detection apparatus, and the like. The two-dimensional scan is a method of scanning the probe in two dimensions on the surface of the subject to detect the flaw, and the flaw detection signal is incident on the subject and the X and Y components that are the scan direction of the subject surface. It consists of a time t component (the defined area is a total three-dimensional signal) that is the depth direction of the ultrasonic wave.

  The phased array flaw detection method is a method of flaw detection while changing the ultrasonic incident angle on the subject in addition to the two-dimensional scan described above, and the flaw detection signal is the scan direction of the subject surface X , Y components, a time t component that is the depth direction of the ultrasonic wave incident on the subject, and an incident angle θ component that is the incident direction of the ultrasonic wave (the defined area is a total of four-dimensional signals). .

<Filtering processing by spatial frequency>
FIG. 2 is an explanatory diagram for explaining filtering processing by spatial frequency. In the figure, among the obtained flaw detection signals, a display showing the value V of the flaw detection signal when the time component t and the component X in the scan direction are changed, that is, a B scope display (V = V (t, X )). The display of the value V of the flaw detection signal with respect to the time component t while fixing the components X and Y in the scan direction is referred to as A scope display (V = V (t)). Further, the display indicating the value V of the flaw detection signal when the components X and Y and the time component t are changed is referred to as C scope display (V = V (t, X, Y)).

  Conventionally, for example, when the flaw detection signal is V = V (t, X), a band pass filter or the like is applied to the flaw detection signal for each component X, that is, the time series signal V (t). This method has the problems described above.

  As shown in FIG. 2, when the domain is defined as a two-dimensional component of time component t and component X, the defect echo 30 has a somewhat wide distribution (low frequency), whereas electrical noise and forest echo Etc. have a narrow distribution (high frequency).

  Therefore, in this case, the spatial filter device 11 is a two-dimensional spatial frequency filter that filters high frequencies, thereby reducing the pseudo echoes 31 such as electrical noise and forest echoes. Thus, the S / N ratio of the extracted defect echo 30 can be improved. Note that when a flaw detection signal having a dimension larger than two dimensions is targeted, a filtering process may be performed using a multidimensional spatial filter.

<Morphological and median filter devices>
The spatial filter device 11 is preferably a shape filter device such as a morphological filter device 12a or a median filter device 12b.

  For the morphological filter and the median filter, for example, the result of applying these filters to the flaw detection signal V (t, X) having a two-dimensional domain is expressed as follows. .

（Ｗｔ，Ｗｘは定数） Morphological filter

(Wt and Wx are constants)

（Ｗｔ，Ｗｘは定数） Median filter

(Wt and Wx are constants)

  FIG. 3 is a conceptual diagram showing an example of filtering processing by the shape filter in comparison with the bandpass filter. In addition, the same figure demonstrates the effect of the filtering in A scope display. As shown in the figure, the original signal is a waveform having a peak on an obliquely drifting base, and a bandpass filter or a shape filter is used as filtering for matching the obliquely drifting base to the time component t-axis. There is a way.

  The filtering by the band filter can extract the peak waveform of the original signal, but the peak waveform is deformed, whereas the filtering by the shape filter extracts the peak waveform while suppressing the deformation of the peak waveform. Can do. By using the shape filter, it is possible to suppress the deformation of the waveform feature of the defect echo due to filtering.

<Shape noise filter device>
FIG. 4 is a diagram illustrating an example of a specific pseudo echo derived from the shape of the subject. The figure shows a pipe-shaped subject having a weld (FIG. 4A), a cross-sectional view taken along the line AA in FIG. 4A, and a surface of the subject. A flaw detection signal V (X, Y, t) ((c) in the figure) when performing a dimension scan is shown. The shape noise filter device 13 will be described with reference to FIG.

  When the welded portion 22 is present in the pipe-shaped subject 20, there may be a seasoning 23 that is a portion that is cut and thinned in the vicinity of the welded portion 22. Since the seaming 23 is formed on the inner surface of the pipe over the entire circumference (X component direction) of the pipe-shaped subject 20, the pseudo echo (seaning echo 32) from the seaming 23 is generated in the flaw detection signal. It is generated as a pseudo echo along the direction of the X component which is the direction.

  Accordingly, the seaming echo 32 in the direction along the welded portion 22 can be filtered and removed, and the S / N ratio of the defect echo 30 can be improved. The seaming echo 32 is a low-frequency flaw detection signal in the direction of the X component in FIG. 4C, and the defect echo 30 can be extracted by removing the low-frequency component in the direction along the welded portion 22. it can.

  As a pseudo echo signal from the subject, in addition to a unique seaming echo existing in the welded part, a pseudo echo derived from the shape of the subject, for example, from a change in pipe diameter or pipe shape of a pipe-shaped subject It is known that there is a pseudo echo.

  With respect to such a specific pseudo echo derived from the shape of the subject, similarly to the seasoning echo 32, the pseudo echo in the direction in which the shape distribution is known to be uniform is removed, and the S / N of the defect echo is obtained. The ratio can be improved.

<Flaw detection inspection by two-dimensional scanning>
For example, the S / N ratio of a crack defect can be improved by performing various kinds of spatial filtering processing as described above on a flaw detection signal obtained by flaw detection inspection by two-dimensional scanning.

  In the case of flaw detection inspection for the purpose of detecting a flaw defect, a flaw detection signal by two-dimensional scanning has a continuous distribution in the direction of the flaw defect. By reducing the high frequency component of the spatial frequency in this direction, the S / N ratio of the crack defect can be improved.

<Flaw detection inspection using phased array>
Further, with respect to the flaw detection signal obtained by the flaw detection inspection by the phased array, it is possible to further reduce the weld noise and improve the S / N ratio of the defect echo. FIG. 5 is a diagram for explaining an example of the flaw detection signal by the phased array sensor. In the case of oblique flaw detection, there are two types of echoes from defects, a corner echo (FIG. 1A) and an end echo (FIG. 2B).

  As shown in the figure, it is known that the corner echo and the end echo do not cause an abrupt echo change with respect to the change in the incident angle θ of the ultrasonic wave (θ1 → θ2). On the other hand, for example, in the case of a welded part echo by surface reflection from the welded part 22, the echo change rate with respect to the change in the incident angle θ is large.

  Therefore, by applying a spatial filter to the flaw detection signal to which the incident angle dimension is added as the signal definition area, the weld echo and other shape echoes are reduced, and the S / N ratio of the defect echo is reduced. Can be improved. The fact that the rate of change in echo with respect to the incident angle is high means that the frequency is high with respect to the incident angle component. Use a spatial filter.

<Automatic analyzer>
As described above, the multi-dimensional spatial filter device 11 can extract a defect echo or a pseudo echo from a flaw detection signal obtained by flaw detection. The flaw detection signal or the flaw detection signal after the filtering process is automatically analyzed by the automatic analyzer 14.

  The sample feature quantity database 15 is a database of feature quantities of sample signals of defect echoes and pseudo echoes known in advance by empirical rules or simulations. The automatic analysis device 14 can improve the accuracy of the analysis by comparing the feature quantity of the obtained flaw detection signal with the database information.

  As a matching method, for example, first, the feature quantity of each signal is calculated by analyzing the waveform characteristics of the sample signal and the flaw detection signal. The feature amount includes a signal intensity ratio, a frequency range, and the like. Next, automatic analysis is performed by comparing and comparing the feature quantities of the signals.

<Optimum display device>
The optimum display device 16 is, for example, in a space composed of optimum X, Y components and time components so that a flaw detection signal obtained by a two-dimensional scanning flaw detection method or a flaw detection method using a phased array sensor is displayed including a defect echo. As for the optimal cross-section selection device 16a (optimal cross-section selection means) that automatically selects and processes the optimal cross-section of the sensor, and the flaw detection method using the phased array sensor, the incident angle component is automatically selected so as to be in the incident direction toward the defect echo There is an optimum incident angle selecting device 16b (optimum incident angle selecting means).

  In general, in the analysis of flaw detection signals, the flaw detection signals are analyzed by displaying all cross sections by A scope display, B scope display or C scope display. However, processing a huge amount of data or performing these displays manually. It takes a lot of time and effort.

  Therefore, the optimum section selection device 16a obtains the optimum projection direction in two dimensions of the two-dimensionally scanned flaw detection signal (three-dimensional signal composed of the X, Y components and the time component t) and sets it as the analysis target. That is, when displaying the flaw detection signal or the flaw detection signal B scope after the filtering process, the optimum cross section is automatically selected and processed so that the display includes the defect echo.

  As a method for automatically selecting and processing the optimum cross section composed of the X, Y component and the time component t, for example, it can be realized by repeating convergence calculation or performing principal component analysis. Reduce labor and reduce the amount of data to be processed.

  The method of repeating the convergence calculation uses the echo density projected onto each display surface as an evaluation value, and performs the convergence calculation so that the evaluation value is maximized. By targeting the defect echo detected by the automatic analyzer 14 and changing the X and Y components in the projection direction and the time component t and by iterating the cross section that projects the distribution space, the echo density contained in the cross section is obtained. Calculate the highest surface and select the optimal cross section.

  In the principal component analysis method, principal component analysis is performed on defect echoes detected by the automatic analyzer 14 with a flaw detection signal value as a probability density (for example, a high probability density is obtained when the signal value is high). Thus, the main component axis and the second axis perpendicular to the main axis are obtained, and the plane including these axes is selected as the optimum cross section.

  Further, the optimum incident angle selection device 16b automatically selects and processes the incident angle component θ so that the incident direction is directed to the defect echo when the flaw detection signal or the flaw detection signal after the filtering process is displayed on the B scope. Examples of a method for obtaining the optimum incident angle component θ include a method of repeating the above-described convergence calculation and a method by principal component analysis.

<Relative position display device>
The relative position display device 17 has a function of displaying the flaw detection signal displayed on the screen or the flaw detection signal after filtering processing (each scope display) and the model of the subject in association with each other. A screen for displaying flaw detection signals because it is difficult to determine from which part of the subject the echo to be analyzed is the echo reflected from which angle, etc. with only these flaw detection signals. On the top, for example, an image of the subject model is superimposed, a flaw detection signal or a filtering signal is attached to the subject model, or displayed in another window.

  The relative position display device 17 associates the flaw detection signal with the position of the three-dimensional shape (examination space) of the subject, and confirms the position of the signal source with the subject model displayed on the same screen, thereby providing a unique characteristic of the subject. It is possible to easily determine the possibility of a pseudo echo source or a defect echo source, such as estimating a pseudo echo based on a shape, thereby improving analysis accuracy and reducing labor of analysis.

  In the above-described embodiment, the flaw detection method using ultrasonic waves has been described. However, the present invention is not limited to this, and the flaw detection method using radar can also be applied. In other words, the flaw detection signal obtained by using the radar may be input to the analyzer 10 as the flaw detection signal 18 shown in FIG.

It is a schematic block diagram of the analyzer which implements the analysis method of the flaw detection signal concerning embodiment of this invention. It is explanatory drawing explaining the filtering process by a spatial frequency. It is a conceptual diagram which shows an example of the filtering process by a shape filter in comparison with a band filter. It is a figure explaining an example of the peculiar pseudo echo originating in the shape of a subject. It is a figure explaining an example of the flaw detection signal by a phased array sensor. It is a schematic block diagram which shows the structure of a general ultrasonic flaw detector.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Flaw detector 1a Ultrasonic transmission / reception part 1b Display part 1c Recording part 2 Subject 2a Defect 3 Probe 4 Analyzer

10 Analyzer 11 Spatial filter device (for multidimensional)
12a Morphological type filter device 12b Median type filter device 13 Shape noise filter device 13a Morphological type / Median type 14 Automatic analyzer 15 Sample feature database 16 Optimal display device 16a Optimal cross-section selection device 16b Optimal incident angle selection device 17 Relative position display device 18 Flaw Detection Signal 19 Analysis Result 20 Subject 21 Probe 22 Welding Portion 23 Seaning 24 Defect 30 Defect Echo 31 Pseudo Echo 32 Seaning Echo 33 Forest Echo

Claims (18)

In an analysis method for analyzing a flaw detection signal obtained by moving a sensor or changing an incident angle of an ultrasonic wave or a radar to make the ultrasonic wave or radar incident on a subject,
A flaw detection signal analysis method, wherein the flaw detection signal is filtered by a multidimensional spatial filter to extract a defect echo. In the flaw detection signal analysis method according to claim 1,
The flaw detection signal analysis method, wherein the multi-dimensional spatial filter has a function of filtering a pseudo echo derived from a shape or tissue of the subject. In the analysis method of the flaw detection signal according to claim 1 or 2,
The flaw detection signal analysis method, wherein the multidimensional spatial filter is a morphological filter or a median filter. In the analysis method of the flaw detection signal in any one of Claim 1 thru | or 3,
Further, the defect echo sample signal and / or the pseudo echo sample signal or its feature quantity database is matched with the flaw detection signal, its feature quantity, or the filtered flaw detection signal or its feature quantity. The analysis method of the flaw detection signal characterized by analyzing. In the analysis method of the flaw detection signal in any one of Claim 1 thru | or 4,
The flaw detection signal comprises an X and Y component that is a scan direction of the subject surface and a time component that is an ultrasonic wave or radar depth direction incident on the subject. Method. In the analysis method of the flaw detection signal in any one of Claim 1 thru | or 4,
The flaw detection signal includes an X and Y component that is a scanning direction of the subject surface, a time component that is an ultrasonic wave or radar depth direction incident on the subject, and an ultrasonic wave or radar incident direction. A method for analyzing flaw detection signals, comprising an incident angle component. In the analysis method of the flaw detection signal according to claim 5 or 6,
When the flaw detection signal or the flaw detection signal after the filtering process is displayed on a B scope, the optimum cross section in the space composed of the X, Y component and the time component is automatically selected so as to display the defect echo. A method for analyzing a flaw detection signal. The method for analyzing a flaw detection signal according to claim 6,
Analyzing flaw detection signals, wherein when the flaw detection signals or flaw detection signals after the filtering process are displayed on a B scope, the incident angle component is automatically selected so as to be in the incident direction toward the defect echoes. Method. In the analysis method of the flaw detection signal in any one of Claim 1 thru | or 8,
A method for analyzing a flaw detection signal, wherein the flaw detection signal or the flaw detection signal after the filtering process is displayed in association with the model of the subject. In an analyzer for analyzing a flaw detection signal obtained by moving a sensor or changing an incident angle of an ultrasonic wave or a radar to make the ultrasonic wave or radar incident on a subject.
A flaw detection signal analyzer, comprising: filtering means for filtering the flaw detection signal with a multidimensional spatial filter, and extracting the defect echo. In the flaw detection signal analyzer according to claim 10,
The multi-dimensional spatial filter has a function of filtering a pseudo echo derived from the shape or tissue of the subject, and a flaw detection signal analyzer. In the flaw detection signal analyzer according to claim 10 or 11,
The flaw detection signal analyzing apparatus, wherein the multidimensional spatial filter is a morphological filter or a median filter. In the flaw detection signal analyzer according to any one of claims 10 to 12,
Furthermore, it has a database composed of the sample signal of the defect echo and / or the sample signal of the pseudo echo or the feature amount thereof, the database, the flaw detection signal or the feature amount thereof, or the flaw detection signal after the filtering process or the An apparatus for analyzing flaw detection signals, comprising automatic analysis means for matching and analyzing feature quantities. The flaw detection signal analyzer according to any one of claims 10 to 13,
The flaw detection signal comprises an X and Y component that is a scan direction of the subject surface and a time component that is an ultrasonic wave or radar depth direction incident on the subject. apparatus. The flaw detection signal analyzer according to any one of claims 10 to 13,
The flaw detection signal includes an X and Y component that is a scanning direction of the subject surface, a time component that is an ultrasonic wave or radar depth direction incident on the subject, and an ultrasonic wave or radar incident direction. A flaw detection signal analyzer comprising an incident angle component. In the flaw detection signal analyzer according to claim 14 or 15,
When the flaw detection signal or the flaw detection signal after the filtering process is displayed on a B scope, the optimum cross section in the space composed of the X, Y component and the time component is automatically selected so as to display the defect echo. An analysis device for flaw detection signals, characterized by comprising means for selecting an optimal cross section for the flaw detection signal. In the flaw detection signal analyzer according to claim 15,
When displaying the flaw detection signal or the flaw detection signal after the filtering process in a B-scope display, it has optimum incident angle selection means for automatically selecting and processing the incident angle component so that the incident direction is directed toward the defect echo. Characteristic flaw detection signal analyzer. In the flaw detection signal analyzer according to any one of claims 10 to 17,
An apparatus for analyzing a flaw detection signal, comprising: relative position display means for displaying the flaw detection signal or the flaw detection signal after the filtering process in association with the model of the subject.

Images