JP2010054292A - Method of measuring internal defect - Google Patents

Abstract

An internal defect measurement method capable of measuring and evaluating an internal defect of a structure quantitatively and with high accuracy is provided.
This internal defect measurement method includes a detection data collection step of collecting detection data by scanning the outer surface of a structure made of a ferromagnetic material with a detection sensor having magnetic flux detection means having a plurality of channels (S10). ), A phase detection processing step (S20) for performing filtering processing after multiplying the collected detection data by a reference signal, and a coordinate conversion step (S30) for performing coordinate conversion of vector coordinates on the obtained signal, A thinning portion detection step (S40) for detecting a thinning portion of the structure from the coordinate-transformed signal, and a defect diameter estimation step (S40) for estimating the defect diameter of the structure from the way of attenuation between each channel of the detection sensor ( S50) and a remaining thickness calculation step (S60) for calculating the remaining thickness of the structure from the obtained defect diameter and a predetermined relational expression.
[Selection] Figure 2

Description

  The present invention measures and evaluates the size and depth of internal defects of a structure having a plate shape, box shape, pipe shape, or the like made of a ferromagnetic material from the outer surface of the structure by an electromagnetic induction method. The present invention relates to an internal defect measurement method.

  Detecting internal defects in structures made of ferromagnetic materials and trying to evaluate them can be difficult or costly even if it is difficult to directly measure the inside of the structure. There are many cases. Therefore, an ultrasonic flaw detection method is generally used as a method for detecting and evaluating an internal defect of a structure by measurement from the outer surface of the structure. As an ultrasonic flaw detection method, for example, there is a technique described in Patent Document 1. In this technique, a plurality of ultrasonic probes are juxtaposed to detect defects existing in or on the surface of a steel material and multi-channel ultrasonic flaw detection is performed.

  However, in the ultrasonic flaw detection method shown in Patent Document 1, it is necessary to apply a probe to the outer surface of the structure. Therefore, rust on the surface to which the probe is applied, foreign matter such as dust, and coating depending on circumstances. A ground treatment for removing the film is necessary. Therefore, there is a problem that it takes a lot of time for measurement. Further, a contact medium is required between the structure to be measured and the measurement location is pinpointed (in Patent Document 1, this is achieved by arranging a plurality of ultrasonic probes in parallel. There are problems such as dealing with shortcomings).

On the other hand, an internal defect detection technique for a structure using an electromagnetic induction method has been proposed. As this type of internal defect detection technology, for example, there is a disclosure of Patent Document 2. The technique described in this document excites a steel material to be measured by two excitation means provided apart so that magnetic fluxes having opposite transmission directions and the same transmission magnetic flux density can pass through the steel material. The magnetic flux density leaking from the steel material at an intermediate position between the two excitation means is detected by the leakage magnetic flux density detection means to detect corrosion or cracking of the steel material.
Japanese Patent No. 3228132 Japanese Patent Laid-Open No. 11-44674

However, the method disclosed in Patent Document 2 does not need to remove foreign matter such as rust on the surface to which the probe is applied, dust, or a coating film, unlike the ultrasonic flaw detection method illustrated in Patent Document 1. However, it only detects corrosion or cracks in the steel material. That is, the size and depth of the internal defect of the structure cannot be quantitatively measured and evaluated, which is insufficient for evaluating the soundness of the structure.
Therefore, the present invention has been made in view of the above circumstances, without removing foreign matter such as rust and dust on the surface to which the probe is applied, or the coating film from the outside of the structure made of a ferromagnetic material. An object of the present invention is to provide an internal defect measurement method capable of quantitatively and accurately measuring the size and depth of an internal defect of a structure and evaluating it.

  In order to solve the above problems, a first invention of the present invention is an internal defect measurement method for measuring internal defects of a structure made of a ferromagnetic material using a low-frequency electromagnetic induction method, A detection data collection step of collecting detection data by scanning the outer surface of the structure with a detection sensor having an excitation device and a magnetic flux detection means having a plurality of channels, and a signal serving as a reference for the collected detection data And a phase detection processing step for performing a filtering process after separately multiplying signals each having a phase of a reference signal shifted by 90 °, and a zero point of a vector coordinate for drawing a signal obtained in the phase detection processing step. The coordinate conversion step of moving the coordinate axis in the major axis and minor axis direction of the ellipse of the drawn signal trajectory and moving the coordinate axis in the drawn signal locus, and the coordinate-converted detection data A defect diameter estimating step of estimating a defect diameter of an internal defect of the structure, and a remaining thickness calculating step of calculating a remaining thickness of the structure, the remaining thickness calculating step including the defect diameter For each defect diameter obtained in advance by an experiment using the defect diameter obtained in the estimation step and the object to be measured in which the pseudo defect portion corresponding to the obtained defect diameter is formed (original thickness × ellipse length) And the remaining thickness of the structure is calculated based on the relational expression of (remaining thickness / original thickness).

  The second invention of the present invention is an internal defect measuring method for measuring an internal defect of a structure made of a ferromagnetic material by using a low frequency electromagnetic induction method, wherein the external surface of the structure is measured. , A detection data collection step of collecting detection data by scanning with a detection sensor having an excitation device and a magnetic flux detection means having a plurality of channels, and a reference signal and a reference signal for the collected detection data After individually multiplying the signals whose phases are shifted by 90 °, the phase detection processing step for performing the filtering process, and the zero point of the vector coordinates for drawing the signal obtained in the phase detection processing step are moved to an arbitrary position. At the same time, a coordinate transformation process that rotates the coordinate axis in the major axis and minor axis direction of the ellipse of the drawn signal trajectory, and a steep rising portion of the coordinate-converted signal are detected. A defective portion detecting step for determining the presence or absence of a defective portion, and a remaining thickness calculating step for calculating a remaining thickness of the structure, wherein the remaining thickness calculating step is obtained in the defect diameter estimating step. (Original thickness × long ellipse length) and (remaining meat) for each defect diameter obtained in advance by an experiment using a measured object in which a pseudo defect portion corresponding to the obtained defect diameter was formed. The remaining thickness of the structure is calculated on the basis of the relational expression (thickness / original thickness).

According to the internal defect measuring method according to the present invention, by performing a predetermined process including the above-described series of steps, there is noise that draws a random signal locus in the detection signal, or a locus close to an ellipse is obtained. Even if it is drawn, it is possible to eliminate the influence of a disturbance (for example, a lift-off change between the detection sensor and the object to be measured) that swings in a direction unrelated to the defect. Therefore, the size of the internal defect (defect diameter) and the depth (remaining thickness) of the structure can be measured quantitatively and with high accuracy and evaluated.
Here, in the internal defect measurement method according to the present invention, the defect diameter estimation step obtains the size of the defect from the inclination of the attenuation amount of the detection data between the channels of the sensor or the detection data of the sensor in the scanning direction. It is preferable. Such a configuration is suitable for estimating the size of the defect quantitatively and with high accuracy.

  As described above, according to the internal defect measuring method of the present invention, foreign matter such as rust and dust on the surface to which the probe is applied or the coating film is removed from the outside of the structure made of a ferromagnetic material. Therefore, the size (defect diameter) and depth (remaining thickness) of the internal defect of the structure can be measured quantitatively and with high accuracy, and this can be evaluated.

First, the low-frequency electromagnetic induction method used in the internal defect measurement method according to the present invention and the electromagnetic induction device for performing the method will be described. FIG. 1 is a diagram showing a configuration example of an electromagnetic induction device for performing the low-frequency electromagnetic induction method. In the figure, reference numeral 10 denotes a detection sensor, S denotes an object to be measured (structure), K Indicates a defective part.
As shown in the figure, a detection sensor 10 of this electromagnetic induction device includes a detection coil 3 that is a magnetic flux detection means for detecting magnetic flux, and a U-shaped ferromagnetic core 4 around which an excitation coil 2 is wound. And an exciting device (magnetizer) 1. The detection sensor 10 includes an encoder (not shown) below, and a rolling wheel with a built-in bearing and a height adjustment that can be individually adjusted to maintain a constant lift-off between the detection sensor 10 and the object S to be measured. And a screw type adjusting mechanism. In this electromagnetic induction device, the ferromagnetic core 4 of the detection sensor 1 is disposed so as to face the object to be measured S, which is a ferromagnetic structure, and an excitation signal is applied to the excitation coil 2 to strengthen the electromagnetic sensor. Magnetic flux is generated in the magnetic core 4, and thereby leakage magnetic flux generated in the defective portion K of the object S to be measured is detected by the detection coil 3. In addition, in order to improve the measurement efficiency of the leakage magnetic flux which generate | occur | produces in the defect part K, this detection coil 3 has arranged 16 coils in a fixed pitch. In addition to the detection coil 3 exemplified above, for example, a magnetic sensing element can be used as the magnetic flux detection means.

  Here, the biggest difference between the general electromagnetic induction method and the low frequency electromagnetic induction method is the frequency of the excitation signal to be applied. That is, in a general electromagnetic induction method, the magnetic flux is concentrated on the surface layer portion of the object S to be measured by a high frequency excitation signal of several hundred kHz to several MHz. On the other hand, in the low frequency electromagnetic induction method, the frequency of the excitation signal to be applied is set to a low frequency of about several hundred Hz or less, thereby spreading the magnetic flux distribution in the plate thickness direction of the object S to be measured, Thereby, the sensitivity in the thickness direction of the measurement object S is improved.

Hereinafter, by using the above-described electromagnetic induction device, the internal defect measurement method according to the present invention detects the internal defect (for example, the corrosion thinning portion) of the measured object S (the structure made of a ferromagnetic material) and the portion thereof. A processing procedure for calculating the remaining thickness in FIG. FIG. 2 is a diagram showing an example of a processing flow of the internal defect measuring method according to the present invention.
As shown in FIG. 2, this internal defect measuring method includes a detection data collection step (step S10), a phase detection processing step (step S20), a coordinate conversion step (step S30), a defect portion detection step (step S40), a defect The diameter estimation step (step S50) and the remaining thickness calculation step (step S60) are executed in this order.

  Here, in the present embodiment, a general-purpose notebook PC and an AD converter are used as the detection data collection and analysis device. This general-purpose notebook PC is equipped with a series of automatic analysis processing programs including steps S10 to S60, and the thickness of the measured object S and the diameter of the defect portion K are automatically determined by the installed programs. Can be instantaneously calculated and the result can be displayed on the screen of a general-purpose notebook PC. In addition, for any other part of the object to be measured S (for example, a part not having the defect portion K), if the measurement / analyzer judges, an arbitrary position on the screen of the general-purpose notebook PC is clicked with the mouse. In the same manner as described above, the remaining thickness and diameter of the measurement object S at that position can be calculated and displayed.

  Moreover, in this embodiment, the several steel plate with a plate thickness of 6-19 mm was prepared as a test piece as a ferromagnetic structure. Then, for each steel plate, the length of the test piece is a flat bottom circular hole having a diameter of φ10 mm, φ20 mm and φ30 mm and having a remaining thickness of 2 mm (original thickness of the object to be measured S-2 mm) as a pseudo defect K. They were provided appropriately spaced in the direction. Then, these test pieces are used as the object to be measured S, and the defect portion K of the steel plate is sequentially sensed by the detection sensor 10, and the measurement accuracy by the internal defect measurement method according to the present invention using the detection signal from the detection coil 3. Evaluated.

  Specifically, in the detection data collection step (step S10), the detection sensor 1 shown in FIG. 1 described above is used, and the detection coil 3 scans the outer surface of the object S to collect the detection data. . The collection of the detection data at this time may be performed by performing a series of processing steps of the following steps S20 to S60 for each detection data, or once the detection data is temporarily stored in a storage medium. A series of processing procedures may be performed.

Next, in the phase detection processing step (step S20), predetermined phase detection processing was performed. FIG. 3 is a diagram for explaining the phase detection processing.
In this phase detection process, as shown in FIG. 3, after the detection signal from the detection coil 3 is multiplied by a reference signal (in many cases, an excitation signal is used) as a reference, filtering that is normally performed such as noise removal is performed. A signal (referred to as “Y signal”) that has undergone processing (for example, low-pass filtering) is acquired, and similarly, the phase of the reference signal serving as a reference is delayed by 90 ° with respect to the detection signal from the detection coil 3. A signal obtained by multiplication (referred to as “X signal”) is acquired. FIG. 4A shows an image obtained by plotting the Y signal and the X signal obtained by this phase detection processing on vector coordinates in which the vertical axis represents Y and the horizontal axis represents X.

Here, in order to measure the internal defect of a structure made of a ferromagnetic material using a low-frequency electromagnetic induction method, the present inventors have a flat-bottom hole-like pseudo-defect portion K having various dimensions. As a result of repeating the experiment using the object S to be measured, the following knowledge was obtained.
That is, as shown in FIG. 4A, the signal trajectory in the vector coordinates of the detection signal at the defective portion of the object S to be measured has an elliptical shape (Knowledge 1). If the object to be measured S is a plate, the inclination angle of the ellipse of the signal locus shown in FIG. 4A changes depending on the thickness (original thickness) of the plate. Are the same (the original thickness is the same), the inclination angle of the ellipse is substantially constant and does not change depending on the presence or absence of the defect K (Knowledge 2). Furthermore, the knowledge that the length of this ellipse changes with a certain correlation by the magnitude (defect diameter or its depth) of the defect part K (corrosion thinning part etc.) changed was acquired (knowledge 3).

  Therefore, in the processing procedure for detecting the defective portion K of the object to be measured S and calculating the remaining thickness at the portion in the present invention, the detection signal is obtained by executing a predetermined processing based on these findings 1 to 3. Even if there is noise that draws a random signal trajectory inside or a trajectory close to an ellipse, a disturbance that swings in a direction not related to a defect (lift-off change between the detection sensor 10 and the object S to be measured, etc.) The influence of can be eliminated.

  That is, in the subsequent coordinate conversion step (step S30), the zero point of the vector coordinates is moved to an arbitrary position (see FIG. 4B), and the thickness (original thickness) of the object S to be measured is further increased. If they are the same, the knowledge that the angle of the ellipse is almost constant is used, and the coordinate transformation of the vector coordinates obtained earlier, that is, the rotation of the coordinate axis, is performed in the major axis and minor axis directions of the ellipse. Is what you do. FIG. 4B plots the signal trajectory on the vector coordinates (Y ′ axis in FIG. 4A) obtained by moving and rotating the zero point in FIG. The image is shown. Hereinafter, the Y axis and the X axis that have moved and rotated the zero point are referred to as a Y ′ axis and an X ′ axis, respectively.

Next, in the defect detection step (step S40), when a signal path of a long ellipse is drawn in the vector coordinates in the defect part K (thinning part), the angle of the long ellipse is the original thickness of the object S to be measured. Focusing on the fact that they are constant if they are the same, a part satisfying the following (Equation 1) is defined as a defective portion K, and “corrosion thinning” is determined. This is to detect the maximum voltage (hereinafter referred to as “peak value”) of the detection coil 3 at the steep rising portion of ΔY ′ shown in FIG. It is set in advance depending on the target material of the measuring object S, a defective portion to be detected, and the like.
| ΔY ′ / ΔX ′ | ≧ constant (Equation 1)

Then, in the subsequent defect diameter estimation step (step S50), from the inclination of the attenuation amount of the signal between the channels of the detection coil 3 having a plurality of channels (16 coils), the “defect reduction” is performed in the defect portion detection step. The size of the defect diameter of the defect portion K determined to be “thick” is obtained.
That is, among the 16 coils of the detection coil 3, the coil that is closest to the center of the defect portion K and has a large peak value is set to 0 channel (ch), and the 1 channel, 2 ch,. When called, the peak value detected by each coil attenuates in order from 0ch. When it is plotted on the graph, it becomes as illustrated in FIG. The example shown in the figure is an object to be measured in which a pseudo-defect portion having a diameter of φ10 mm, φ20 mm, or φ30 mm is formed out of the object to be measured S having a flat-bottom hole-like pseudo defect portion K having various dimensions. In the example of S, the vertical axis represents the attenuation ratio of the peak value (the peak value of each coil / the peak value of the coil of 0ch), and in FIG. The horizontal axis indicates the number of channels (ch) of each coil.

As shown in FIG. 5, the rate of attenuation of the peak value between each channel (the slope of the graph in FIG. 5) indicates the defect diameter of the defect portion K (in the example of FIG. 5, the diameter of the pseudo defect portion is φ10 mm, φ20 mm, φ30 mm). ) According to each. Therefore, it is possible to estimate the defect diameter of the defect portion K from the way of attenuation of the peak value between each channel (that is, “the slope of the graph”).
In the present embodiment, the detection coil 3 is provided with multiple channels (16 coils), and the example in which the size of the defect diameter of the defect portion K is obtained from the slope of the attenuation of the signal between the channels has been described. For example, the size of the defect can be estimated quantitatively and with high accuracy by performing the same processing on the time-division signal in the scanning direction of the detection sensor 10 (detection coil 3).

  In the subsequent remaining thickness calculation step (step S60), the defect diameter of the defect portion K obtained in the defect diameter estimation step (step S50) and the object on which the pseudo defect portion corresponding to the obtained defect diameter is formed. Relationship between (original thickness of measured object S × long ellipse length (peak value of Y ′)) and (residual thickness / original thickness) for each defect diameter obtained in advance by an experiment using measured object S Based on the equation, the remaining thickness of the measurement object S is calculated.

  Here, the calculation of the remaining thickness in the remaining thickness calculation step (step S60) of the present embodiment is based on the following knowledge. That is, since the length of the ellipse (the peak value of Y ′) becomes higher as the remaining thickness becomes smaller, there is clearly a correlation with the thickness (Knowledge 4). Moreover, even if the remaining thickness ratio is the same, the longer the original thickness, the lower the ellipse length (the peak value of Y ′) becomes (Knowledge 5). Therefore, a more preferable approximate expression can be obtained by performing correction in consideration of this relationship.

  A part of the results relating to this finding is shown in Table 1 below. In the table, for each defect diameter (diameter φ10 mm, φ20 mm, φ30 mm) of the measurement object S in which the pseudo defect portion is formed, the original thickness (true value), the depth of the defect diameter (true value), Remaining thickness (true value), remaining thickness ratio (remaining thickness / original thickness), long ellipse length (peak value of Y ′) and (original thickness of measured object S × long elliptical length (Y ′ Peak value)).

  Further, FIG. 6 is a diagram that considers the correction in consideration of the above-described relationship, that is, “original thickness of measured object S × long ellipse length (peak value of Y ′)” and “remaining thickness / original thickness”. Shows the relationship. In addition, the same figure (a)-(c) makes each defect diameter (diameter φ10mm, φ20mm, φ30mm) a parameter, and there are three types of graphs for each defect diameter. It is an example of the results obtained by plotting the measurement results in Table 1 with “original thickness of the object S × long ellipse length (peak value of Y ′)” and “horizontal thickness / original thickness” on the horizontal axis. .

  FIGS. 7A to 7C are diagrams for comparison showing the relationship between the ellipse length (peak value of Y ′) and the dimensionless amount (defect diameter / remaining thickness). As in FIG. 6, FIG. 7 uses the defect diameter (diameter φ10 mm, φ20 mm, φ30 mm) as a parameter, and the vertical axis indicates “long ellipse length (Y ′ peak value)” and the horizontal axis indicates “remaining thickness / original thickness”. ”And the measurement results were plotted. The comparative example shown in FIG. 7 is based on the internal defect measurement method disclosed in a previous application (Japanese Patent Laid-Open No. 2006-208312) by the present applicant.

Here, it can be seen that any of the results shown in FIGS. 6A to 6C and FIGS. 7A to 7C show a high correlation with the linear approximation line shown in each figure. 6 and FIG. 7 are compared with each other, R ² (the square value of the correlation coefficient) is compared, and a linear approximation line shown in FIGS. 6A to 6C (that is, “the original thickness of the measured object S × The primary approximate line in the case of adopting “long ellipse length (peak value of Y ′)” is the primary approximate line shown in FIGS. 7A to 7C (that is, “long ellipse length on the vertical axis”). It can be seen that the measured value is more approximate to the true value than the (primary approximate line when “(peak value of Y ′)” is taken) (“R ² ” is between 0 and 1). The value is taken, and the closer this value is to 1, the closer the linear approximation line is to the true value.)

  Thereby, in this application, when calculating the remaining thickness in the remaining thickness calculation step (step S60), the first-order approximate straight line of FIG. 6 is adopted, which is estimated in the previous defect diameter estimation step. From the “defect diameter”, select one of the primary approximate lines in FIG. 6, and obtain “remaining thickness / original thickness” from this first approximate line, thereby further approximating the “remaining thickness” to the true value. ”Is being acquired.

  That is, in the first-order approximate straight line of FIG. 6 that responds to the estimated “defect diameter”, when FIG. 6A is selected, the “long ellipse length (the peak value of Y ′) that responds in FIG. Assuming that “x original thickness” is A, “remaining thickness / original thickness” B is obtained from the primary approximate line of FIG. Then, the “remaining thickness” to be calculated can be calculated by “B × original thickness”. Note that when selecting any of the first-order approximate straight lines in FIG. 6, each defect diameter (diameter φ10 mm) of the measurement object S in which the estimated “defect diameter” value is rounded to form a pseudo defect portion. , Φ20 mm, φ30 mm) is made to respond.

In the internal defect measuring method according to the present invention, FIG. 8 shows the result of evaluating the total detection error with respect to the “remaining thickness” value approximated to the true value. 6A shows the result of adopting the primary approximation line of FIG. 6 in the internal defect measuring method according to the present invention described above, and FIG. 7B shows the primary approximation of FIG. 7 as a comparative example. This is a result of adopting a straight line. As can be seen from FIG. 8, when comparing R ² (correlation coefficient square value) between FIGS. 8A and 8B, the R ² value of FIG. The result was better than the R ² value of FIG. In the internal defect measurement method according to the present invention, the total detection error is a remaining thickness of ± 20%, and the range in which automatic defect detection can be reliably performed is a plate thickness of 24 mm or less, The diameter was 10 mm or more and 30% depth or more of the plate thickness.

It is a figure which shows one structural example of the electromagnetic induction apparatus for performing the low frequency electromagnetic induction method used with the internal defect measuring method which concerns on this invention. It is a figure which shows an example of the processing flow of the internal defect measuring method which concerns on this invention. It is a figure explaining the phase detection process in the processing flow of FIG. FIG. 6A is a diagram in which signals after phase detection processing are plotted on vector coordinates. FIG. 5B is a diagram showing an image in which signal trajectories on the vector coordinates obtained by moving and rotating the zero point in FIG. It is a figure showing an example of the method of attenuation of the peak value between each channel of a detection coil. The figure which showed the relationship between "long ellipse length (peak value of Y ') x original thickness" and "(residual thickness / original thickness)" in the internal defect measuring method which concerns on this invention ((a)- (C)). 6 is a comparative example of FIG. 6 and is a diagram ((a) to (c)) showing a relationship between “long ellipse length (peak value of Y ′)” and “(remaining wall thickness / original thickness)”. is there. It is the figure which showed the relationship between the actual remaining thickness and the calculated remaining thickness, The figure (a) shows the result by the internal defect measuring method which concerns on this invention shown in FIG. ) Shows the results of the method of the comparative example shown in FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Excitation apparatus 2 Excitation coil 3 Magnetic flux detection means 4 Ferromagnetic core 10 Detection sensor S Measured object K Defect part

Claims (3)

An internal defect measurement method for measuring internal defects of a structure made of a ferromagnetic material using a low frequency electromagnetic induction method,
A detection data collection step of collecting detection data by scanning the outer surface of the structure with a detection sensor having an excitation device and a magnetic flux detection means having a plurality of channels, and a signal serving as a reference for the collected detection data And a phase detection processing step for performing a filtering process after separately multiplying signals each having a phase of a reference signal shifted by 90 °, and a zero point of a vector coordinate for drawing a signal obtained in the phase detection processing step. A coordinate conversion step of rotating the coordinate axis in the major axis and minor axis direction of the ellipse of the drawn signal trajectory and moving it to an arbitrary position, and the internal defect of the structure based on the coordinate-converted detection data A defect diameter estimating step for estimating a defect diameter, and a remaining thickness calculation step for calculating a remaining thickness of the structure,
In the remaining thickness calculation step, the defect diameter obtained in the defect diameter estimation step, and the defect diameter obtained in advance by an experiment using a measurement object in which a pseudo defect portion corresponding to the obtained defect diameter is formed. An internal defect measuring method, comprising: calculating a remaining thickness of the structure based on a relational expression between (original thickness × long elliptical length) and (remaining thickness / original thickness). An internal defect measurement method for measuring internal defects of a structure made of a ferromagnetic material using a low frequency electromagnetic induction method,
A detection data collection step of collecting detection data by scanning the outer surface of the structure with a detection sensor having an excitation device and a magnetic flux detection means having a plurality of channels, and a signal serving as a reference for the collected detection data And a phase detection processing step for performing a filtering process after separately multiplying signals each having a phase of a reference signal shifted by 90 °, and a zero point of a vector coordinate for drawing a signal obtained in the phase detection processing step. A defect is detected by moving to an arbitrary position and detecting a steep rising portion of the coordinate-converted signal and a coordinate conversion step of rotating the coordinate axis in the major axis and minor axis direction of the ellipse of the drawn signal locus. The defect diameter of the internal defect of the structure based on the defect detection process for determining the presence / absence of the defect and the detection data in the detected defect Includes a defect size estimation step of estimating, the remaining wall thickness calculating step of calculating a remaining wall thickness of the structure,
In the remaining thickness calculation step, the defect diameter obtained in the defect diameter estimation step, and the defect diameter obtained in advance by an experiment using a measurement object in which a pseudo defect portion corresponding to the obtained defect diameter is formed. An internal defect measuring method, comprising: calculating a remaining thickness of the structure based on a relational expression between (original thickness × long elliptical length) and (remaining thickness / original thickness). In the internal defect measuring method according to claim 1 or 2,
In the defect diameter estimating step, the defect diameter of the internal defect of the structure is estimated from the inclination of the attenuation amount of the detection data between the channels of the detection sensor or the detection data in the scanning direction of the detection sensor. Internal defect measurement method.

Images