JP2003270214A - Method for eddy current flaw detection and flaw detection probe - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for an eddy current flaw detection by which flaw inspection of a thick plate of a structure such as an atomic power plant and an aircraft can be executed, and detection of flaws not only on the surface but also on a rear face opposite to the inside and a flaw detection face can be executed, and a flaw detection probe which is optimum therefor. <P>SOLUTION: The method for the eddy current flaw detection is characterized in that oppositely directed currents are fed to two excitation coils, the eddy currents of the respective coils are overlapped in a region between the excitation coils, and the flaw detection is executed by a detection coil arranged between the excitation coils. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an eddy current flaw detection method and flaw detection probe capable of inspecting flaws on the front surface, the inside and the back surface of a thick plate metal material as well as a thin plate by eddy current flaw detection. The flaw described in this specification means a defect of a metal material that can be inspected by eddy current flaw detection such as a crack, a hole, a dent, and a nonmetallic inclusion, and all of these defects are targeted.

[0002]

2. Description of the Related Art In general, an eddy current flaw detection test, which is one of non-destructive electromagnetic applications, has 1) high surface sensitivity, 2) high-speed non-contact flaw detection is possible, and 3) sensor structure is simple and designed. 4) Easy to process and record because the signal is directly obtained as an electric signal 5) Suitable for automatic flaw detection / remote operation 6) Easy to be affected by noise factors such as change of material and shape , Has the characteristics. Although the eddy current flaw detection test has good surface sensitivity in principle, it is not suitable for flaw detection on the back surface of thick plates because the eddy current is attenuated in the thickness direction due to the skin effect. It has been used limited to.

An example of application of the eddy current flaw detection test is in-service inspection of a steam generator heat transfer tube of a pressurized water type LWR power plant. Especially regarding the flaw detection of the steam generator heat transfer tube, the eddy current flaw detection has made remarkable progress in the numerical analysis technology and the development of the new probe in the past ten years. Current,
In the flaw detection of a heat transfer tube with a plate thickness of 1.27 mm, it is possible to detect a backside 20% crack and reconstruct the crack shape by inverse problem analysis. It is expected that these technologies will be applied to other inspection targets in the future. For structures such as nuclear power plants and aircraft, it is desirable to perform nondestructive inspection based on the defect tolerance standard or damage tolerance design, and thus high defect detection capability and defect shape evaluation are required.

On the other hand, these structures have a large number of parts made of thick-walled materials instead of thin plates such as heat transfer tubes of steam generators. There are cracks in such places, such as welded parts of austenitic stainless steel, which are difficult to detect by the conventional ultrasonic flaw detection method, and application of electromagnetic non-destructive inspection is also being considered. Therefore, it is required to apply the eddy current flaw detection test to thick-walled materials and reflect the features in the inspection to improve the sophistication.

A problem in applying the eddy current flaw detection test to thick-walled materials is the detection of flaws on the back surface opposite to the flaw detection surface. In general, it is difficult to detect flaws on the back surface due to the attenuation of eddy current due to the skin effect, and it is necessary to overcome this problem. In addition,
Conventionally, the eddy current flaw detection test method has been applied to inspected objects with a plate thickness of 1 to 1.5 mm, but if the thickness exceeds that, especially with a plate thickness of 6 mm or more, it is possible to inspect for scratches by eddy current. It was practically difficult.

[0006]

DISCLOSURE OF THE INVENTION The present invention is capable of inspecting a thick plate for a structure such as a nuclear power plant or an aircraft, and checking not only the front surface but also the inner and back surfaces (inner surface), that is,
An object of the present invention is to provide an eddy current flaw detection method capable of detecting a flaw on the back surface opposite to the flaw detection surface and a flaw detection probe most suitable for the flaw detection method.

[0007]

In order to solve the above-mentioned problems, the inventors of the present invention have conducted diligent research, and as a result, improved the structure of the eddy current flaw detection probe so that only the surface of the material to be inspected However, it was found that the detection of scratches inside and especially on the back surface opposite to the flaw detection surface can be more accurately measured. The flaw detection method of the present invention is particularly effective for flaw inspection of thick plate materials, but it goes without saying that it can also be applied to thin plate materials. The present invention is based on this knowledge, and applies currents in opposite directions to 1.2 exciting coils, superimposes eddy currents of the respective coils in a region between the exciting coils, and arranges the detecting coils arranged between the exciting coils. 1. Eddy current flaw detection method characterized by flaw detection by 2. The eddy current flaw detection method according to the above 1, wherein the detection coil is a differential coil. The distance between the exciting coils is set so that the ratio of the back surface eddy current density and the surface eddy current density directly below the detection coil is 0.5 or more (eddy current density / surface eddy current density). 3. The eddy current flaw detection method described in 1 or 2 above. The ratio of the inner radius r of the exciting coil to the height h is r / h =
1 to 1 above, which is in the range of 0.5 to 1.5
4. Eddy current flaw detection method described in each of 3. 5. The eddy current flaw detection method according to each of 1 to 4 above, wherein the surface of the plate and the surface of the plate are inspected for scratches on the inside and the back of the plate. 7. The eddy current flaw detection method according to each of 1 to 5 above, wherein a flaw inspection of a thick plate having a plate thickness of 6 mm or more is performed. The eddy current flaw detection method according to each of the above 1 to 6, wherein the crack shape is reconstructed by an inverse problem analysis method. I will provide a.

The present invention also relates to 8. 8. An eddy current flaw detection probe characterized by comprising two exciting coils for passing currents in opposite directions and a detecting coil arranged in a region between the exciting coils at a position where eddy currents caused by the exciting coils overlap each other. 10. The eddy current flaw detection probe described in the above 8, wherein the detection coil is a differential coil. The ratio of the back surface eddy current density and the surface eddy current density directly below the detection coil is 0.5 (back surface (eddy current density / surface eddy current density)).
11. The eddy current flaw detection probe according to the above 8 or 9, wherein the exciting coil is installed at the above-mentioned separated positions. The ratio of the inner radius r of the exciting coil to the height h is r / h =
The range from 0.5 to 1.5, and the range from 8 to
12. The eddy current flaw detection probe described in each of 10. 13. The probe for eddy current flaw detection according to each of 8 to 11 above, which is for inspecting the surface of the plate and the inside and the back of the plate for flaws. The probe for eddy current flaw detection according to each of the above 8 to 12, wherein a flaw inspection of a thick plate having a plate thickness of 6 mm or more is performed.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention proposes an eddy current flaw detection probe capable of detecting a flaw on the back surface of a thick-walled material with the aim of expanding the field of application of the eddy current flaw detection test. For this purpose, the modified magnetic vector potential method (A.Kameari, Solution of Asy
mmetric Conductor with a Hole by FEM Using Edge-el
ement, COMPEL, 9, (1999), pp.230-232.) and edge element finite element method (A.Kameari, Three Dimensional Eddy Curren
t Calculation Using Edge Element for Magnetic Vect
or Potential, Applied Electromagnetics in Material
s, pp.225-236, (1988).), based on the eddy current analysis, various parameters of the probe were set, and the design and prototype were carried out. Verify the performance of the probe developed by experiment,
Inverse problem analysis was performed using the obtained results, and quantitative shape evaluation of cracks was performed. The design of the probe and the experimental results of the probe will be described below. Moreover, the inverse problem analysis result with respect to the experimental result is shown, and the result is similarly described below.

When applying an eddy current flaw detection test to a thick plate, a problem is the skin effect of the eddy current. Specifically, it is considered that the flaws on the back surface are hindered by the following items. That is, 1) it is difficult to obtain a strong eddy current for detecting scratches on the back surface on the back surface due to the attenuation of the eddy current due to the skin effect in the plate thickness direction. 2) Due to the effect of the skin effect, the eddy current on the front surface causes the eddy current on the back surface. It is significantly larger than the current, and the defect signal on the back surface may be affected by noise due to changes in the surface shape or material.

The above problem can be solved by setting the frequency to an appropriate value for a thin plate.
With thick plates, it is difficult to solve two points at the same time. If the frequency is set low and the skin depth is increased, the eddy current becomes weak. Since the intensity of the defect detection signal is greatly influenced by the intensity of the eddy current induced in the location where the defect exists, the signal also weakens accordingly. On the contrary, if the frequency is set large and the strength of the eddy current is increased, the eddy current is significantly attenuated by the skin effect, and the defect signal on the back surface is greatly affected by the noise on the surface. Especially, this tendency is remarkable in the vicinity of the exciting coil. Therefore, not only the frequency adjustment but also the search for the probe shape capable of penetrating the strong eddy current flatly in the plate thickness direction is required for the flaw detection of the thick material.

In order to generate a strong eddy current on the back surface, in the present invention, currents in opposite directions are made to flow through the two exciting coils, and the eddy currents by the respective coils are superposed in the region between the exciting coils. This excitation method has a remarkable advantage that it can reduce excessive enlargement of the excitation coil and increase in lift-off accompanying it. This point is one of the major features of the present invention. The dimensions of the individual exciting coils were selected using three-dimensional eddy current analysis so that a strong eddy current can be generated on the back surface even when used alone.

On the other hand, in order to prevent the defect signal on the back surface from being strongly affected by the noise on the surface, a position where a strong eddy current can be flatly permeated is searched for, and the signal is detected there. This problem results in adjusting the distance between the excitation coils to change the eddy current distribution, since the detection coils must be located in the region between the excitation coils. Again, by using 3D eddy current analysis,
The distance between the exciting coils was selected. FIG. 1 shows a conceptual diagram of an eddy current flaw detection probe developed with the above two points as a design policy. In FIG. 1, reference numeral 1 is an exciting coil, reference numeral 2 is a detection coil (two differential detection coils), reference numeral 3 is an electric current, reference numeral 4 is the front surface of the inspection object 10, and reference numeral 5 is the back surface of the inspection object 10. , Reference numeral 6 indicates a magnetic flux.

(Determination of probe dimensions by numerical analysis)
In the present invention, dimensional values were determined using three-dimensional eddy current analysis during probe development. In this analysis method, the edge element finite element method based on the modified magnetic vector potential method is adopted. By using edge elements, it is not necessary to remove the electrical scalar potential from the governing equations and to have simultaneous equations for eddy current divergence. Also,
By using the modified magnetic vector potential method,
The exciting coil can be treated independently of the conductor and the space surrounding it. Therefore, it is possible to obtain an effect of significantly reducing the storage capacity required for the numerical analysis. In addition to this, there is a high-speed solution method that applies a database to this analysis method, which was used for comparison and comparison of the experimental results and the analysis results described later. In the calculation of the eddy current flaw detection signal, this is to realize a high-speed calculation by creating a database and limiting the analysis region to a region where a crack is predicted.

A specific example of the present invention will be shown below. An eddy current flaw detection probe for thick plates was designed using INCONEL having a plate thickness of 7 mm as a test material (inspection object). The lift-off is 0.2 mm. When the frequency is 5 kHz, the skin depth is about 7 mm, but the signal strength was taken into consideration and set to 10 kHz for calculation. The current density of the exciting coil is 1.0 × 10 ⁶ [A
fixed to / m ² ]. Hereinafter, these values are common to all analyzes in this specific example.

There are many parameters for the exciting coil, but here we focused on the inner diameter of the coil (specifically, the inner radius is used), the winding width, and the height. The excitation coil (inner diameter 1 mm, winding width 0.5 mm, height 0.5 mm) of the existing probe was magnified at the same size, and a parameter survey was conducted on the expanded coil for the internal radius, winding width, and height. . FIG. 2 shows the dimensional shape of the exciting coil 1. In FIG. 2, reference numeral 7 indicates the height, reference numeral 8 indicates the winding width, and reference numeral 9 indicates the inner radius. Code 10
Is an object to be inspected. The winding width has a great influence on the back surface eddy current from the relationship between the change of each parameter obtained by the analysis and the back surface eddy current density. In this specific example, the winding width is set to 6 mm in consideration of spatial restrictions. However, this winding width can be arbitrarily changed according to the type and size (thickness) of the material to be inspected and the design of the eddy current flaw detection probe.

After selecting the design value of the winding width, the change of the back surface eddy current when the inner diameter and the height were changed at the same time was examined. The analysis result is shown in FIG. In the figure, it can be seen that the backside eddy current changes relatively steeply along the straight line of (inner diameter) / (height) = 1. Generally, it is desirable to further expand the exciting coil along this straight line, but the preferable range is (inner diameter) / (height) = 0.5-1.5. In consideration of these analysis results, in the present specific example, the dimensions of the exciting coil were finally set to an inner diameter of 10 mm, a winding width of 6 mm, and a height of 12 mm. The specific dimensions can be arbitrarily changed according to the type and size (thickness) of the material to be inspected and the design of the eddy current flaw detection probe. The number of turns was set to 1995 turns based on the cross-sectional area of the winding. By incorporating the above process in the process of implementation, it became possible to manufacture an exciting coil capable of easily inducing a strong eddy current on the back surface.

After the size of the exciting coil is determined, the distance between the two exciting coils is adjusted to change the eddy current density distribution. The change of the eddy current when the distance between the exciting coils was changed was investigated by three-dimensional eddy current analysis. Fig. 4 shows the relationship between the distance between the exciting coils, the eddy current densities on the front surface and the back surface, and their ratio (back surface eddy current density / surface eddy current density). It can be seen from FIG. 4 that when two exciting coils are brought close to each other, a large difference occurs in the eddy current densities on the front surface and the back surface. On the other hand, when the distance between the exciting coils is, for example, 10 mm or more, particularly 12 mm or more, the eddy current density ratio of (back surface eddy current density) / (surface eddy current density) increases slightly, but the gradient You can see that becomes slower. The backside eddy current density continues to decrease asymptotically. Considering both the strength of the backside eddy current and the skin depth,
In this example, the distance between the exciting coils was 12 mm.
Further, in this specific example, (back surface eddy current density) / (surface eddy current density) is 0.7, but when it is 0.5 or more, good sensitivity for back surface crack detection is obtained.

After setting various design values of the exciting coil, the design value of the detecting coil is selected. When designing the excitation coil, the detection intensity immediately below the detection coil is high, and the ratio of the back surface eddy current density to the surface eddy current density (back surface eddy current density / surface eddy current density) is 0.5 or more. The inner radius, height, and width of the exciting coil, and the exciting coil, so that the ratio of the inner radius r of the exciting coil to the height h is in the range of r / h = 0.5 to 1.5. Set the distance between. Generally, with respect to the detection coil, the greater the number of turns, the higher the sensitivity to cracks. The detection coil was designed by filling the area between the exciting coils and then expanding in the height direction. In this example, the height of the detection coil was finally 6 mm, and the number of turns was 1300. From the eddy current distribution by the probe determined by the above numerical analysis results,
The eddy current in the plane is shown when the plane of symmetry of the lobe is taken as the cut surface. Regarding the real part and amplitude of the eddy current induced in the DUT, (back surface eddy current density) / (surface eddy current) The current density) was 0.75 and 0.72, which were high values. The imaginary part is 0.35, but since the real part is more dominant, it is considered that it will not significantly affect flaw detection. Accordingly, even if the eddy currents on the front surface and the back surface are substantially in phase with each other, it can be predicted that the newly designed probe will not significantly hinder the detection signal on the back surface due to the noise on the front surface.

(Eddy current flaw detection system) The eddy current flaw detection system used in the specific example of the present invention was carried out by controlling a two-dimensional electric stage on which a test piece was mounted by a personal computer via a GPIB board. The probe is fixed in place.
The signal from the probe is processed by the flaw detector, and divided into a real part and an imaginary part, and is taken into the personal computer by the A / D board. ASSORT-PC2 made by Aswan Electronics was used as the flaw detector. In order to obtain discrete data, the stage control PC was used to move the stage while simultaneously capturing data such as flaw detection signals and coordinates of measurement points.

(Tests and Results) The test pieces used in this example were pipes of a purely domestic H-IIA rocket simulated by using the same material (INCONEL718) and a flat plate of the same plate thickness (7 mm). used. There was a weld line in the longitudinal direction at the center of the test piece, and the weld line had a surplus removed. For this 7 mm thick test piece, there are three semi-elliptical artificial cracks on the boundary between the weld line and the base metal. The crack length is 10 mm and the width is
It is 0.2 mm. When the short radius of the ellipse is the crack depth, the depth of each crack is 1.00 mm (14.3%), 0.50 mm (7.1
%) And 0.25 mm (3.6%). Internal cracks (ID) and external cracks (OD) can be obtained by detecting these cracks from the surface where cracks are open or from the surface where they are not open.
And the flaw detection. The settings of the flaw detector in the experiment were a test frequency of 10 kHz and a gain of 79 dB. The setting of stage control was such that the scan pitch was 0.5 mm in both X and Y directions, and the lift-off between the probe and the test piece was 0.2 mm.

(Comparison-Absolute value type coil) For comparison, an experiment was conducted using an absolute value type coil. The defect detection capability of this probe is 20% ID, OD6
It is 0%. In addition to 10 kHz, 5 kHz and 1 kHz were applied as test frequencies, and flaw detection was performed on external cracks with a depth of 1 mm. 10kHz, 5kHz,
It was confirmed that cracks could not be detected in any of the cases of 1 kHz. As a result, it can be said that the absolute value type coil is not suitable for flaw detection of thick-walled materials. Further, in this test piece, it can be seen that the changes in the physical properties such as the electric conductivity and the magnetic permeability between the welded portion and the base material are minute.

(Detection Results Using the Probe of the Present Invention) Using a probe which is a specific example of the present invention, an experiment of a semi-elliptical inner surface crack having a depth of 0.25 mm was conducted by the present probe. As detection conditions, frequency: 10 kHz, lift-off: 0.2 mm,
Phase: 283.0 degrees. As a result of two-dimensional display of Vy signal obtained by two-dimensional scanning, there are four peaks for one crack. This is because the differential between the two detection coils is used as the detection signal, and the probe is bilaterally symmetric and has a self-differential characteristic. The crack exists at the center of the four peaks, but the set of peaks appearing in the X direction across the crack is due to the differential characteristics of the detection coil.
On the other hand, the set of peaks appearing in the Y direction is due to the self-differential characteristic. In this experiment, the crack is also symmetrical, and the absolute value of the signal is the smallest at the center of the crack due to the self-differential characteristic.
Note that the flaw detection signal was set so as to appear significantly in the y direction, and a signal containing noise was obtained in the x direction. Similarly, for the external crack, four peaks can be confirmed.
Figure 5 shows the experimental results for an outer surface crack with a depth of 0.5 mm. Also,
Extract the B-scan signal along the line shown in the figure,
This is shown in FIGS. 6 and 7. Two positive and negative peaks can be confirmed in each of the two B scan signals. This shows that the developed probe was able to detect an external crack with a depth of 0.5 mm.

(Comparison of Numerical Analysis Results with Experimental Results) Under the same conditions as the experiments, eddy current flaw detection signals were calculated by forward problem analysis by the edge element deformation magnetic vector potential method using a database and compared with the experimental signals. Where the width of the artificial crack is
Since it is known to be 0.2 mm, it is sufficient to use a one-dimensional signal in the crack direction for crack discrimination. The crack signal becomes maximum when one of the two detection coils is on the extension line of the crack, so the center of the crack in the crack direction is 0 mm, and the center point of one of the detection coils is +17.5 mm to -17.5 mm. when moving up to mm
The signals of 22 one-dimensional data at 1 mm intervals (partial intervals of 2 mm) are used for the following comparison.

Since the exciting current and the phase difference due to the filter are unknown in the flaw detector output, it is impossible to compare the absolute value and the phase with the calculated signal. Therefore, first, it is necessary to convert the experimental result into a voltage value obtained from the calculation result. Rotational expansion is performed by the following equation so that the amplitudes and phases of the peaks of the experimental results and the calculated results match the signals of the previous 22 points. S ′ = αe ^jθ S (1) Here, S and S ′ are experimental signals before and after conversion, α is an expansion coefficient, and θ is a rotation angle. The coefficients were set as follows so that the experimental signal of a semi-elliptical inner surface crack with a depth of 1 mm would match the analytical signal. α = 2.15 (2) θ = −52.1 [degree]

The respective experimental results were converted using the coefficients of the equations (1) and (2). The corresponding calculation results and converted experimental results are compared and shown in FIGS. 8 (a)-(d). Figure 8 (a)-(d)
As shown in Fig. 5, it can be seen that the experimental results and the calculated results show good agreement regarding the internal crack. There are some errors in the amplitude and phase of the backside crack, but not so large that it hinders the inverse problem analysis. From the experimental results, it can be considered that the crack shape can be reconstructed.

(Inverse Problem Analysis Method) The calculation procedure of the inverse problem analysis used for the quantitative evaluation of the crack shape is shown below. 1) The crack shape is given to the analysis model, and the eddy current flaw detection signal is obtained by the previous high-speed forward problem analysis. 2) Compare the experimental signal with the analytic signal, and if they do not match, correct the shape by the steepest descent method and calculate the signal again. 3) Perform this procedure until the error between the experimental and analytical signals becomes smaller than a predetermined value or the change becomes smaller. Since this analysis uses high-speed forward problem analysis using a database, the inverse problem analysis itself is also characterized by being high-speed.

(Reconstruction of Crack Shape) The crack shape is estimated using the signal converted in the above. Crack reconstruction is achieved by iterative computations targeting the transformed input signal until a closely resembling signal is obtained. The width of the crack is fixed at 0.2 mm, and a rectangle of 18 mm × 0.2 mm × 7 mm is set as the Suspect Region where the crack exists. Approximately represent the shape of a crack with a rectangular array, and set 16 parameters for each 1 mm (2 mm in part). There are 22 measurement points in the direction passing on the center line of the detection coil as described above. Inverse problem analysis method developed by Huang et al. (H. Huang, T. Takagi,
H.Fukutomi and J. Tani, Forward and Inverse Analysis
of ECT Signals Based on Reduced Vector Potential
Method Using A Database, Electromagnetic Nondestru
ctive Evaluation (II), IOS press, (2001), pp.313-32
Using 1), the crack shape was estimated. This estimation result is
Shown in FIGS. 9 (a)-(d).

Regarding the inner surface cracks, although some crack lengths were slightly deviated, the crack depths showed good agreement with the actual cracks. The maximum error in the crack length direction is 18%, whereas the maximum error in the crack depth is 1.1%. On the other hand, with regard to the estimation of external cracks, the crack shape could be reconstructed accurately for the two detectable cracks. Maximum depth error is 4.1%, length error is 9%
Is. By reconstructing these crack shapes, the developed probe can be said to be suitable for inverse problem analysis for cracks on the back surface in flaw detection of thick materials.

[0030]

INDUSTRIAL APPLICABILITY The present invention can accurately detect defects such as cracks existing on the surface, inside and back surface of a thin plate as well as the surface of a thick plate by utilizing three-dimensional eddy current analysis. An eddy current flaw detection method and an eddy current probe therefor are provided. The feature of the eddy current flaw detection method is that strong eddy currents that penetrate flatly in the plate thickness direction are generated by passing currents in opposite directions to the two exciting coils. It is possible to provide an eddy current flaw detection probe which generates a large eddy current and in which the defect signal on the back surface is not affected by the noise on the front surface. For the position of the detection coil, in the case of this probe (back surface eddy current density) / (surface eddy current density)
Shows 0.5 or more, and further shows 0.7 or more, and shows a high value in a state where a strong eddy current density is maintained. As a result of conducting an experiment using this probe, the detection capability of the INCONEL test piece with a thickness of 7 mm was 0.25 mm for the internal crack and 0.5 mm for the external crack.
You can reach the level of being. Furthermore, it is possible to analyze the inverse problem by comparing the experimental results with the analysis results.
By performing this inverse problem analysis, there is an excellent advantage that the crack shape can be restored well even for the outer surface crack. From the above, the probe of the present invention has a high defect detection capability in flaw detection of a thick material, is excellent in defect dimension evaluation, and has a remarkable effect suitable for defect detection.

[Brief description of drawings]

FIG. 1 is a diagram showing an overview of an eddy current flaw detection probe of the present invention.

FIG. 2 is an explanatory diagram of dimensions and shapes of an exciting coil.

FIG. 3 is a diagram showing an analysis result of a change in backside eddy current when the inner diameter and the height are simultaneously changed.

FIG. 4 is a diagram showing a relationship between a distance between exciting coils, an eddy current density on a front surface and a back surface, and a ratio thereof.

FIG. 5 is a diagram showing an image of a Vy signal obtained by two-dimensionally scanning (C scan) a semi-elliptical inner surface (back surface) crack using the probe of the present invention.

6 is a graph of the Vy signal extracted from the B-scan (scan line 1) signal along the line shown in FIG.

7 is a graph of the Vy signal extracted from the B-scan (scan line 2) signal along the line shown in FIG.

FIG. 8 is a diagram showing a comparison between numerical analysis results and experimental results.

FIG. 9 is a diagram showing a result of defect shape reconstruction by inverse problem analysis.

[Explanation of symbols]

1 Excitation coil 2 detection coil 3 current 4 Surface of inspection object 5 Back side of the object 6 magnetic flux 7 height 8 width 9 Inner radius 10 Inspected object

   ─────────────────────────────────────────────────── ─── Continued front page    (72) Inventor Tetsuya Uchiichi             2-1-1 Katahira, Aoba-ku, Sendai City, Miyagi Prefecture             Inside the Institute for Fluid Science, University (72) Inventor Kazuhiko Sato             2-1-1 Katahira, Aoba-ku, Sendai City, Miyagi Prefecture             Inside the Institute for Fluid Science, University (72) Inventor Huang Kou             2-1-1 Katahira, Aoba-ku, Sendai City, Miyagi Prefecture             Inside the Institute for Fluid Science, University F-term (reference) 2G053 AA11 AB07 AB21 AB27 BA02                       BA15 BB11 BC02 BC07 BC11                       BC14 CA03 DA01

Claims (13)

[Claims] 1. A method in which currents in opposite directions are applied to two exciting coils, eddy currents caused by the respective coils are superposed in a region between the exciting coils, and flaw detection is performed by a detecting coil arranged between the exciting coils. Characteristic eddy current flaw detection method. 2. The eddy current flaw detection method according to claim 1, wherein the detection coil is a differential coil. 3. The ratio of the back surface eddy current density and the surface eddy current density immediately below the detection coil to the back surface (eddy current density / surface eddy current density).
The eddy current flaw detection method according to claim 1 or 2, wherein the distance between the exciting coils is set so that is 0.5 or more. 4. The ratio of the inner radius r of the exciting coil to the height h is
4. The eddy current flaw detection method according to claim 1, wherein r / h is in the range of 0.5 to 1.5. 5. The eddy current flaw detection method according to claim 1, wherein the surface of the plate and the surface of the plate are inspected for scratches on the inside and the back of the plate. 6. The eddy current flaw detection method according to claim 1, wherein a flaw inspection of a thick plate having a thickness of 6 mm or more is performed. 7. The eddy current flaw detection method according to claim 1, wherein the crack shape is reconstructed by an inverse problem analysis method. 8. An eddy current comprising two exciting coils for passing currents in opposite directions, and a detection coil arranged at a position where eddy currents caused by the exciting coils overlap in a region between the exciting coils. Probe for flaw detection. 9. The eddy current flaw detection probe according to claim 8, wherein the detection coil is a differential coil. 10. The exciting coil is installed at a separated position where the ratio of the back surface eddy current density and the surface eddy current density directly below the detection coil is 0.5 or more (eddy current density / surface eddy current density). The probe for eddy current flaw detection according to claim 8 or 9. 11. The vortex according to each of claims 8 to 10, wherein the ratio of the inner radius r of the exciting coil to the height h is in the range of r / h = 0.5 to 1.5. Current flaw detection probe. 12. The scratch inspection of the inside and the back of the plate is performed from the front surface of the plate and the front surface of the plate.
1. The probe for eddy current flaw detection according to each of 1. 13. The probe for eddy current flaw detection according to claim 8, wherein a flaw inspection of a thick plate having a plate thickness of 6 mm or more is performed.