JPS5824728B2 - Defect length measurement method and device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method and apparatus for accurately measuring the size of a defect using ultrasonic flaw detection, and more specifically, the present invention relates to a method and apparatus for accurately measuring the size of a defect using ultrasonic flaw detection. The present invention relates to a method and apparatus for correcting the apparent length of a defect and accurately measuring the defect length.

Conventionally, as a method for measuring defect length using ultrasonic flaw detection, an image of the defect is displayed on an X-Y display corresponding to the scanning plane of a probe, such as the C-scope method, and the image on the display is There is a way to measure the dimensions of.

However, with this method, once the image is displayed on the display, humans have to measure the dimensions of the image, so it takes time to sort out the flaw detection results. There were drawbacks such as the inability to display numerical values.

A method to solve this problem is to detect the reflected waves from the defective part while scanning the ultrasonic probe with a pulse motor, and then count the driving pulses of the pulse motor while the reflected waves are being detected and absorbed. proposed a method for measuring the size of defects (Japanese Patent Application No. 39347/1982).

Figures 1 and 2 are diagrams for explaining the principle.
FIG. 1 is a diagram showing the arrangement of a sample having a defect and a probe, and FIG. 2 is a diagram showing the scanning path of the probe and the position of the defect on the scanning surface.

In FIG. 1, a sample 1 with a defect 2 inside is shown.
02, when the probe 101 is scanned along the sample surface, when the ultrasonic beam 1 emitted from the probe 101 is incident on the defect 2, a reflected wave from the defect is detected.

The probe 101 scans along a path 103 shown in FIG.

Scanning in the X direction and the Y direction is performed by driving separate pulse motors.

The scanning distances in the X and Y directions per drive pulse of each pulse smoke are ΔX and ΔY, respectively.

Further, the scanning of the probe 101 from the scanning position al to a, / is performed by driving the X-direction scanning pulse motor with Cmax drive pulses.

At this time, for example, when flaw detection is performed on the scanning line from al to a, /, the reflected wave from defect 2 is detected for the first time at position b, and after b2, the reflected wave from defect 2 is no longer detected.

Here, 2' indicates the outline of the defect 4 projected onto the scanning plane.

The distance from the position to bl can be determined from the number of drive pulses of the X-direction scanning pulse motor in that section.

Similarly, in flaw detection from position a2' to a2, the distance from position b2' to b2 is determined, and compared with the distance from b, to b,/, the longer value is stored in the memory.

By sequentially scanning in this manner, the maximum length of the defect, in the case of FIG. 2, the distance from b4' to b4 can be determined.

The reflected wave from the defect was detected in the scan from position a1 to ai, but the reflected wave is no longer detected in the scan from a1+1 to a'i+1, so it is determined that the flaw detection for defective part 2' has been completed. can.

Since the number of driving pulses of the Y-direction scanning pulse motor up to this point is one, the length of the defect in the Y-direction is determined as i×ΔY.

By the way, the above defect length measuring method has a problem in that when the ultrasonic beam 1 is spread out, only the apparent length of the defect, which is different from the true length of the defect 2, can be measured.

For example, based on the reflected wave amplitude detected when an ultrasonic beam is incident on a defect surface of infinite size, the Consider the case where the defect length of defect 2 shown in FIG. 3 is measured.

In FIG. 3, x and y indicate the true length and width of defect 2, respectively.

A circle with a radius Δr indicates the size of the ultrasound beam 1.

A dotted line 3 indicates a boundary line where the amplitude of the reflected wave from the defect 2 matches the detection limit of the probe when the beam center is on this line.

In such a case, the defect length measuring method described in FIGS. 1 and 2 measures the apparent length X and width Y, both of which are longer than the length X and width y of the defect.

On the contrary, as shown in Fig. 4, when an extremely short defect 2 (length is the apparent size of the defect, and the apparent length X' of the defect is longer than the actual length X', and conversely, the apparent width Y' is shorter than the actual width y'.

In this way, the defect length measurement method proposed earlier is affected by the spread of the ultrasonic beam, the length and width of the defect, and the measured defect length is observed to be different from the true defect length. It was difficult to measure the length correctly.

An object of the present invention is to correct the apparent length and width of a defect, which is measured differently from the true defect length or width, by the above-mentioned ultrasonic beam spread, and to obtain accurate defect length and width. An object of the present invention is to provide a method and apparatus for measuring defect length.

The principle of the present invention will be explained below.

For example, it is well known that when a focused ultrasound probe is used, the ultrasound flux distribution on the focal plane becomes the distribution shown by the solid line 4 in Figure 5 with respect to the radial distance r from the beam center. It is being

This distribution can be expressed as follows.

Here, ■o is the sound flux at the beam center, and J I (r) is the first-order Bessel function of the first kind.

Further, Δr is called the Rayleigh resolution, which indicates the radial distance at which the ultrasound flux on the focal plane becomes zero, and is generally regarded as the radius of the beam.

This Δr is the aperture D of the probe and the focal length f of the acoustic lens.
, can be determined from the ultrasonic wavelength λ using the following equation.

The sound velocity distribution determined by equation (1) coincides within 3% with the normal distribution shown by the dotted line 5 in FIG. 5 within the beam radius 31.

Therefore, by approximating the sound velocity distribution as a normal distribution shown in FIG. 5, the relationship between the true length and apparent length of the defect can be expressed by an error function, as will be shown later.

Expressing the approximated sound velocity distribution as a formula, we get becomes.

Approximating the sound velocity distribution of the ultrasonic beam using equation (3), and assuming that the ultrasonic reflectance of the defect surface is uniform, the amplitude of the reflected wave from the defect is the acoustic flux of the ultrasonic beam incident on the defect surface. is proportional to.

As shown in FIG. 6, from one point of defect 2 with area S, ro
The reflected wave amplitude Ca from the defect 2 when the center of the ultrasonic beam 1 with a radius Δr exists at a position separated by Δr is expressed as follows.

Here, I (r) is the sound flux distribution of the ultrasonic beam 1, and ε is the ultrasonic reflectance of the defect surface.

By the way, in many cases in ultrasonic flaw detection, the defect size is evaluated by assuming that the defect shape is rectangular or circular.

For example, the defect 6 shown in FIG. There is.

Therefore, consider a case where the defect shape is rectangular.

The reflected wave amplitude Ca from defect 1 when the center of the ultrasonic beam exists at position a shown in FIG.
), and by substituting this into equation (4), it can be expressed as a simple equation as follows.

, and ε is the ultrasonic reflectance of the defect surface.

Similarly, when the center of the ultrasonic beam is located at the position shown in FIG. 8, the reflected wave amplitude Cb from the defect I is as follows.

The reflected wave amplitude C6 of the ultrasonic wave when the ultrasonic beam is incident on an infinitely large defect surface is proportional to the total sound flux of the ultrasonic beam, and is therefore given by the following equation.

However, the apparent lengths X and Y of the defect shown in FIG. This is the length obtained by flaw detection using a probe of η.

Therefore, in equation (5) and equation (5)7, equation (6)
By substituting the equations, the relationship shown in the following equation can be obtained.

becomes.

In Equation (7) and Equation (7), Δr and η are determined by the probe used and are known. Therefore, if the apparent length X and width Y of the defect are measured,
7) The actual length x, width y, and force j of the defect can be calculated using equation (7)'.

FIG. 9 illustrates the relationship between the apparent length X and width Y of the defect and the actual length X and width y of the defect using equations (7) and (7)' above.

For convenience, Figure 9 shows the relationship when η = 0.1, where the vertical axis is the ratio between the defect width y and the ultrasonic beam diameter 2Δr, and the horizontal axis is the defect length X and the beam diameter 2Δr. shows the ratio of

The solid line is a curve whose parameter is the ratio between the apparent length X and the beam diameter 2Δr, and the dotted line is a curve whose parameter is the ratio between the apparent width Y and the beam diameter 2Δr.

For example, if the diameter of the ultrasonic beam is 1071 m, the apparent length 2Δr =0.45, y/2Δr
=0.7 is found, and the actual defect length X is 4.5w/l
, the width y is 771gl1t.

Similarly, the apparent length X of the defect is 9wl1. @If Y is 671gIt, the actual defect length X is 5.5
III! It.

The width y is 2 shoes.

Next, the relationship between the defect radius and the apparent radius when the defect shape is circular will be described.

As shown in FIG. 10, the center of the defect 9 with radius R and the radius Δr
When the centers of the ultrasound beams are separated by R/, the reflected wave amplitude Ca is expressed as follows from equations (3) and (4).

In the above equation, if R' is the apparent radius of the defect, then C
a corresponds to the detection limit amplitude η, and the relationship expressed by the following equation is obtained.

Therefore, when the relationship between the apparent radius R of the defect and the actual radius is illustrated using equation (9), it becomes as shown in FIG.

Figure 11 shows the relationship between the actual radius of the defect and the apparent radius in the case of η-o,i, with the ratio R'/Δr of the apparent radius of the defect and the beam radius on the vertical axis, and the defect The horizontal axis is the ratio R/Δr of the radius of the beam to the radius of the beam.

For example, the table in Figure 11 shows that the diameter of the ultrasound beam is lQm.
In this case, if the diameter of the circular defect is 6 rings, the apparent diameter is 107×11t.

Therefore, by reversing the above procedure, the apparent diameter becomes l
In the case of Qa, the actual diameter of the defect is found to be 67 Mt using the table in FIG.

However, when measuring the apparent length and width of a defect, it is difficult to determine whether the shape of the defect is square or circular just by matching the length and width.

However, for example for a circular defect the diameter 1071
If the apparent length and width measured with an ultrasonic beam of 0.07l are each 1.071gh, it is regarded as a square defect and FIG. 9 is used.

The true length and width of the defect each obtain 5.771111.

On the other hand, if FIG. 11 is used, the true diameter of the defect will be 6 shoes, which is only Q, 37 MIt different from the case where FIG. 9 is used, and they match within 5%.

As described above, when the conversion table shown in FIG. 9 is used in the detection of circular defects, the true defect length can be determined with an error of less than 5%.

In this way, regardless of whether the defect shape is circular or rectangular, the defect length can be accurately determined from the apparent length and width of the defect using the conversion table shown in FIG.

In particular, the conversion curve shown in FIG. 9 normalizes the apparent length and width of the defect by the diameter of the ultrasound beam.

Therefore, even when the diameter of the ultrasonic beam is changed depending on the probe, or when the diameter of the ultrasonic beam is different in the length and width directions of the defect, such as in oblique angle detection, the same conversion curve can be used. It has the characteristic of being applicable.

Hereinafter, the defect length measuring method according to the present invention will be explained by giving an example.

Defects are detected using a probe with an ultrasonic beam diameter of 2Δr.

The scanning distance of the probe from the position where the reflected wave from the defect starts to be detected until it is no longer detected, that is, the apparent length X and width Y of the defect, is explained using, for example, Figures 1 and 2. Measure using the method described above.

Next, the ratios X/2Δr and Y/2Δr of the apparent length X and width Y of the defect to the diameter 2Δr of the ultrasonic beam are calculated.

Then, using the conversion table in Figure 9 prepared in advance,
Coordinates of the intersection of the solid line indicating the value of X/2Δr and the dotted line indicating the value of Y/2Δr (X/2Δr.

Find X/2Δr).

The values x/2Δr and X/2 found in this way
The correct length X and width y of the defect are determined by multiplying each Δr by the diameter 2Δr of the ultrasound beam.

For example, suppose that when the diameter of the ultrasonic beam is 101101l, the apparent length and width of the defect are 9 and 611a, respectively.

The coordinates (0,65,0,20) of the intersection of the solid line at 0.9 and the dotted line at 0.6 are determined from the conversion curve in FIG.

Therefore, in this case we obtain 6.5 shoes and 2 wIt as the correct length and width of the defect, respectively.

Now, store the conversion curve shown in Figure 9 in a secondary matrix in the storage device in advance, and read out the previously stored values from the address corresponding to the apparent length and width of the measured defect. The exact length and width of the defect can be determined automatically.

An example of the address structure and storage contents of the storage device described above is as follows.

The address structure of the storage device is a two-dimensional matrix with i rows and j columns.

For example, when the layout is 150 rows x 150 columns, the i row and 3rd column are the apparent length Xi and width Yj of the defect, respectively.
and are associated as shown in the following formula.

In the above formula, [ ] indicates a Gauss symbol, and indicates that the decimal part of the value in [ ] is rounded down to make it into an integer.

Further, 2Δr is the diameter of the ultrasonic beam. Address (i, j
) stores a 6-digit integer Mi,j.

The integers Mi and j stored at addresses (i, j) are 3-digit integers Mi and M that represent the length and width of the defect, respectively.
Create from i using the following formula.

The three-digit integers Mi and M· in the above equation are determined using, for example, the conversion curve shown in FIG.

That is, find the defect lengths Xi and yj corresponding to the apparent lengths Xi and X of the defect using the conversion table shown in FIG.
Integers Mi and Mj are determined by the following equations.

However, [ ] is a Gaussian symbol, and 2Δr is the diameter of the ultrasound beam.

In other words, the 6-digit integer M7. stored at address (i, j) in the storage device. In j, the upper three digits of the integer Mi indicate the ratio between the actual length Xi and the ultrasonic beam diameter, which corresponds to the ratio between the apparent length of the defect Xl and the ultrasonic beam diameter 2Δr, and the lower three digits The integer Mi indicates the ratio between the actual width Xi and the ultrasound beam diameter, which also corresponds to the ratio between the apparent width Yj and the ultrasound beam diameter.

For example, in the conversion table of FIG. 9, the diameter of the ultrasonic beam is 2Δr=10Za.

The apparent length Xi and width Yj of the defect are each 0.9
and the true length Xi and width y of the defect corresponding to 0.6
are 0.66 and 0.18, respectively.

Therefore, the integer 06 is placed at address (90, 60) in the storage device.
You only need to memorize 6018, and the upper 3° digits are 0.
The true length Xi of the defect can be found from 66, and the true width y of the defect can be found from the lower three digits 018.

Following the same procedure, the number @Mi, j is the address (i, j)
By storing two pieces of information regarding the length and width of a defect in one address, it is possible to store two pieces of information in one numerical value.

FIG. 12 shows an embodiment of the defect length measuring device of the present invention, which is equipped with the storage device described above.

In FIG. 12, the block 10 is designed to measure the apparent length of the defect Xi % by the method described in FIGS. 1 and 2, etc.
This device measures the apparent width Y, and outputs the apparent length Xi and width Yj measured by the device 10 as digital quantities to the arithmetic devices 11 and 11', respectively.

The arithmetic units 11 and 11' calculate the (
As shown in equations 10) and (10)', < 100Xi/2
Δr and 100Y·/2Δr are calculated, and integer values i and j are output to the reading device 13 by rounding down the decimal places of each value.

The reading device 13 reads the numerical values Mi, j stored at the address (i, j) of the storage device 14 and outputs them to the numerical value converting device 15.

In the numerical conversion device 15, an integer value Mi (as shown in equation (11)) obtained by dividing the input value M.
i is obtained by shifting Mi, j by 3 digits to the right) to the device 16
Further, the obtained Mi is multiplied by 1000, and the obtained integer value Mi is subtracted from Mi.
’.

Devices 16 and 16' are arithmetic devices that calculate their respective input values Mi using a value 2Δr set in device 12.
and M. are multiplied by 2Δr/100, and the values are output to the display device 1T.

The defect length display device 17 displays the output value of the device 16 as the true defect length X, and the output value of the device 16' as the defect width y.

Using the method of the present invention explained above, we measured the length X and width y of a rectangular plane defect made by electric discharge machining on a stainless steel plate (plate thickness 10wIt), and the measured values were as shown in the following table. Obtained.

In the above measurements, it is 6.9 mm in the length direction and 4.97 mm in the width direction.
The size of the defect appears to increase due to the spread of the ultrasonic beam using a column-shaped ultrasonic beam that spreads to 11M, so as shown in the table above, the conventional defect length measurement method can The actual defect length is overestimated by a factor of three.

On the other hand, in the method of the present invention, since the length is corrected from the apparent length of the defect by the length measured largely due to the spread of ultrasound, accurate defects with an error within 10 factors of the ultrasound beam diameter are displayed. I was able to measure the length.

As described above, the features of the defect length measuring method and apparatus of the present invention are as follows.

1) The accuracy of defect length measurement is greatly improved compared to conventional defect length measurement methods and devices.

2) Even if the diameter of the ultrasound beam changes, the same method and equipment can be applied as is.

3) Similarly, it can be applied to cases where the diameter of the ultrasonic beam is different in the length direction and width direction of the defect, such as in oblique angle flaw detection.

[Brief explanation of the drawing]

Figure 1 shows the arrangement of a sample with a defect and the probe, Figure 2 explains the conventional defect length measurement method, and Figures 3 and 4 show the size of the defect and the spread of the ultrasonic beam. Figure 5 shows the relationship between the area where the reflected wave from the defect is detected with an amplitude above the detection limit of the probe, and the sound pressure distribution in one radial direction on the focal plane of the ultrasonic beam. Figure 6 is a diagram showing the positional relationship between the defect and the ultrasonic beam, Figure 7 a and b
Figure 8 shows the actual defect shape and the size when the shape is approximated as a rectangle or a circle. Figure 8 shows the size of a rectangular defect, the spread of the ultrasonic beam, and the reflected wave from the defect at the detection limit of the probe. A diagram showing the relationship between the areas detected with the above amplitudes, and FIG. 9 is a curve diagram showing the conversion for determining the true defect length from the apparent defect length by the defect length measurement method of the present invention.
Figure 10 shows the relationship between the size of a circular defect, the spread of the ultrasonic beam, and the area where the reflected wave from the defect is detected with an amplitude above the detection limit of the probe. Figure 11 shows the radius of the circular defect. FIG. 12 is a diagram showing the structure of an embodiment of the defect length measuring device of the present invention. DESCRIPTION OF SYMBOLS 1... Ultrasonic beam, 2... Defect, 10... Conventional defect length measuring device, 11.11'... Arithmetic device, 12
...Gain setting device, 13...Reading device, 14...
・Storage device, 15...Numeric value conversion device, 16.16'・
... Arithmetic device, 17... Display device.

Claims (1)

[Claims] 1. Move the ultrasonic probe along the sample surface in the X-axis direction and in the Y-axis direction.
The apparent size of the defect in the X-axis direction and Y-axis direction is calculated from the scanning distance of the probe, which is scanned two-dimensionally in the axial direction, from when the reflected wave of the ultrasonic beam from the defect starts to be detected until it is no longer detected. The measured value is normalized by the size of the spread of the ultrasonic beam in the X-axis direction and the Y-axis direction. A defect length measuring method characterized in that the true defect lengths in the X-axis direction and the Y-axis direction are determined by performing correction based on the spread of the ultrasonic beam using a conversion table. 2 Move the ultrasonic probe along the sample surface in the X-axis direction and Y-axis direction.
A measurement device that measures the apparent size of a defect by scanning two-dimensionally in the axial direction, and an apparent defect length that is measured differently from the true defect length due to the spread of the ultrasonic beam of the probe. a storage device that stores a conversion table for the X-axis direction and the Y-axis direction for correction; and a storage device that stores a conversion table for the X-axis direction and the Y-axis direction for correction; a first arithmetic device for normalizing the outputs in the X-axis direction and the Y-axis direction from the storage device; a reading device that reads out conversion data; and a reading device that matches the standardized values in the X-axis direction and Y-axis direction with the conversion data in the X-axis direction and Y-axis direction to identify true defects in the X-axis direction and Y-axis direction. A defect length measuring device comprising: a second arithmetic device for calculating the length.