JP2000227422A - Eddy current examination - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an eddy current examining method of high defect detecting performance affected hardly by fluctuation of lift-off and a noise magnetic flux. SOLUTION: A frequency f1 of an alternating current power source 6a for exciting an eddy current probe 5a is made low, and a frequency f2 of an alternating current power source 6b for exciting an eddy current probe 5b is made high. Both a noise generated in the vicinity of a surface of a thin steel plate 1 and a signal generated by an internal defect 8 are detected thereby from the probe 5a. On the other hand, the singal generated by the internal defect 8 is not detected or detected slightly from the probe 5b, and the noise generated in the vicinity of the surface of the thin steel plate 1 is detected mainly from the prove 5b. A noise component is removed to enhance an S/N ratio by operating the these signals thereof.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of applying an alternating magnetic field to a metal object, and changing a magnetic field caused by an eddy current generated by the defect existing inside the metal object near a surface of the metal object. The present invention relates to an eddy current flaw detection method for detecting a flaw by detecting a defect with a magnetic sensor disposed in the flaw detection sensor.
Eddy current flaw detection method with improved ratio.

[0002]

2. Description of the Related Art Eddy current flaw detection is widely used as a method for detecting a defect existing inside a ferromagnetic material such as iron. As an example, FIG. 7 shows an example of the configuration of an eddy current flaw detector installed in an inspection line in an iron making plant.

An eddy current flaw detector 14 is provided along a transport path of a metal object 11 such as a thin steel strip which is transported at a substantially constant speed V on a product inspection line by transport rollers 12 and 13. The eddy current flaw detector 14 is provided with an eddy current probe (coil) 15 disposed near the surface of the metal object 11 in a running state, an AC current source 16 for supplying an AC current to the eddy current probe 15, and detection signals from the eddy current probe 15. And a signal processing device 17 for detecting a defect 18 inside or on the surface of the metal object 11 based on the signal processing device 17.

When a defect 18 exists in the metal object 11,
The eddy current of the metal object 11 and the magnetic field generated thereby are disturbed by the defect 18. The eddy current probe 15 detects this change in the magnetic field. Since the intensity of the signal generated by the magnetic field change corresponds to the size of the defect 18, the size of the defect 18 can be evaluated based on the signal level of the detection signal of the eddy current probe 15.

[0005]

As described above, in the conventional defect detection of a metal object, the defect is detected by the signal level of the detection signal of the eddy current probe. However, the magnetic signal detected by the eddy current probe varies not only due to the eddy current disturbance due to the above-described defect but also to various factors. For example, due to variations in the distance (lift-off) between the sensor and the defect,
The intensity and distribution of the eddy current generated in the metal object change, and at the same time, when detecting the magnetic field generated by the eddy current, the relative position between the sensor and the metal object is different, so the detection is performed by the sensor. The resulting magnetic signal changes. The signal generated by such a lift-off fluctuation becomes noise, and deteriorates the detectability.

In addition, the signal of the eddy current probe includes disturbances in the magnetic flux distribution outside the metal specimen due to local changes in magnetic and electrical characteristics and unevenness in the metal specimen, and disturbances in the magnetic field caused by surface roughness. May be included. From the point of view of defect detection, these signals
This is an unnecessary magnetic flux (noise magnetic flux), which deteriorates the detectability.

As a countermeasure against the former lift-off fluctuation, for example, Japanese Patent Application Laid-Open No. 61-2065 describes a method using phase discrimination in order to eliminate the influence of a signal change accompanying the lift-off fluctuation. . However,
The latter "noise flux" cannot be reduced by such a method.

[0008] Regarding measures against noise magnetic flux, in order to avoid the influence of noise magnetic flux, after phase detection is performed, the defect caused by the difference between the frequency of the signal caused by the defect and the frequency of the signal caused by the noise magnetic flux is detected. Judgment methods may be used.

FIG. 8 is a diagram showing an example of the measurement results of the frequency characteristics of the defect signal and the signal after phase detection of the noise magnetic flux. That is, FIG. 8 shows the frequency characteristics of a defect signal when leakage magnetic flux due to a defect is detected by a magnetic sensor and the frequency characteristics when noise magnetic flux is detected by a magnetic sensor when a thin steel plate is run at a constant speed. Is exemplified.

As can be seen from FIG. 8, the defect signal generally has a higher frequency distribution than the noise flux. Therefore, it is possible to incorporate a high-pass filter having a cutoff frequency f into the signal processing device and relatively emphasize and extract the defect signal from the detection signals output from the magnetic sensor to the signal processing device as compared with the noise magnetic flux. It is possible.

However, since the frequency characteristic of the defect signal and the frequency characteristic of the noise magnetic flux overlap each other, if the defect to be detected is small and the level of the defect signal is small, or if the noise magnetic flux is large, even if the defect magnetic flux is large, Even if the frequency is discriminated, it is difficult to reduce the signal level of the noise magnetic flux to a level at which a defect can be detected.

[0012] The present invention has been made in view of the above situation, and an AC magnetic field is applied to a metal object, and the magnetic field caused by the eddy current generated by the AC magnetic field is present inside the metal object. An eddy current flaw detection method for detecting flaws by detecting a fluctuation due to a defect to be detected by a magnetic sensor arranged near the surface of the metal object, which is less susceptible to fluctuations in lift-off and noise magnetic flux. It is an object to provide a high-performance eddy current flaw detection method.

[0013]

According to a first aspect of the present invention, an AC magnetic field is applied to a metal object, and a magnetic field caused by an eddy current generated by the AC magnetic field is generated inside the metal object. In the eddy current flaw detection method for detecting flaws by detecting the variation due to the defects present in the magnetic sensor disposed near the surface of the metal object, the measurement is performed at different excitation frequencies, under different conditions The present invention relates to an eddy current flaw detection method (claim 1), in which the measurement signals are calculated and the defect signal is detected based on the result.

In order to describe the reason why the defect detection performance can be improved by the present means, the nature of the noise magnetic flux will be described first. The following is an example of the results of a survey using a steel plate having a thickness of 1 mm. From the surface of the steel sheet, it was chemically cut little by little so as not to introduce distortion, and the change in the noise magnetic flux level was examined. As shown in FIG. 2, the noise magnetic flux gradually decreased, and the surface layer was cut by about 20 μm. By the way, it turned out that it is less than half of the state before cutting. That is, it turned out that the main source of the noise flux was in the surface layer. This is because
For example, the magnetic signal fluctuates due to the surface roughness, and unevenness in the magnetic properties is concentrated on the surface due to local variations in the surface layer structure caused by cooling from the surface during steel plate manufacturing. It is thought to be due to this. Such a phenomenon is observed in samples other than the sample used here, and can be considered as one of the properties of the noise magnetic flux.

Now, assuming that the main part of the noise magnetic flux has its source in the surface layer, the internal defect is generally deeper than that. The following describes what happens when the behavior of the noise magnetic flux and the defect signal level is considered under two kinds of measurement conditions having different excitation frequencies.

When an AC magnetic flux is applied to a conductor such as steel,
The magnetic flux penetrating into the conductor due to the skin effect becomes weaker as it becomes deeper. Therefore, the deeper the signal source, the lower the detection signal level. This phenomenon appears more remarkably as the excitation frequency is higher. For example, as the low excitation frequency, select a frequency at which a defect is sufficiently detected, and, compared to a low excitation frequency as a high excitation frequency, the noise flux having its source on the surface layer is smaller than the signal from the defect at a deep position. A frequency that can be emphasized relatively can be selected.

An example is shown in FIG. FIG. 3 shows the excitation frequency (detection frequency) on the horizontal axis, and the ratio of the defect signal level to the noise level (S / N ratio) on the vertical axis. As the excitation frequency increases, the defect signal becomes a noise magnetic flux signal. It can be seen that it becomes smaller than. Since the difference between the magnetization level of the defect signal and the change in the magnetization level of the noise magnetic flux is generated according to the above-described principle as shown in FIG. 3, two kinds of appropriate excitation frequency conditions are selected, and the sample is obtained from the two kinds of conditions. By performing an appropriate calculation between signals corresponding to the same position, it is possible to reduce noise magnetic flux which is largely present in two types of measurement conditions and relatively strengthen a defect signal. As the calculation, an appropriate one may be selected so as to improve the detectability in accordance with the properties of the defect signal and the noise.

In addition, the change in the magnetic sensor signal due to the lift-off fluctuation is also eliminated when calculating signals having different measurement conditions, and as a result, the influence of the lift-off fluctuation can be reduced.

Here, the case where two kinds of excitation frequency conditions are used has been described. However, when measurement is performed under three or more kinds of excitation frequency conditions, and when the data corresponding to the same position of the measurement target is calculated as a result, Needless to say, the defect detection ability can be improved by the same concept and method.

A second means for solving the above-mentioned problem is as follows.
The first means, wherein a measurement value at a higher excitation frequency is weighted and subtracted from a measurement value at a lower excitation frequency, and a defect signal is detected based on the result. (Claim 2).

The magnitude of the noise magnetic flux signal does not change so much when a high-frequency excitation frequency is used or when a low-frequency excitation frequency is used. On the other hand, as described above, the magnitude of the magnetic flux signal caused by the internal defect greatly changes depending on the excitation frequency. In this means, by taking advantage of this property, by weighting and subtracting the measurement value obtained when performing flaw detection at a high frequency from the measurement value obtained when flaw detection is performed at a low frequency, it is common to both. The included noise magnetic flux signal is eliminated. As a result, almost only the magnetic flux signal caused by the internal defect remains and appears as a detection signal, so that the internal defect can be detected with a large S / N ratio.

[0022]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing an outline of an apparatus for performing an eddy current flaw detection method according to the present embodiment, and shows an example of application to flaw detection of internal inclusions in a thin steel plate. In FIG. 1, 1 is a thin steel plate, 2 and 3 are transport rolls, 4 is an eddy current flaw detector, 5a and 5b are eddy current probes serving both as an AC magnetizer and a magnetic sensor, 6a and 6b are AC power supplies, and 7 is a signal processing device. , 8 are internal defects.

In this product inspection line, an eddy current flaw detector 4 is installed along the conveying path of the steel sheet 1. The eddy current flaw detector 4 is mainly composed of eddy current probes 5a and 5b and a signal processing device 7. Eddy current probe 5a, 5
b is excited by AC power supplies 6a (frequency f1) and 6b (frequency f2), respectively. This eddy current probe 5a,
The lift-off value, which is the distance between 5b and the steel sheet 1, is L, which is the same for both.

Although not shown, if the object to be measured is a ferromagnetic material, a DC magnetizer is provided corresponding to each of the eddy current probes 5a and 5b, and the thin steel sheet 1 is magnetized by applying a DC bias magnetic field. It is desirable to magnetize to the saturation level to reduce the differential magnetic permeability, increase the penetration depth of the eddy current, and reduce the influence of the change in the material of the thin steel plate 1.

For example, the frequency f1 of the AC power supply 6a for exciting the eddy current probe 5a is low, and the frequency f2 of the AC power supply 6b for exciting the eddy current probe 5b is high. Thereby, the signal generated by the internal defect 8 is detected from the eddy current probe 5a together with the noise generated near the surface of the thin steel plate 1. On the other hand, from the eddy current probe 5b, a signal generated by the internal defect 8 is not detected or is small even if it is detected, and noise generated mainly near the surface of the thin steel plate 1 is detected.

In order to determine the amplitude of the magnetic signal (AC) detected by the eddy current probes 5a and 5b, the AC power supplies 6a and 6b
Is subjected to phase detection processing by the signal processing device 7 using the reference signal from the output of (1), but since this is a conventional means in the eddy current flaw detection method, the description thereof is omitted. further,
For the signal after the phase detection, the signal from the same position on the steel plate is operated by the signal signal processing device 7 to reduce noise magnetic flux and relatively emphasize a defect signal to improve S / N.

For example, if the signal of the eddy current probe 5a after the phase detection is Va (t) and the signal of the eddy current probe 5b is Vb (t), V (t) = Va (t) −k · Vb (t). , And calculates the detection signal. Where k is a constant,
It is experimentally required to eliminate the noise magnetic flux included in Va (t) and Vb (t). Actually, since the distance between the eddy current probe 5a and the eddy current probe 5b is large, in the above equation, the signal Va (t) corresponds to the time during which the thin steel plate 1 travels from the position of the eddy current probe 5a to the position of the eddy current probe 5b. ) Will be used. The lift-off L does not necessarily have to be the same for the eddy current probes 5a and 5b, and may be different from each other.

Processing such as subtraction of a measured value at high-frequency excitation and a measured value at low-frequency excitation, delay processing, and filtering may be performed on an analog signal, or may be performed after converting an analog signal into a digital signal. You may. Further, even when the conversion is performed after converting into a digital signal, the conversion may be performed by hardware or software.

FIG. 4 shows an outline of another apparatus for performing the eddy current flaw detection according to the embodiment of the present invention. In the following drawings, the same components as those already appearing in the drawings after the drawings used for describing the embodiments of the present invention are denoted by the same reference numerals, and description thereof will be omitted. In FIG. 4, 6 is an eddy current probe, and 6 is an AC power supply.

The device shown in FIG. 4 is different from the device shown in FIG. 1 only in that there is only one eddy current probe 5 and one AC power source 6, and from the AC power source 6, an AC having a frequency f1 and an AC having a frequency f2 are provided. The point that the mixture is applied to the coil of the eddy current probe 5. The output signal from the eddy current probe 5 is subjected to phase detection separately in the signal processing device 7 using the component of the frequency f1 and the component of the frequency f2 from the AC power supply 6, thereby obtaining the excitation frequency f1.
And a signal component caused by the excitation frequency f2. The other operations are the same as those shown in FIG. 1, and the description thereof will be omitted. According to the device shown in FIG. 4, not only one set of the eddy current probe 5 and the power supply device 6 may be used, but also the defect detection positions at the two excitation frequencies are the same, so that one signal is delayed for comparison. There is no need, and the circuit configuration is simplified.

When the object to be measured is a ferromagnetic material, it is needless to say that direct current magnetization (for example, magnetizing to a saturation level and decreasing magnetic permeability) can be applied to the alternating magnetic field. .

[0032]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment in which the present invention is applied to an apparatus for detecting minute internal inclusions in a thin steel sheet (thin steel strip) online will be described below with reference to FIG. The eddy current flaw detector shown in FIG. 5 is the same as that shown in FIG. 1, but in FIG. 1, a delay processing circuit 9 which is placed inside the signal processor 7 and is not shown is an independent element. The only difference is that they are configured and illustrated.

The thickness of the thin steel sheet 1 transported on the product inspection line was 1 mm. Further, the steel plate 1 has a conveying roller 2,
According to FIG. 3, the sheet was conveyed at a substantially constant speed V = 20 m / min.

The lift-off L1, which is the distance between each of the eddy current probes 5a and 5b and the surface of the thin steel plate 1, was set to 0.7 mm. Although not shown, a plurality of eddy current probes 5
a and 5b are linearly arranged at a pitch of 10 mm in the width direction of the plate, and a total of 200 sets of eddy current probes 5a and 5b cover 1 m in the width direction of the plate. In addition,
A DC magnetizer (not shown) is provided on the opposite side of the eddy current probes 5a and 5b with the thin steel plate 1 interposed therebetween.
Is DC magnetized to near the saturation magnetization level.

As the low-frequency excitation frequency, it is necessary to select a frequency at which both the defect signal and the noise magnetic flux are greatly detected. However, in order to make the noise detection situation similar, it is necessary to set the condition so that the noise flux signal under this condition does not become unnecessarily low so that the noise flux signal also exists in the detection signal at the high frequency excitation frequency. In consideration of these conditions, here, 5 kHz was selected, and the frequency of the power supply 6a was set to this value. It is sufficient to consider the defect signal frequency at about 700 Hz at the upper limit, and an excitation frequency of 5 kHz can sufficiently detect the frequency. As high-frequency excitation conditions, it is necessary to select conditions under which detection sensitivity is poor for a defect signal and noise magnetic flux can be detected. Here, 100 kHz is selected.

The signals from the eddy current probes 5a and 5b are phase-detected in the signal processing device 7 using a reference signal from an excitation power supply. The detection signals Va (t) and Vb (t) after the phase detection are applied to a band-pass filter in order to reduce a direct current component, a formation noise component having a low frequency, and to cut an electric noise higher than a defect signal frequency. The passband is the same under both magnetization conditions, 200-500 Hz.

The signal processing device 7 converts the phase-detected detection signals Va (t) and Vb (t) of the eddy current probes 5a and 5b to 10 kHz.
Analog-to-digital conversion is performed at the sampling frequency of z, and subsequent calculations are performed using digital values.

Further, the time difference Δt corresponding to the same steel sheet position is obtained by dividing the positional deviation d between the eddy current probe 5a and the eddy current probe 5b in the moving direction of the thin steel sheet 1 by the sequentially measured steel sheet speed V, and delay processing is performed. The circuit 10 relatively compares the signal Va (t) of the eddy current probe 5a with the signal Vb of the magnetic sensor 11b.
Va (t−Δt) and Vb (t) are made to correspond to each other by delaying (t). That is, the signal used for defect determination
V (t) = Va (t−Δt) −k · Vb (t) is used as V (t). Here, k is a constant, and Va (t−Δt) and V
The noise flux contained in b (t) was experimentally determined to be eliminated.

FIG. 6 shows the effect of improving the detection ability. Under low frequency excitation conditions (measured with the eddy current probe 5a), noise due to the material is large and the S / N ratio is 1.5. Under the high frequency excitation condition (measured by the eddy current probe 5b), the noise magnetic flux generated under the low frequency excitation condition also appears, but it can be seen that the signal corresponding to the defect signal is not detected. The result obtained by multiplying the signal under the high frequency excitation condition by 0.2 and subtracting it from the signal under the low frequency excitation condition at the corresponding position is the difference processing result. The noise magnetic flux is drastically reduced, and the defect signal is relatively emphasized.
It can be seen that the S / N ratio has improved (3.3).

[0040]

As described above, according to the first aspect of the present invention, measurement is performed at two different excitation frequencies, and measurement signals under the two different conditions are calculated. Since the defect signal is detected based on the result, it is possible to reduce the noise magnetic flux which is largely present under each measurement condition and relatively strengthen the defect signal.

In the invention according to claim 2, the measured value at the higher excitation frequency is weighted and subtracted from the measured value at the lower excitation frequency, and the defect signal is detected based on the result. Is performed to eliminate the noise magnetic flux signal included in both, so that the internal defect signal can be detected with a high S / N ratio by a simple calculation.

[Brief description of the drawings]

FIG. 1 is a diagram showing an outline of an apparatus for performing an eddy current flaw detection method according to the present embodiment.

FIG. 2 is a diagram illustrating an example of a relationship between a deleted thickness of a surface of a steel sheet and a normalized noise magnetic flux signal level.

FIG. 3 is a diagram showing a relationship between an excitation frequency and an S / N ratio of an internal defect.

FIG. 4 is a diagram showing an outline of another apparatus for performing the eddy current flaw detection method according to the present embodiment.

FIG. 5 is a schematic diagram showing a configuration of an eddy current flaw detection device used in an embodiment of the present invention.

FIG. 6 is a diagram showing an improvement result of an S / N ratio in an example of the present invention.

FIG. 7 is a schematic view showing an example of a conventional eddy current flaw detection device.

FIG. 8 is a diagram illustrating an example of measurement results of frequency characteristics of a defect signal and a noise magnetic flux.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Thin steel plate 2, 3 ... Conveyance roll 4 ... Eddy current flaw detector 5, 5a, 5b ... Eddy current probe 6, 6a, 6b ... AC power supply 7 ... Signal processing device 8 ... Internal defect 9 ... Delay processing circuit

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Akiyo Nagato 1-2-1, Marunouchi, Chiyoda-ku, Tokyo F-term (reference) 2G053 AA11 AB21 BA15 BB03 BC02 BC07 CA03 CB10 CB24 DA01 DA06 DB02

Claims (2)

[Claims] 1. An AC magnetic field is applied to a metal object, and a variation in a magnetic field caused by an eddy current generated by the defect inside the metal object is arranged near a surface of the metal object. An eddy current flaw detection method for flaw detection by detecting a defect using a detected magnetic sensor.The measurement is performed at different excitation frequencies, the measurement signals under different conditions are calculated, and a defect signal is detected based on the result. Eddy current flaw detection. 2. The method according to claim 1, wherein the measured value at the higher excitation frequency is weighted and subtracted from the measured value at the lower excitation frequency, and a defect signal is detected based on the result. The eddy current flaw detection method according to claim 1.