JP6327185B2 - Sensitivity correction method and sensitivity correction apparatus for eddy current flaw detector - Google Patents

Description

  The present invention relates to a sensitivity correction method and a sensitivity correction apparatus for an eddy current flaw detection apparatus.

  Conventionally, an eddy current flaw detection method is widely known as a method for detecting a defect present in the surface layer of a metal band that is a conductor. In the eddy current flaw detection method, an eddy current is generated on the surface layer of a metal band by exciting a primary coil, a magnetic field generated by the eddy current induces an induced voltage in the secondary coil, and an output voltage due to the induced voltage is measured. To detect metal band defects.

  When there is a defect on the surface of the metal strip, the flow of eddy current changes due to the defect, so that the generation of a magnetic field due to the eddy current changes, and the induced voltage induced in the secondary coil is different from usual. For this reason, when the secondary coil is differentially connected, the output voltage between the secondary coils is not zero due to the defect, and therefore it is possible to detect a defect in the metal band.

  For example, Patent Document 1 discloses an E-type sensor in which a primary coil and two secondary coils are wound around an E-shaped core as a non-contact eddy current sensor used in an eddy current flaw detector. In the E-type sensor, a primary coil is wound around a central leg, and two secondary coils wound around both side legs are differentially connected. The E-type sensor is arranged at a position that is not in contact with the metal band so that the three legs protrude toward the surface of the metal band to be inspected.

  Patent Document 2 describes a sensitivity calibration method for an eddy current flaw detector in which the production line is periodically stopped and the sensitivity of the E-type sensor in the sensor array is calibrated. As described above, since the eddy current flaw detector is required to detect a defect in the metal band with high accuracy and stability over a long period of time, it is necessary to periodically calibrate the sensitivity of the eddy current sensor.

JP-A-7-116732 JP 2006-234750 A

  However, if the eddy current flaw detector is continuously used, a sensitivity change or a change with time in a eddy current sensor or an electric circuit is caused. Therefore, even if sensitivity calibration is performed periodically after the production line is stopped, the sensitivity of the eddy current sensor is not always constant until the next sensitivity calibration is performed.

  Further, when the production line must be stopped, it is difficult to frequently calibrate the sensitivity in a large-scale line such as a steel production line. Sensitivity calibration is performed while the line is stopped for a short period of 3 to 4 months, and the sensitivity of the eddy current sensor may change greatly during that period.

  The present invention has been made in view of the above circumstances, and provides a sensitivity correction method and sensitivity correction apparatus for an eddy current flaw detector capable of correcting sensitivity during operation of a production line with respect to sensitivity changes caused by an eddy current sensor. For the purpose.

  The sensitivity correction method for an eddy current flaw detection apparatus according to the present invention is the sensitivity correction method for an eddy current flaw detection apparatus having a flaw detection sensor comprising a non-contact type eddy current sensor that is a test object for a metal strip running on a non-metal roll. During the operation of the metal band production line, a correction sensor comprising a non-contact eddy current sensor is provided outside the edge of the metal band in the width direction of the metal band on the roll surface of the non-metal roll. A detection step of detecting the detected metal piece, a calculation step of calculating a correction coefficient based on a detection value of the metal piece by the correction sensor, and the metal band output from the flaw detection sensor during the operation And a correction step of correcting the sensitivity of the detected value by the correction coefficient.

  The sensitivity correction method of the eddy current flaw detection apparatus is an information that can identify that the correction coefficient exceeds the predetermined threshold without performing the sensitivity correction in the correction step when the correction coefficient exceeds a predetermined threshold. It is preferable to further include a notifying step for notifying at.

  In the sensitivity correction method of the eddy current flaw detection apparatus, the detection step includes detecting the metal piece by the correction sensor when the production line resumes operation, and obtaining a detection value of the metal piece as a reference value; Detecting the metal piece a plurality of times in a predetermined cycle after obtaining the reference value, and obtaining a detection value of the metal piece as a current detection value, and the calculating step includes the step of calculating the current detection value with respect to the current detection value. It is preferable to include a step of calculating the correction coefficient that is a ratio of the reference values.

  In the sensitivity correction method of the eddy current flaw detector, the detection step includes a step of detecting the metal piece by a plurality of correction sensors, and the calculation step includes a detection value of the metal piece by the plurality of correction sensors. Preferably, the method includes a step of calculating the correction coefficient using an average value based on.

  The sensitivity correction apparatus for an eddy current flaw detection apparatus according to the present invention is the sensitivity correction apparatus for an eddy current flaw detection apparatus having a flaw detection sensor comprising a non-contact type eddy current sensor for inspecting a metal strip running on a non-metal roll. A metal piece provided on a non-contact surface located outside the edge portion of the metal band in the width direction of the metal band among the roll surfaces of the non-metal roll, so as to face the non-contact surface and the metal piece A correction sensor composed of a non-contact type eddy current sensor disposed on the metal strip, a detection means for detecting the metal piece by the correction sensor while the metal strip production line is in operation, and the correction sensor A calculating means for calculating a correction coefficient based on the detection value of the metal piece; and a correction means for correcting the sensitivity of the detection value of the metal band output from the flaw detection sensor during the operation by the correction coefficient. Characterized in that it comprises and.

  The sensitivity correction apparatus of the eddy current flaw detection apparatus further includes notification means for notifying the fact that the correction coefficient exceeds a predetermined threshold with identifiable information when the correction coefficient exceeds a predetermined threshold. The correction means does not perform sensitivity correction when the correction coefficient exceeds the predetermined threshold value, and corrects the sensitivity of the flaw detection sensor with the correction coefficient when the correction coefficient is within the predetermined threshold value. Is preferred.

  The sensitivity correction apparatus for the eddy current flaw detection apparatus is characterized in that the detection means detects the metal piece by the correction sensor when the production line resumes operation, and acquires the detection value of the metal piece as a reference value. Detecting the metal piece a plurality of times in a predetermined cycle, and acquiring a detection value of the metal piece as a current detection value, wherein the calculation means is a ratio of the reference value to the current detection value. It is preferable to calculate a correction coefficient.

  In the sensitivity correction apparatus for the eddy current flaw detector, the metal piece is a metal tape attached on the non-contact surface, and the metal tape does not contact the metal band and the axis of the non-metal roll. It is preferable that it is formed in a predetermined length along the direction.

  The sensitivity correction device for the eddy current flaw detector includes the first metal piece provided on one non-contact surface located outside the one edge part of the metal band in the width direction of the metal band. And a second metal piece provided on the other non-contact surface located outside the other edge portion of the metal band in the width direction of the metal band, and the correction sensor includes the first metal piece. A first correction sensor for detecting the second metal piece, and a second correction sensor for detecting the second metal piece, wherein the calculation means includes the detection value of the first metal piece from the first correction sensor and the second metal piece. Preferably, the correction coefficient is calculated using an average value based on a detection value of the second metal piece from the second correction sensor.

  According to the present invention, during operation of the production line, the metal piece provided on the non-metal roll is detected by the correction sensor, and the sensitivity correction is performed for the flaw detection sensor that is inspecting the metal band based on the detected value. Can do. Thereby, since the sensitivity of the flaw detection sensor can be corrected in a short cycle without stopping the production line, the inspection accuracy of the metal band by the flaw detection sensor can be improved.

FIG. 1 is a flowchart showing a processing flow of a sensitivity correction method for an eddy current flaw detector according to this embodiment. FIG. 2 is a schematic diagram showing a sensitivity correction device of the eddy current flaw detector according to the present embodiment. FIG. 3 is a diagram schematically illustrating a correction sensor of the sensitivity correction apparatus. FIG. 4 is a diagram schematically showing a flaw detection sensor of the eddy current flaw detection apparatus. FIG. 5 is a block diagram showing an electronic control unit of the sensitivity correction apparatus. FIG. 6 is a graph showing the sensor sensitivity ratio in the periodic off-line calibration. FIG. 7 is a graph showing the sensor sensitivity ratio after correction. FIG. 8A is a schematic diagram illustrating an example of an off-line calibration method. FIG. 8B shows a jig for off-line calibration. FIG. 8C and FIG. 8D are diagrams showing the relative positional relationship between the sensor leg and the jig when the jig is detected by the flaw detection sensor.

  Hereinafter, a sensitivity correction apparatus for an eddy current flaw detector according to an embodiment of the present invention will be described in detail with reference to the drawings.

(1. Sensitivity correction device)
FIG. 2 is a schematic diagram showing a sensitivity correction device of the eddy current flaw detector according to the present embodiment. The sensitivity correction device (hereinafter simply referred to as “sensitivity correction device”) 1 of the eddy current flaw detection device is a copper tape 3 attached to a non-metal roll 2 and a non-contact type eddy current sensor that detects the copper tape 3. Sensor 4 and an electronic control unit 20 (not shown in FIG. 2) to which a detection signal from the correction sensor 4 is input.

  The sensitivity correction apparatus 1 corrects the sensitivity for the flaw detection sensor 11 of the eddy current flaw detection apparatus 10 that inspects the steel strip 5 traveling on the production line. The flaw detection sensor 11 inspects a defect of the steel strip 5 traveling on the non-metallic roll 2. In the eddy current flaw detector 10, a plurality of (for example, 50 channels) flaw detection sensors 11 are arranged in the width direction of the steel strip 5 to form a sensor array, and therefore the steel strip traveling on the non-metallic roll 2 is used. 5, the entire width direction can be included in the inspection region.

  In the production line, the steel strip 5 formed in various widths can run on the non-metallic roll 2, and therefore the axial length of the non-metallic roll 2 is formed larger than the maximum width of the steel strip 5. Yes. Therefore, in the non-metallic roll 2, both end sides in the axial direction of the roll surface formed on the cylindrical outer peripheral surface are non-contact surfaces 2a that do not contact the traveling steel strip 5.

  The copper tape 3 is affixed to the non-contact surface 2a. As shown in FIG. 2, since there are non-contact surfaces 2a on both outer sides in the width direction of the steel strip 5 with respect to the running steel strip 5, the first copper tape 3A and the second copper tape 3B are provided on both outer sides. And are provided.

  The first copper tape 3A is affixed to one non-contact surface 2a located on the outer side in the width direction with respect to one width direction end portion (hereinafter referred to as “one edge portion”) 5a of the steel strip 5. The 2nd copper tape 3B is affixed on the other non-contact surface 2a located in the width direction outer side rather than the other width direction edge part (henceforth "the other edge part") 5b of the steel strip 5. FIG.

  Each copper tape 3 </ b> A, 3 </ b> B is formed in a rectangular shape having a predetermined thickness that can be detected by the correction sensor 4, and is non-contact so that its longitudinal direction is parallel to the axial direction of the non-metallic roll 2. It is affixed on the surface 2a. As shown in FIG. 2, the copper tapes 3 </ b> A and 3 </ b> B extend linearly with a predetermined length from both ends in the axial direction of the roll surface to the center side, and the center side end portions are both edge portions 5 a and No contact with 5b. That is, in the width direction of the steel strip 5, the first copper tape 3A is located outside the one edge portion 5a, and the second copper tape 3B is located outside the other edge portion 5b. In addition, the 1st copper tape 3A and the 2nd copper tape 3B may be provided in the same position in the rotation direction (circumferential direction of a roll surface) of the nonmetallic roll 2, or are provided in different positions. May be.

  The correction sensor 4 includes a first correction sensor 4A that detects the first copper tape 3A, and a second correction sensor 4B that detects the second copper tape 3B. Since the first correction sensor 4A is provided at a position facing the first copper tape 3A and the one non-contact surface 2a, the first correction sensor 4A is wider than the one edge portion 5a in the relative positional relationship with the steel strip 5. Located outside in the direction. Since the second correction sensor 4B is provided at a position facing the second copper tape 3B and the other non-contact surface 2a, the relative position relative to the steel strip 5 is wider than the other edge portion 5b. Located outside in the direction.

  Further, since the correction sensor 4 is a non-contact sensor, the first correction sensor 4A is separated from the first copper tape 3A and one non-contact surface 2a by a predetermined distance in the radial direction of the non-metal roll 2. . Similarly, the second correction sensor 4B is separated from the second copper tape 3B and the other non-contact surface 2a by a predetermined distance.

  Further, the correction sensor 4 is not a sensor for inspecting the steel strip 5. Therefore, each correction sensor 4A, 4B is provided at a position where the steel strip 5 cannot be detected so as not to detect the steel strip 5 by mistake. In other words, each copper tape 3A, 3B is arranged adjacent to each correction sensor 4A, 4B in the sensor row in the width direction, that is, each flaw detection sensor arranged on the outermost side in the width direction in the sensor row. 11 and 11 are pasted at positions that do not affect the detection of the steel strip 5.

  The flaw detection sensor 11 is a non-contact eddy current sensor, and is provided at a position facing the surface of the steel strip 5 and a predetermined distance away from the surface. As the sensor array, a plurality of flaw detection sensors 11 are arranged at predetermined intervals in the width direction of the steel strip 5. Further, the flaw detection sensor 11 in the sensor array and the correction sensors 4A and 4B are arranged in a line in the width direction of the steel strip 5, and the correction sensors 4A and 4A are arranged on both sides in the width direction of the flaw detection sensor 11 in the sensor array. 4B is provided.

  In this description, the first copper tape 3 </ b> A and the second copper tape 3 </ b> B are described as the copper tape 3 unless particularly distinguished. Similarly, the first correction sensor 4 </ b> A and the second correction sensor 4 </ b> B are referred to as correction sensors 4 when they are not particularly distinguished.

  FIG. 3 is a schematic diagram for explaining the sensor structure of the correction sensor 4. The correction sensor 4 is an E-type sensor including an E-shaped core 7 having three legs 7a, 7b, and 7c, a primary coil (excitation coil) 8, and a secondary coil (detection coil) 9. . The correction sensor 4 is supported by a support member (not shown).

  The primary coil 8 is wound around the central leg portion 7a of the core 7, and both ends thereof are electrically connected to an AC power source (not shown). The secondary coil 9 is composed of two secondary coils 9A and 9B that are differentially connected. One secondary coil 9 </ b> A is wound around one side leg 7 b of the core 7. The other secondary coil 9 </ b> B is wound around the other side leg 7 c of the core 7. Furthermore, both ends of the differentially connected secondary coils 9A and 9B are electrically connected to a phase detector (not shown) for phase detection of the output between one secondary coil 9A and the other secondary coil 9B. It is connected to the.

  Also, since the three legs 7a, 7b, 7c are all formed in parallel and the same length, the gap width between the center leg 7a and one of the both side legs 7b is equal to the center leg 7a and the other side. The width is the same as the gap width of the leg 7c. In this case, the width of each copper tape 3A, 3B (the length in the circumferential direction of the roll surface) is formed to be smaller than the gap between the central leg 7a and the both side legs 7b, 7c. Then, the correction sensor 4 is arranged so that all the leg portions 7a, 7b, 7c protrude toward the non-contact surface 2a and are aligned along the tangential direction of the non-contact surface 2a. As shown in FIG. 3, one side leg 7 b, the center leg 7 a, and the other side leg 7 c are arranged in this order in the rotational direction of the non-metallic roll 2.

  Here, the detection principle of the copper tape 3 by the correction sensor 4 will be described. The correction sensor 4 is an eddy current type displacement sensor that detects the position of the copper tape 3. During operation of the production line, the non-metallic roll 2 rotates and the circumferential position (rotational direction position) of the copper tape 3 changes, but the correction sensor 4 is fixed, so the copper tape is fixed to the correction sensor 4. 3 is displaced. In short, the copper tape 3 periodically passes through a region (opposing region) facing the correction sensor 4. It can be said that the facing region is a region where the correction sensor 4 is projected onto the non-contact surface 2a.

  Specifically, the copper tape 3 first passes through a region (first region) facing the one side leg 7b around which one secondary coil 9A is wound. Thereafter, an opposing region (third region) facing the central leg portion 7a around which the primary coil 8 is wound is interposed through an opposing region (second region) between the one side leg 7b and the gap between the central leg 7a. pass. Then, through a region (fourth region) between the central leg 7a and the gap between the other side legs 7c, a region opposite to the other side legs 7c around which the other secondary coil 9B is wound ( 5th area).

Thus, when the copper tape 3 passes through the region (first to fifth regions) facing the correction sensor 4, the primary coil 8 is excited to excite the magnetic field generated by the primary coil 8. Therefore, an eddy current is generated in the copper tape 3. An induced voltage is induced in each of the secondary coils 9A and 9B by the magnetic field generated by the eddy current of the copper tape 3. Then, the correction sensor 4 detects the copper tape 3 and outputs a detection signal to the electronic control unit 20. The detection signal (the detection signal of the copper tape 3) output from the correction sensor 4 includes the output voltage V _out of the secondary coils 9A and 9B that represents the detection value of the copper tape 3.

For example, when the copper tape 3 is located in a region (second region) opposite to the gap between the one side leg 7b and the center leg 7a, the first induced voltage V induced in the one secondary coil 9A. _{i_1} is larger than the second induced voltage V _{i_2} induced in the other secondary coil 9B. That is, differentially connected secondary coils 9A, in 9B, since the value greater than the difference between the first induced voltage _{V i_1} a second induced voltage _{V i_2} is 0, the output voltage _{V out} is not zero.

Moreover, when the copper tape 3 is located in the area | region (3rd area _| region) facing the center leg part 7a, the 1st induction voltage _{Vi_1} in one secondary coil 9A is the 2nd induction in the other secondary coil 9B. The voltage V _{i_2} has the same magnitude. Since the two secondary coils 9A and 9B are differentially connected, the first induced voltage V _{i_1} and the second induced voltage V _{i_2} cancel each other, and the output voltage V _out becomes zero.

And when the copper tape 3 is located in the opposing area | region (4th area _| region) of the center leg part 7a and the gap | interval part of the other both-sides leg part 7c, 2nd induced voltage V _{i_2} in the other secondary coil 9B is It is larger than the first induction voltage V _{i_1 in} one secondary coil 9A. That is, the output voltage V _out of the differentially connected secondary coils 9A and 9B is not zero.

The correction sensor 4 that has detected the copper tape 3 in this way outputs a detection signal (detection signal of the copper tape 3) representing the output voltage _Vout to the electronic control unit 20. When the copper tape 3 is not located within the magnetic field generated by the correction sensor 4, the magnetic field caused by the copper tape 3 does not act on the secondary coils 9A and 9B, and the output voltage _Vout becomes zero.

  FIG. 4 is a diagram schematically showing the sensor structure of the flaw detection sensor 11. In this eddy current flaw detector 10, the sensor structure of the flaw detection sensor 11 is the same as that of the correction sensor 4. Therefore, for the flaw detection sensor 11, the description of the same configuration as that of the correction sensor 4 is omitted, and the reference numerals thereof are cited. Moreover, the conventional eddy current flaw detection method may be used for the flaw detection principle which inspects the defect of the steel strip 5 by the flaw detection sensor 11.

When there is no defect in the surface layer of the steel strip 5, the first induced voltage V _{i_1} and the second induced voltage V _{i_2} are equal in the flaw detection sensor 11, and therefore the output voltage Vout of the secondary coils 9A and 9B, which is the difference between them. 0. On the other hand, when the surface layer of the steel strip 5 has a defect, the manner in which the eddy current is generated in the steel strip 5 changes, so that the induced voltages V _{i_1} and V _{i_2} generated in the two secondary coils 9A and 9B of the flaw detection sensor 11 are changed. And the output voltage V _out is not zero. At that time, the flaw detection sensor 11 outputs an inspection signal (detection signal of the steel strip 5) indicating the output voltage _Vout to the electronic control unit 20. Thereby, in the electronic control apparatus 20, it can be determined whether the surface layer of the steel strip 5 has a defect.

  Thus, since the correction sensor 4 and the flaw detection sensor 11 have the same sensor structure, the change with time of the flaw detection sensor 11 is the same as the change with time of the correction sensor 4. Therefore, the sensitivity correction apparatus 1 can correct the sensitivity of the flaw detection sensor 11 while the production line is in operation, with the change in sensitivity due to the change over time by the correction sensor 4 as a representative value. As a result, a plurality of flaw detection sensors 11 (for example, a sensor array) can be brought closer to an inspection with a constant sensitivity.

(1-1. Electronic control device)
FIG. 5 is a block diagram for explaining the configuration of the electronic control unit 20. The electronic control unit 20 is mainly composed of a microcomputer, and is configured to execute an operation according to a predetermined program based on input data and data stored in advance. The electronic control unit 20 receives a signal from the correction sensor 4 (detection signal of the copper tape 3) Sg1 and a signal from the flaw detection sensor 11 (inspection signal of the steel strip 5) Sg2.

The electronic control device 20 includes a detection unit 21, a calculation unit 22, a sensitivity correction unit 23, a correction coefficient determination unit 24, and a notification control unit 25. The detection unit 21 is a means for detecting the detection signal Sg1 of the copper tape 3 input from the correction sensor 4 and the inspection signal Sg2 of the steel strip 5 input from the flaw detection sensor 11. The calculation unit 22 is a means for calculating the correction coefficient α based on the detection value V ₁ (the output voltage V _out of the correction sensor 4) of the copper tape 3 included in the detection signal Sg1 from the correction sensor 4. The correction coefficient α is a value for correcting the sensitivity of the flaw detection sensor 11, and the detection value V ₂ of the steel strip 5 included in the inspection signal Sg2 input from the flaw detection sensor 11 (the output of the steel strip 5). This is a coefficient for correcting the voltage V _out ).

Sensitivity correcting unit 23, by multiplying the correction coefficient α to the detected value V ₂ which is input from the wound sensor 11 probe is a means for correcting the sensitivity of flaw detection sensor 11. The correction coefficient determination unit 24 is a means for determining whether or not the correction coefficient α exceeds a predetermined threshold value β set in advance as upper and lower limit values. The predetermined threshold value β is a threshold value for determining whether the correction coefficient α is a normal value or an abnormal value. When the correction coefficient α is an abnormal value that exceeds the predetermined threshold β, the notification control unit 25 is a means for notifying the outside that the correction coefficient α is abnormal. An output signal from the notification control unit 25 is input to the notification device 6 that is communicably connected to the electronic control device 20.

  The notification device 6 includes a buzzer, a lamp, a monitor, and the like. When a signal is input from the notification control unit 25 to the notification device 6, it is notified to the outside by information that can identify that the correction coefficient α is abnormal, for example, alarm sound, warning light, warning screen, or the like.

(2. Sensitivity correction method)
FIG. 1 is a flowchart showing a processing flow of a sensitivity correction method for an eddy current flaw detector according to this embodiment. This sensitivity correction method is performed after the previous offline calibration is performed until the next offline calibration is performed, that is, while the production line is in operation. Note that offline means that the production line is stopped.

As shown in FIG. 1, the sensitivity correction apparatus 1 detects the copper tape 3 by the correction sensor 4 immediately after the periodic off-line calibration, and acquires the detected value V ₁ as the correction reference value V ₃ ( Step S1). The reference value V _3, the sensitivity reference value to be used for correcting the sensitivity of flaw detection sensor 11 until the next off-line calibration from a previous off-line calibration. The reference value V ₃ obtained in step S1 is stored in the storage means of the electronic control unit 20. Step S1 is performed at the timing of resuming the production line after performing offline calibration, that is, immediately after the production line is in operation and immediately after offline calibration (for example, during the day of resumption).

Sensitivity correction device 1 during production line operations, periodically by correction sensor 4 detects the copper tape 3, and acquires the detected value V ₁ as the current detection value V ₄ (step S2). The current detection value V _4, after obtaining the reference value V ₃ for correction, and the reference value V ₃ is data acquired as a value separate representing the current sensitivity of the correction sensor 4. This step S2 is performed regularly (for example, once a week) while the production line is in operation. Therefore, until the next off-line calibration is performed, the sensitivity correction apparatus 1 to step S2 performed multiple times, stores a plurality acquired current detected value V ₄ in the storage means.

The sensitivity correction device 1 calculates a correction coefficient α that is a ratio of the correction reference value V _{3 to} the current detection value V ₄ (step S3). In the sensitivity correction sensor 4, the reference value V ₃ is a value close to the value of the previous off-line calibration, the current detection value V ₄ is the value after the production line has sensitivity change during operation. Therefore, dividing the current detection value V ₄ at the reference value V _3, the sensitivity of the correction sensor 4 from the time of detecting the reference value V ₃ at step S1 until the time of detecting the current detection value V ₄ in step S2 What A sensitivity change ratio (current detection value V ₄ / reference value V ₃ ) indicating whether or not the degree has changed is obtained. Therefore, in step S3, it is possible to calculate the sensitivity change ratio, calculate the reciprocal of the sensitivity change ratio, and use the reciprocal as the correction coefficient α (reference value V ₃ / current detection value V ₄ ). In short, since may be a value divided by the current detection value V ₄ the correction coefficient α is a reference value V ₃ calculated in step S3, the calculation process, the after calculating the sensitivity change ratio as described above may seek reciprocal, or a reference value V ₃ without calculating the sensitivity change ratio may be divided by the current detection value V _4.

Steps S1 to S3 described above will be described in the case where the correction sensor 4 includes the first correction sensor 4A and the second correction sensor 4B. For example, in step S1, the first detection value V _{1_A} from the first correction sensor 4A is set as the first reference value V _{3_A} , and the second detection value V _{1_B} from the second correction sensor 4B is _set as the second reference value V. _Acquired as _{3_B} . In step S2, and it acquires the second current detection value _{V 4_B} for the first current detection value _{V 4_A} and second correction sensor 4B for the first correction sensor 4A. In step S3, the first sensitivity change ratio (first current detection value _{V4_A} / first reference value _{V3_A} ) for the first correction sensor 4A is calculated, and the second sensitivity sensor 4B for the second correction sensor 4B is calculated. A sensitivity change ratio (second current detection value V 4 — _B / second reference value V 3 — _B ) is calculated. Further, in step S3, an average value of the first sensitivity change ratio and the second sensitivity change ratio is calculated, and the reciprocal of the average value is set as a correction coefficient α. Thus, the sensitivity correction accuracy of the flaw detection sensor 11 can be improved by calculating the correction coefficient α using the average value based on the detection values V ₁ from the plurality of correction sensors 4.

Returning to FIG. 1, the sensitivity correction apparatus 1 determines whether or not the correction coefficient α is a normal value (step S4). In step S4, the correction coefficient α is compared with the predetermined threshold value β, and when the correction coefficient α does not exceed the predetermined threshold value β, it is determined as a normal value, and conversely, when the correction coefficient α exceeds the predetermined threshold value β. Is determined to be an abnormal value. Further, the predetermined threshold β may be an upper limit value and a lower limit value. In this case, when the correction coefficient α is within the range of the predetermined threshold β as the lower limit and the predetermined threshold β as the upper limit, it is a normal value and exceeds the predetermined threshold β as the upper and lower limits Is an abnormal value. For example, correction sensor 4 there is a possibility that some inconvenience obtains an abnormal detection value V ₁ occurring when detecting the copper tape 3. If, when the sensitivity correction of the wound sensor 11 probe using the calculated on the basis of the abnormal detection value V ₁ correction coefficient alpha, appropriate correction can not be made to the original sensitivity change, of the eddy-current flaw detection device 10 Reliability decreases. Step S4 is a process performed in order to suppress such an abnormal state.

When the correction coefficient α is affirmative determination is made in step S4 as a normal value, the sensitivity correction device 1, the sensitivity of the wound sensor 11 probe by integrating the correction coefficient α to the detected value V ₂ from the flaw detection sensor 11 Correction is performed (step S5). For example, wound sensor 11 probe is the case of all 50 channels, by integrating the correction coefficient α to each of all the 50 detection value V _2, thereby to correct the sensitivity of each flaw detection sensor 11. While the production line is in operation, the flaw detection sensor 11 always inspects the steel strip 5, so that the inspection signal Sg <b> 2 from the flaw detection sensor 11 is continuously input to the electronic control unit 20. In step S5, by using the correction coefficient alpha, the sensitivity correction for all of the detected values V ₂ from the plurality of flaw detection sensor 11. Here, when the probe sensitivity of flaw sensor 11 after being calibrated by the last off-line calibration to a predetermined value Va, a value obtained by integrating the correction coefficient α to the detected value V ₂ at step S5, i.e. inspection of the corrected The sensitivity of the sensor 11 is close to a predetermined value Va that is the sensitivity after calibration.

  On the other hand, if the correction coefficient α is determined to be an abnormal value in a negative determination in step S4, the sensitivity correction apparatus 1 notifies the notification means 6 to the outside information that can identify that the correction coefficient α is an abnormal value. (Step S6). When the notification is given, it is possible to take measures such as stopping the sensitivity correction apparatus 1. In short, the eddy current flaw detection apparatus 1 is configured not to perform sensitivity correction of the flaw detection sensor 11 based on the correction coefficient α which is an abnormal value when the correction coefficient α exceeds a predetermined threshold value β.

Further, in the present embodiment, the steps S2 to S5 described above are repeatedly performed at a predetermined cycle (for example, once a week) to correct the sensitivity change actually occurring in the flaw detection sensor 11. It is possible to update to a suitable correction coefficient α. Thereby, the inspection accuracy of the steel strip 5 by the flaw detection sensor 11 can be improved. Specifically, the sensitivity correction device 1, carried out weekly step S2, calculates the correction coefficient alpha _{_1} based on the current detection value V _{4_1} acquired week at step S3, the correction coefficient alpha _{_1} in step S5 correcting the detected value V ₂ from scratches sensor 11 probe week. Then, the sensitivity correction device 1, next week newly calculating the correction coefficient alpha _{_2} in step S3 based on the current detection value V _{4_2} acquisition, the detection value from the wound sensor 11 probe week by the correction coefficient alpha _{_2} in step S5 to correct the V _2. Thus, even when the correction coefficient α is updated in a predetermined period, the reference value V ₃ is the same.

Furthermore, in the present embodiment, the correction coefficient α can be calculated using an average value based on past data. For example, in step S3, the calculation method that the mean value is calculated the average detection value V _{4_Ave,} and described above with reference to average detection value V _{4_Ave} based on the current detection value V ₄ of the past few times acquired in step S2 Thus, the correction coefficient α can be obtained. When the correction coefficient α is calculated using an average value based on the detection values V ₁ from the plurality of correction sensors 4, the correction coefficient (first reference value V _{3_A} / first) for the first correction sensor 4A is calculated. 1 current detection value V _{4_A} ) and a correction coefficient (second reference value V _{3_B} / second current detection value V _{4_B} ) for the second correction sensor 4B is calculated, and then the average value of the correction coefficients Is calculated as the correction coefficient α.

  Here, a sensitivity calibration method that is performed by periodically stopping the production line will be described. The periodic off-line calibration method may be a conventionally known method. FIG. 8A shows an example of a conventionally known periodic offline calibration method. FIG. 8B shows a jig used in the off-line calibration. FIG. 8C and FIG. 8D are diagrams for explaining the relative positional relationship between the sensor leg and the jig when the jig is detected by the flaw detection sensor. 8A to 8D do not show the primary coil 8 and the secondary coil 9 of the flaw detection sensor 11.

  As shown in FIG. 8A, in this off-line calibration method, the off-line calibration jig 40 is arranged below the legs (lower in the third direction) of the flaw detection sensor 11 forming the sensor array. The jig 40 includes a non-metal plate 41 and a conductor 42.

  The non-metal plate 41 is, for example, an acrylic plate or a bake plate, and is formed in a rectangular shape with a predetermined thickness. The length in the longitudinal direction (first direction length) of the non-metal plate 41 is larger than the sensor row width (first direction width) in the flaw detection sensor 11 of the sensor row. In the jig 40, the surface 41a of the non-metal plate 41 is disposed so as to face each leg portion of the flaw detection sensor 11 in the third direction. A tape-like conductor 42 made of a material such as copper, iron, or aluminum is attached to the back surface 41b of the non-metal plate 41.

  As shown in FIG. 8B, the conductor 42 attached to the back surface 41b extends linearly over the entire area along the longitudinal direction (first direction) of the non-metal plate 41. Since the length (first direction length) of the conductor 42 is equal to the length (first direction length) of the non-metal plate 41, it is longer than the sensor row width (first direction width) of the flaw detection sensor 11. As shown in FIGS. 8C and 8D, the width of the conductor 42 (the length in the second direction) is the gap width between one side leg 7b and the center leg 7a in each flaw detection sensor 11 (see FIG. 8C). (Interval in the second direction) and the gap width (interval in the second direction) between the other side leg 7c and the center leg 7a.

  Then, by arranging the jig 40 so that the conductor 42 is located in the gap between the legs of the flaw detection sensor 11 in the second direction, and acquiring the detection value of the conductor 42 by the flaw detection sensor 11, Perform offline calibration. In the measurement state at the time of calibration, each gap portion of the flaw detection sensor 11 and the conductor 42 are arranged so as to be parallel along the first direction.

First, as shown in FIG. 8C, each flaw detection sensor 11 is in a state (first measurement state) in which the conductor 42 is positioned between one of both leg portions 7b and the central leg portion 7a in the second direction. The output voltage _{Vout at} is measured, and the output voltage _Vout is obtained as the first inspection value _{V5_1} . For example, an output signal is input from each flaw detection sensor 11 to the electronic control unit 20 (not shown in FIG. 8). Next, as shown in FIG. 8 (d), each flaw detection sensor 11 is in a state (second measurement state) in which the conductor 42 is positioned between the central leg 7a and the other side leg 7c in the second direction. _Is measured, and the output voltage _Vout is obtained as the second inspection value _{V5_2} .

Then, the electronic control unit 20 calculates a difference Vs between the first inspection value V _{5_1} and the second inspection value V _{5_2} for each flaw detection sensor 11, and calibrates the sensitivity so that the difference Vs becomes the above-described predetermined value Va. To implement. The difference Vs represents the sensitivity of the flaw detection sensor 11 before sensitivity calibration, and the predetermined value Va represents the sensitivity of the flaw detection sensor 11 after sensitivity calibration.

(3. Comparison of sensitivity ratio)
Next, the sensor sensitivity ratio in off-line calibration will be described with reference to FIGS. FIGS. 6 and 7 show cases in which the flaw detection sensors 11 serving as sensor arrays for all 50 channels are targeted.

  FIG. 6 is a diagram showing the sensitivity ratio without correction. The previous value shown in FIG. 6 is the sensitivity (predetermined value Va) of the flaw detection sensor 11 calibrated by the previous offline calibration. Further, the current value shown in FIG. 6 is the sensitivity (difference Vs) of the flaw detection sensor 11 immediately before the current offline calibration.

  FIG. 6 shows how much the sensitivity of the flaw detection sensor 11 has actually decreased due to the change over time. As shown in FIG. 6, when the sensitivity correction by the sensitivity correction apparatus 1 is not performed, the sensitivity of the flaw detection sensor 11 gradually changes as the production line operates during regular offline calibration (for example, for three months). However, immediately before the next off-line calibration, the steel strip 5 is inspected with a sensitivity about 22% lower than the previous off-line calibration (sensitivity represented by the predetermined value Va) with the sensor array average sensitivity of the flaw detection sensor 11. It would have been.

  FIG. 7 is a diagram illustrating the sensitivity ratio with correction when the sensitivity correction is performed by the sensitivity correction apparatus 1. The sensitivity ratio with correction is the sensitivity correction performed by the sensitivity correction device 1 at the timing immediately before the current offline calibration with respect to the sensitivity of the flaw detection sensor 11 at the previous offline calibration (sensitivity represented by the predetermined value Va). It is the ratio of the sensitivity (sensitivity after being corrected with the correction coefficient α) of the subsequent flaw detection sensor 11. That is, FIG. 7 shows the sensitivity ratio as a result of performing the correction close to the predetermined value Va. As shown in FIG. 7, the sensitivity correction apparatus 1 performs sensitivity correction, so that the value after sensitivity correction is a change value within several percent from the predetermined value Va even immediately before the next offline calibration. I understand that. The sensitivity-corrected value is within a range where the deviation from the predetermined value Va is 10%. Thus, by carrying out the sensitivity correction by the sensitivity correction device 1, the inspection accuracy of the steel strip 5 by the flaw detection sensor 11 can be significantly improved (for example, sensitivity accuracy is about 20%).

  As described above, according to the sensitivity correction method and sensitivity correction apparatus of the eddy current flaw detector according to the present embodiment, the sensitivity of the flaw detection sensor can be corrected based on the detection value acquired from the correction sensor during operation of the production line. Thereby, the influence by the time-dependent change of the sensor sensitivity which arises with the sensor for a flaw detection during regular off-line calibration can be suppressed. Therefore, the defect inspection accuracy of the steel strip by the flaw detection sensor can be improved.

  In addition, when a sensor array flaw detection sensor is already provided in the production line, it is easy to add the sensitivity correction apparatus of the present embodiment. That is, it is only necessary to provide correction sensors on both ends of the sensor array, which can be introduced in a space-saving and low-cost manner.

  For example, it is possible to apply this embodiment mentioned above to the pickling line in a steel manufacturing process. In this case, it is possible to detect the bald defect before inserting it into the cold rolling line, which is a subsequent process. Therefore, various measures such as removing the shaving before cold rolling, transmitting the shaving mixing information to the cold rolling line, opening the rolling mill for the shaving, etc. It becomes possible. Therefore, it is possible to prevent the plate breakage and the damage to the rolling roll due to the bald defects.

  In addition, this invention is not limited to embodiment mentioned above, In the range which does not deviate from the objective of this invention, it can change suitably.

  For example, in the above-described embodiment, the two first and second correction sensors 4A and 4B are provided on both sides in the sensor row width direction of the flaw detection sensor 11. However, in the present invention, the number of the correction sensors 4 is the same. Is not particularly limited. In the case of the sensitivity correction apparatus 1 having one correction sensor 4, the configuration may include only the first correction sensor 4A among the correction sensors 4A and 4B, or the second correction sensor 4. The structure provided only with the sensor 4B may be sufficient.

  Further, in the present invention, the detection target of the correction sensor 4 is not limited to the copper tape 3. In short, any conductor that can be detected by the correction sensor 4 can be used, and therefore a metal piece or metal tape made of a metal other than copper can be used. Furthermore, if it is a rectangular metal piece, it does not need to be a tape, and it should just be provided so that the metal piece may not peel from the non-contact surface 2a.

Further, in the correction sensor 4 and the flaw detection sensor 11, an E type in which the secondary coil 9 is wound around the central leg 7a and the two primary coils 8 that are differentially connected are wound around the both leg portions 7b and 7c. It may be a sensor. Even in this case, the output voltage V _out of the secondary coil 9 is because the differential effect is obtained as in the above-described embodiment.

  In addition, the present invention is not a complete replacement for the conventional periodic off-line calibration method, and is preferably used in combination with off-line calibration that is effective for variations in the characteristics of individual flaw detection sensors in the sensor array. This is because in the present invention, the sensitivity of the flaw detection sensor 11 can be corrected in a shorter cycle than when off-line calibration is performed.

DESCRIPTION OF SYMBOLS 1 Sensitivity correction apparatus 2 Nonmetallic roll 3 Copper tape 4 Correction sensor 5 Steel strip 10 Eddy current flaw detection apparatus 11 Flaw detection sensor

Claims (9)

In the method of correcting the sensitivity of the eddy current flaw detection apparatus having a flaw detection sensor comprising a non-contact type eddy current sensor for inspecting a metal strip traveling on a non-metal roll
During the operation of the metal band production line, by a correction sensor comprising a non-contact eddy current sensor, out of the roll surface of the non-metal roll, outside the edge of the metal band in the width direction of the metal band. A detection step of detecting a provided metal piece;
A calculation step of calculating a correction coefficient based on a detection value of the metal piece by the correction sensor;
A sensitivity correction method for an eddy current flaw detection apparatus, comprising: a correction step of correcting the sensitivity of the detected value of the metal band output from the flaw detection sensor during the operation by the correction coefficient.   When the correction coefficient exceeds a predetermined threshold value, the method further includes a notification step of notifying that the correction coefficient has exceeded the predetermined threshold value with identifiable information without performing sensitivity correction by the correction step. The sensitivity correction method of the eddy current flaw detector according to claim 1, wherein The detecting step includes
When the production line resumes operation, detecting the metal piece by the correction sensor, and obtaining a detection value of the metal piece as a reference value;
Detecting the metal piece a plurality of times in a predetermined cycle after obtaining the reference value, and obtaining a detection value of the metal piece as a current detection value,
3. The method of correcting a sensitivity of an eddy current flaw detector according to claim 1, wherein the calculating step includes a step of calculating the correction coefficient that is a ratio of the reference value to the current detection value. The detecting step includes a step of detecting the metal piece by a plurality of the correction sensors,
4. The calculation step according to claim 1, wherein the calculating step includes a step of calculating the correction coefficient using an average value based on detection values of the metal pieces by the plurality of correction sensors. 5. Sensitivity correction method for eddy current flaw detector. In the sensitivity correction apparatus of the eddy current flaw detection apparatus having a flaw detection sensor composed of a non-contact type eddy current sensor for inspecting a metal strip traveling on a non-metal roll,
A metal piece provided on a non-contact surface located outside the edge of the metal band in the width direction of the metal band among the roll surfaces of the non-metal roll;
A correction sensor comprising a non-contact type eddy current sensor arranged to face the non-contact surface and the metal piece;
Detection means for detecting the metal piece by the correction sensor during operation of the metal strip production line;
Calculation means for calculating a correction coefficient based on a detection value of the metal piece from the correction sensor;
A sensitivity correction apparatus for an eddy current flaw detection apparatus, comprising: correction means for correcting sensitivity of the detection value of the metal band output from the flaw detection sensor during the operation by using the correction coefficient. In the case where the correction coefficient exceeds a predetermined threshold value, it further comprises notification means for notifying that the correction coefficient exceeds a predetermined threshold value with identifiable information,
The correction means does not perform sensitivity correction when the correction coefficient exceeds the predetermined threshold, and corrects the sensitivity of the flaw detection sensor by the correction coefficient when the correction coefficient is within the predetermined threshold. The sensitivity correction apparatus for an eddy current flaw detector according to claim 5. The detection means includes
Means for detecting the metal piece by the correction sensor and obtaining the detected value of the metal piece as a reference value when the production line resumes operation;
Means for detecting the metal piece a plurality of times at a predetermined period, and obtaining a detection value of the metal piece as a current detection value,
The sensitivity correction apparatus for an eddy current flaw detector according to claim 5 or 6, wherein the calculation means calculates the correction coefficient that is a ratio of the reference value to the current detection value. The metal piece is a metal tape attached on the non-contact surface,
The eddy current flaw detection according to any one of claims 5 to 7, wherein the metal tape is formed in a predetermined length along an axial direction of the non-metal roll without contacting the metal strip. Device sensitivity correction device. The metal piece is
A first metal piece provided on one non-contact surface located outside one edge of the metal band in the width direction of the metal band;
A second metal piece provided on the other non-contact surface located outside the other edge portion of the metal band in the width direction of the metal band,
The correction sensor is
A first correction sensor for detecting the first metal piece;
A second correction sensor for detecting the second metal piece,
The calculating means calculates the correction coefficient using an average value based on a detection value of the first metal piece from the first correction sensor and a detection value of the second metal piece from the second correction sensor. 9. The sensitivity correction device for an eddy current flaw detector according to claim 5, wherein the sensitivity correction device calculates the sensitivity.

Images