JPH0419558A - Image processing method for ultrasonic flaw detection test - Google Patents

Abstract

PURPOSE:To accurately measure the position and shape of a defect of a material to be inspected by performing ultrasonic flaw detection by a probe 1 from >=2 measurement positions nearby a place where the material inspected is assumed to have the defect, and calculating a composite image pattern only at a position where image patterns overlap with each other. CONSTITUTION:The probe 1 is held on an arm which moves in an X-axial and a Y-axial direction and the surface of the object material 2 is scanned by the probe 1 according to the control signal of a scanner controller 5 to perform the flaw detection by an ultrasonic flaw detector 3. Then a signal processing means generates image patterns of a reflection source with reflection echoes obtained from the measurement positions provided at a coordinate positions where the defect of the object material 2 is estimated. Then the composite image pattern of only the position where the image patterns of the reflection source overlap with each other is calculated and the position and shape of the calculated composite image pattern are measured as the position and shape of the defect.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a pulse reflection type ultrasonic flaw detection testing method,
The present invention relates to an image processing method in an ultrasonic flaw detection test for accurately measuring the position and shape of a defect in a material to be inspected.

[Prior Art] For example, a method of welding steel plates or the like and inspecting the welded portion for defects using an ultrasonic flaw detector is widely used. When detecting defects, it is desired to be able to accurately measure the position and shape of defects in a material to be inspected.

FIG. 6 is a diagram illustrating a conventional ultrasonic flaw detection method, and shows cross sections of the test material 2 and the angle probe IO. Since the angle probe 10 injects ultrasonic waves from an oblique direction to the flaw detection surface of the test material 2, the transducer 1 is placed on a wedge material (also referred to as a shoe) 11 made of, for example, acrylic resin.
2, and the specimen 2 is
This is an element that mutually converts acoustic signals and electrical signals to transmit and receive ultrasonic waves. The test material 2 in this example is made up of two left and right steel plates welded together at the center, and the thickened part at the center is the welded part.

In FIG. 6, in order to detect the welded part of the test material 2 by the angle angle flaw detection method, the angle probe 10 is installed at a position slightly away from the welded part (in this example, on the right side of the welded part), and the pulse The transducer 12 is externally driven by a wave or a burst wave, and the ultrasonic wave is incident on the specimen 2 through the wedge member 12 at the incident angle α. When ultrasonic waves are incident on the test material 2 from this wedge material 12, the ultrasonic waves are refracted at the interface between the two. In this case, the refraction angle θ is theoretically determined by the ultrasonic propagation velocity in both media (acrylic and steel on the upper side) and the incident angle α according to Snell's law.

The maximum energy of the ultrasonic wave propagates at this theoretical refraction angle θ (that is, the sound pressure is at its maximum), but the actual refraction of the ultrasonic wave is centered around this refraction angle θ, and the ultrasonic wave has a certain spatial spread. Formed as a sound beam. FIG. 6 shows how this ultrasonic beam propagates within the specimen 2. If there is a reflection source with a different acoustic impedance (for example, a hollow part due to poor welding) within the test material 2, the ultrasonic wave that has been refracted and incident on the test material 2 will be reflected at the boundary surface (in the previous example, between metal and air). A part of the incident wave is reflected from the boundary surface), passes through the same path as the incident path, is detected by the transducer 12, and is output from the angle probe 10.

The location of defects such as poor welding is determined by the theoretical refraction angle θ(
This is the angle of refraction at which the maximum sound pressure is obtained because the maximum energy of the ultrasound propagates. In Figure 6, W is the beam path diameter from the ultrasound incident point to the defect, and p
When is the surface distance from the ultrasonic incident point to the defect (also referred to as PFD), and d is the depth from the detection surface to the defect, g and d are calculated by the following equations (1) and (2).

47-WΦsinθ-(1)d-W-
cosθ ・(2) and equation (1), (
Since the propagation velocity of W in the medium in 2) is known,
It is calculated from the propagation time.

However, as shown in FIG. 6, due to the spread of the ultrasonic beam, the refraction angle θ at which reflection from the defect is actually obtained.

may be different from the theoretical refraction angle θ, and in this case, the accuracy of measuring the defect position will be reduced.

[Problems to be Solved by the Invention] In measuring the position and shape of a defect using the conventional pulse reflection type ultrasonic flaw detection method as described above, it is difficult to actually Since the refraction angle at which reflection from the defect is obtained and the refraction angle at which the maximum received energy is obtained do not necessarily match, there is a problem in that the accuracy of measuring the position and shape of the defect is poor.

The present invention has been made to solve these problems, and provides an image processing method in an ultrasonic flaw detection test that can accurately measure the position and shape of a defect in a test material in a pulse reflection type ultrasonic flaw detection method. The purpose is to obtain.

[Means for Solving the Problems] In the first image processing method in the ultrasonic flaw detection test according to the present invention, in the pulse reflection type ultrasonic flaw detection test method, the image processing method in the vicinity of a place where a defect in the test material is assumed to be an ultrasonic flaw detection means that performs ultrasonic flaw detection using a probe at at least two or more measurement positions; Create an image pattern of each reflection source from the reflected echoes, then calculate a composite image pattern based only on the overlapping positions of the respective image patterns, and use the position and shape of the calculated composite image pattern as the position and shape of the defect. It is equipped with a signal processing means for measuring as follows.

In the image processing method in the second ultrasonic flaw detection test according to the present invention, in the pulse reflection type ultrasonic flaw detection test method, at least two or more measurement positions in the vicinity of a place where a defect in the test material is assumed to be An ultrasonic flaw detection means that performs ultrasonic flaw detection using a probe from above, measures the distance of each reflection source from the reflected echo obtained from the measurement position, and detects the reflection on the test material from each measurement distance and its measurement beam. Equipped with a signal processing means for approximating each of the estimated position ranges of the sources with circular arcs, then calculating the intersecting position ranges of the respective arcs, and measuring the calculated intersecting position ranges as the position and shape of the defect. It is.

In the image processing method in the third ultrasonic flaw detection test according to the present invention, in the pulse reflection type ultrasonic flaw detection test method, the probe is an ultrasonic flaw detection means that performs ultrasonic flaw detection, and a frequency distribution of reflected echoes obtained from the plurality of measurement positions on each axis of a three-dimensional coordinate position provided at a location where a defect in the test material is assumed. , and measures the position of the defect from the calculated data position near the maximum value of the frequency distribution on each axis, and measures the shape of the defect from the distribution shape of the frequency distribution on each axis. It is something.

[Function] In the present invention, as an image processing method in the first ultrasonic flaw detection test, a pulse reflection type ultrasonic flaw detection method is used to measure at least two or more places in the vicinity of a place where a defect is assumed to be in the test material. We perform ultrasonic flaw detection using a probe depending on the position.
Creating an image pattern of each reflection source from the reflected echo obtained from the measurement position at a coordinate position provided at a location where a defect in the test material is assumed by a signal processing means,
Next, a composite image pattern is calculated using only the overlapping positions of the respective image patterns, and the position and shape of the calculated composite image pattern are measured as the position and shape of the defect.

In addition, as an image processing method in the second ultrasonic test, ultrasonic waves using a probe are applied to at least two measurement positions near the location where a defect in the test material is assumed to be detected using a pulse reflection type ultrasonic flaw detection means. Flaw detection is carried out, and the signal processing means measures the distance of each reflection source from the reflected echo obtained from the measurement position, and the estimated position range of the reflection source in the material to be inspected is calculated from each measurement distance and the measurement beam. Then, the intersecting position range of each of the circular arcs is calculated, and the calculated intersecting position range is measured as the position and shape of the defect.

In addition, as an image processing method in the third ultrasonic flaw detection test, ultrasonic flaw detection is performed using a probe from multiple measurement positions near the location where defects in the test material are assumed to be present using pulse reflection type ultrasonic flaw detection means. , the signal processing means calculates the frequency distribution of reflected echoes obtained from the plurality of measurement positions on each axis of the three-dimensional coordinate position provided at the location where the defect in the test material is assumed, and the frequency distribution of reflected echoes obtained from the plurality of measurement positions is calculated. The position of the defect is measured from the data position near the maximum value of the frequency distribution on each axis,
Further, the shape of the defect is measured from the distribution shape of the frequency distribution on each axis.

[Example] Fig. 1 is a configuration diagram of an ultrasonic flaw detection system showing an example of the present invention. 1 is an ultrasonic probe, 2 is a test material, and in this example, a welded It is a steel plate. 3 is an ultrasonic flaw detector, 4 is a signal connection cable between the probe l and the ultrasonic flaw detector 3, 5 is a scanner controller, 6 is a personal computer (hereinafter referred to as CPU), and 7 is an X-Y scanner. , the probe 1 is held on an arm that moves in the X-axis direction and the Y-axis direction, and the probe 1 is scanned over the surface of the material 2 to be inspected based on control signals from the scanner controller 5 to perform flaw detection. It is a drive mechanism that allows

FIG. 2(a) is a diagram illustrating the scanning of a probe on a material to be inspected, where 2 is the material to be inspected and 10 is an oblique probe. In the figure, the angle probe 10 is scanned for a certain distance along the Y-axis in the direction in which the coordinate value increases on both sides of the welded part of the test material 2, and then in the X-axis direction by the scanning line interval. By repeating the operation of moving the object by the same amount and then scanning the same distance along the Y axis in the direction in which the coordinate value decreases,
All areas requiring flaw detection and inspection are scanned at predetermined scanning line intervals (for example, 3 to 5 intervals).

FIG. 2(b) is a diagram illustrating detection of reflected echoes in oblique flaw detection. The ultrasonic wave that has now been refracted at the refraction angle θ and incident on the test material 2 expands into a range of θ+Δθ to θ−Δθ in the cross section of the Y-Z axis due to beam expansion. If a reflection source (an interface with different acoustic impedance, for example, an interface between metal and air) exists at a distance of the beam path diameter W from the incident point, a reflected echo from this reflection source is detected.

The operation of FIG. 1 will be explained with reference to FIGS. 2(a) and 2(b). The CPU 6 provides the scanner controller 5 with a scanning program for the preliminary probe 1. This probe 1
For example, as shown in FIG. 2(a), the scanning program scans a predetermined area on both sides of the welded part of the inspected material 2, so the scanning program includes a scan start position, a scan end position, a scan line interval, a scan speed, etc. This is a program containing the necessary data. The scanner controller 5 controls the X-Y scanner 7 according to the given scanning program to cause the probe 1 to scan. - As shown in Figure 2 (a), the boat fishing method scans a certain distance from the scanning start position in the direction in which the coordinate value increases, for example along the Y-axis, and then scans in the X-axis direction. After moving by the line interval, the operation of scanning the same distance along the Y-axis in the direction in which the coordinate value decreases is repeated until the scanning end position is reached. During this scanning of the probe 1, the ultrasonic flaw detector 3
As shown in FIG. 2(b), the ultrasonic beam is incident on the specimen 2, and echo signals reflected from internal defects are received. Then, a beam path diameter signal W and amplitude data obtained by quantizing the amplitude of the received echo using binary or multivalued values are supplied to the CPU 6 from the time until the received echo is obtained. At the same time, the scanner controller 5 supplies the x and y coordinate signals of the probe to the CPU 6 at the time when the received echo was obtained.

In this way, the CPU 6 uses the beam path diameter signal W of the received echo from the reflection source obtained at each xsY coordinate position of the probe 1 to determine the target material 2 for each X coordinate where the probe is scanned. The coordinate values of the reflection source in the cross section of Stores received echo amplitude data.

The CPU 6 scans predetermined areas on both sides of the welded portion of the test material 2, acquires respective flaw detection data, and then performs image processing according to the present invention.

FIGS. 3(a) to 3(c) are diagrams for explaining the image processing method in the ultrasonic flaw detection test of the present invention. In each figure,
Test material 2 at a certain X coordinate position where the probe is scanning
A cross section, that is, a Y-Z axis plane, is shown, and an area near the central welding part is digitally subdivided to show the image pattern of the reflection source using the ySz coordinate positions of a large number of subareas. To perform signal processing of this image pattern, this >c
, y, z coordinate values are stored in the image memory (memory provided in the CPU 6 in the example), and this stored data is stored. It is read out and necessary arithmetic processing is performed.

In addition, in FIG. 3, for ease of explanation, the amplitude value of the reflected echo is binarized by comparing it with a predetermined threshold,
The coordinate position where there is a reflection source is a black image pattern, and the coordinate position where there is no reflection source is a white image pattern. However, since the reflected echo originally contains amplitude information proportional to the reflected intensity, the echo amplitude value may be stored in the memory as multi-bit amplitude data via an A/D converter. The image data in this case can be displayed as multi-tone data between black and white.

In FIG. 3(a) in this embodiment, the positions of defect data obtained by performing flaw detection from the right side of the welded portion are shown in black. Similarly, in FIG. 3B, the position of defect data obtained by performing flaw detection from the left side of the welded portion is shown in black. Third
Figure (C) is a diagram in which only the overlapping positions of the defect data in Figures (a) and (b) are extracted and image synthesis is performed. In other words, flaw detection is performed from both sides of the welded part, and a process is performed to extract only defect data that are detected at the same position in the test material (this signal processing involves AND processing of two image data). The resulting third
The parts displayed in black in Figure (e) are measured as the defect position and shape on the Y-Z axis plane.

Therefore, the image processing method in the first ultrasonic flaw detection test according to the present invention is a pulse reflection type ultrasonic flaw detection test method in which a defect is assumed to be present in the test material (in this example, a welded part).
At least 21 in the vicinity of! Ultrasonic flaw detection is performed using a probe from i or more measurement positions (in this example, on both the left and right sides of the welded part), and from the measurement position to the coordinate position provided at the location where the defect in the test material is assumed to be. Create an image pattern of each reflection source from the obtained reflected echoes, then calculate a composite image pattern based only on the overlapping positions of the respective image patterns, and use the position and shape of the calculated composite image pattern as the position of the defect. and shape.

Furthermore, when the image patterns are synthesized as is as shown in FIGS. 3(a) to (C), an image memory is required in the CPU 6, so the image pattern is approximated by a simple figure, such as an arc, to eliminate defects. A simple processing method for calculating the position and shape may also be considered. This is referred to as the second image processing method.

That is, the image processing method in the second ultrasonic flaw detection test according to the present invention is a pulse reflection type ultrasonic flaw detection test method, in which at least two measurement positions in the vicinity of a place where a defect in the test material is assumed to be present are detected. Perform ultrasonic flaw detection using a glass probe, measure the distance of each reflection source from the reflected echo obtained from the measurement position, and calculate the estimated position range of the reflection source in the material to be inspected from each measurement distance and its measurement beam. are approximated by circular arcs, then the intersecting position ranges of the respective circular arcs are calculated, and the calculated intersecting position ranges are measured as the position and shape of the defect.

This second image processing method is a simplified version of the first image processing method, but since it does not require an image memory,
The cost of equipment using this second processing method is significantly reduced.

Further, a third image processing method according to the present invention, which improves measurement accuracy although signal processing is more complicated than the first and second processing methods, will be described next.

In reality, reflection source data obtained by scanning flaw detection with a probe is generated not only on the image side of the weld but also on one side, due to the spread of the ultrasonic beam, at the same x, y, z coordinate position many times. You will get it. In other words, when performing flaw detection by scanning the probe along the Y axis from a position far from the weld to a position near the weld at a certain X coordinate position, the probe will move to a position far from the weld due to the spread of the ultrasonic beam. Received echoes can be detected many times from the same reflection source while moving to a slightly closer position. Therefore, the received signal is
When converted into defect data, the same X% of the inspected material
The number of times defect data is detected at the y%Z coordinate position is important data when determining the position and shape of the defect. The third processing method according to the present invention focuses on this point and calculates the frequency distribution of defect data in each coordinate axis.

In other words, in the Z frequency distribution shown in FIG. 3(C), when the probe is scanned along the Y axis at the same -X coordinate position, the number of times defect data is detected in the Z axis direction at the same -X coordinate position is shows the distribution data. Therefore, the position of the maximum value of the frequency distribution on the Z-axis becomes the depth d of the defect. Also Y
The frequency distribution indicates distribution data of the number of times defect data is detected in the Y-axis direction at the coordinate position d on the Z-axis during the scanning flaw detection of the probe. Therefore, the position of the maximum value of the frequency distribution on the Y-axis may be considered as the X-coordinate position of the defect.

FIGS. 4(a) to 4(C) are diagrams for explaining the image processing method using the frequency distribution of the present invention, each with a distance of 82.
A Y-Z frequency distribution diagram, a Y frequency distribution diagram, and a Z frequency distribution diagram at On+m are shown.

In FIG. 4(a), the Y frequency distribution map and the Z frequency distribution map are each a distribution map based on data of 80% or more of the maximum value of the frequency distribution data. Y-Z frequency distribution map is Y-Z
In the axial plane, the position and shape of the defect is shown only based on data of 80% or more of the maximum value of the frequency distribution data. Using this YZ frequency distribution map of 80% or more, the defect position of the material to be inspected can be measured with high accuracy.

In FIG. 4(b), the Y frequency distribution map and the Z frequency distribution map are each a distribution map based on data of 50% or more of the maximum value of the frequency distribution data. The Y-Z frequency distribution diagram shows the position and shape of defects based on data of 50% or more of the maximum value in the Y-Z plane. This 50% or more Y-Z
An image pattern that is quite close to the actual defect shape in the Y-Z axis plane of the material to be inspected can be measured using the frequency distribution diagram.

In FIG. 4(c), each distribution map is created from all the detection data obtained by the scanning flaw detection of the probe.

In this case, the image pattern in the Y-Z frequency distribution diagram is measured as an image pattern that is slightly larger than the actual defect shape.

The CPU 6 in FIG. 1 uses the detection data obtained by scanning and flaw-detecting the material 2 to be tested according to a pre-built-in program, as explained in FIGS. 4(a) to (e). Various distributions such as Y-2 frequency are calculated, and various signal processing is performed based on these distribution data to accurately measure the position and shape of the defect in the welded part of the test material 2.

Furthermore, in order to accurately understand the shape of the defect, it is necessary to inspect cross-sectional images of the material to be inspected viewed from each direction.

FIGS. 5(a) to 5(e) are diagrams illustrating examples of cross-sectional images of the test material viewed from each direction, and (a) in the figure shows the test material 2.
(b) is an image of the C scope plane, (c) is an image of the B scope plane, (d) is an image of the ST scope plane. The image of
(e) shows the echo height (signal amplitude) on the ST scope surface. In this way, the defect shape can be accurately measured by inspecting cross-sectional images of the material to be inspected viewed from each direction.

In the embodiments shown in FIGS. 2(a) and 2(b), an example of the angle probe 10 was shown as an ultrasonic probe. those with different angles of refraction,
There are many types of probes, including those with multiple built-in transducers and those with variable refraction angles, and the optimal probe is selected and used depending on the shape of the defect in the material to be inspected.

If the shape of the material to be tested is complex, the material to be tested may be placed in a water tank and flaws detected using water as a medium using a vertical probe.

Therefore, although the above embodiments have been described with respect to the case of an oblique probe, the present invention is not limited to this, and can be applied to general ultrasonic probes including vertical probes. Exactly the same operation can be performed.

As explained in detail above, the image processing method in the third ultrasonic flaw detection test according to the present invention is a pulse reflection type ultrasonic flaw detection test method, in which a plurality of images are detected in the vicinity of a location assumed to be a defect in the test material. Ultrasonic flaw detection is performed using a probe from the measurement position, and the reflected echoes obtained from the plurality of measurement positions are measured on each axis of the three-dimensional coordinate position set at the location where the defect in the test material is assumed to be. The frequency distribution is calculated, the position of the defect is measured from the data position near the maximum value of the calculated frequency distribution on each axis, and the shape of the defect is measured from the distribution shape of the frequency distribution on each axis.

[Effects of the Invention] As described above, according to the first image processing method of the present invention,
In the pulse-reflection ultrasonic testing method, ultrasonic flaw detection is performed using a probe from at least two measurement positions near the location where defects in the test material are assumed to be present, and the defects in the test material are detected. Create an image pattern of each reflection source from the reflected echoes obtained from the measurement position at coordinate positions provided in the assumed location, and then calculate a composite image pattern based only on the overlapping positions of each of the image patterns. By doing this, the position and shape of the defect in the material to be inspected is measured, so that the accuracy of measuring the position and shape of the defect can be improved compared to conventional methods.

Further, according to the second image processing method of the present invention, the first
Since the image memory is omitted as a simple processing method in which the image pattern in the processing method is approximated by a circular arc, it is possible to reduce the cost of an ultrasonic flaw detection apparatus using this processing method.

Further, according to the third image processing method of the present invention, the first
In this processing method, ultrasonic flaw detection is performed from a large number of measurement positions on a location where a defect is assumed to be present in the material to be inspected, and each of the three-dimensional coordinate positions provided at the location where a defect is assumed to be present in the material to be inspected is detected. In each axis, the frequency distribution of reflected echoes obtained from the plurality of measurement positions is calculated, the position of the defect is measured from the data position near the maximum value of the calculated frequency distribution in each axis, and the frequency distribution in each axis is calculated. Since the shape of the defect is measured from the distribution shape of the distribution, the first and second
It is possible to obtain the effect that the measurement accuracy of the position and shape of a defect is further improved than that of the image processing method described above.

[Brief explanation of drawings]

Fig. 1 is a block diagram of an ultrasonic flaw detection system showing an embodiment of the present invention, Fig. 2(a) is a diagram illustrating the scanning of a probe on a test material, and Fig. 2(b) is an oblique view. Figures illustrating the detection of reflected echoes in corner flaw detection, Figures 3 (a) to (c) are diagrams explaining the image processing method in the ultrasonic flaw detection test of the present invention, and Figures 4 (a) to (c) are diagrams explaining the detection of reflected echoes in corner flaw detection. Figures 5(a) to 5(e) are diagrams illustrating the image processing method using the frequency distribution of the present invention; Figures 5(a) to 5(e) are diagrams illustrating examples of cross-sectional images seen from each direction of the specimen; Figure 6 is the conventional ultrasound It is a figure explaining the flaw detection method. In the figure, 1 is a probe, 2 is a test material, 3 is an ultrasonic flaw detector, 4 is a connection cable, 5 is a scanner controller,
6 is the CPU, 7 is the X-Y scanner, lO is the angle probe,
11 is a wedge material, and 12 is a vibrator.

Claims (3)

[Claims] (1) In a pulse reflection type ultrasonic flaw detection test method, ultrasonic flaw detection is performed using a probe from at least two or more measurement positions near the location where a defect in the test material is assumed, and the test material is An image pattern of each reflection source is created from the reflected echo obtained from the measurement position at the coordinate position provided at the location where the defect is assumed, and then a composite image is created using only the overlapping positions of the respective image patterns. An image processing method in an ultrasonic flaw detection test, comprising calculating a pattern, and measuring the position and shape of the calculated composite image pattern as the position and shape of a defect. (2) In a pulse-reflection ultrasonic flaw detection test method, ultrasonic flaw detection is performed using a probe from at least two measurement positions near the location where a defect in the test material is assumed, and The distance to each reflection source is measured from the obtained reflected echoes, the estimated position range of the reflection source on the test material is approximated by an arc from each measurement distance and the measurement beam, and then the intersection of each of the abovementioned arcs is approximated. 1. An image processing method in an ultrasonic flaw detection test, characterized in that the calculated intersecting position range is calculated as the position and shape of a defect. (3) In a pulse-reflection ultrasonic flaw detection test method, ultrasonic flaw detection is performed using a probe from multiple measurement positions near the location where a defect in the test material is assumed, and the defect in the test material is assumed to be present. Calculate the frequency distribution of reflected echoes obtained from the plurality of measurement positions on each axis of the three-dimensional coordinate position provided at the location where the An image processing method in an ultrasonic flaw detection test, characterized in that the position of the defect is measured, and the shape of the defect is measured from the distribution shape of the frequency distribution in each of the axes.