JP2015175761A - Surface flaw detection method and surface flaw detection device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a surface flaw detection method and a surface flaw detection device with which a precise approximate circle is found from shape profile data of a steel bar rolled material and the surface flaw caused in the steel bar rolled material can be detected reliably and precisely on the basis of the approximate circle.SOLUTION: A surface flaw detection method of the present invention includes the steps of: extracting a two-dimensional profile of a cross-sectional shape of a steel bar rolled material W; using each plotting coordinate Aconstituting the two-dimensional profile and calculating an approximate circle Cof the two-dimensional profile; finding a distance d between a center coordinate O and a plotting coordinate Aof the approximate circle C, finding a difference between the distance d and a radius R of the approximate circle C, and calculating a depth H in the two-dimensional profile; determining that a surface flaw candidate Zexists in the two-dimensional profile when the depth H is larger than a predetermined threshold value; and determining that the surface flaw Z is caused in the steel bar rolled material W when multiple determined two-dimensional profile are caused in the steel bar rolled material W and the number of the two-dimensional profiles are more than the predetermined threshold value.

Description

  The present invention relates to a surface flaw detection method and a surface flaw detection device for detecting a surface flaw of a rolled material that has been rolled when rolling a rolled material such as a wire, a steel bar, or a steel pipe.

Conventionally, when a rolled material having a substantially circular cross-sectional shape such as a wire rod or steel bar is manufactured from a slab such as a billet, it has been performed in a hot rolling process or a cold rolling process.
In the hot rolling process and the cold rolling process, a rolling facility in which a rough rolling mill, an intermediate rolling mill, a finishing rolling mill, etc. are arranged is provided. Rolling is repeated until a rolled material (product) having a cross-sectional shape is obtained.

In a rolled material having a substantially circular cross-sectional shape manufactured in such a hot and cold rolling process, defects such as surface defects (such as a pressing bar having a concave shape) may occur on the outer peripheral surface thereof. is there. If a fine defect such as a surface flaw occurs on the outer peripheral surface of the rolled material, the quality of the rolled material will deteriorate, and the product cannot be shipped as a product.
Also, if a fine defect generated on the outer peripheral surface of the rolled material is missed during the flaw detection, defective products of the rolled material in which defects exist will be shipped without being excluded, which is very large in the manufacturing process of the rolled material. It becomes a problem.

Therefore, in order to reliably determine and ship a rolled material with no fine defects (surface defects) as a non-defective product, various nondestructive inspection methods are used to detect defects in the rolled material such as wire rods, steel bars, and steel pipes. Is automatically detected, and the detected defects are removed by a grinding device or the like, or the rolling conditions in each rolling process are corrected based on the detection result.
Electromagnetic flaw detection such as eddy current flaw detection and leakage magnetic flux flaw detection has been widely used as means for detecting defects generated on the outer peripheral surface of such rolled materials such as wire rods, steel bars, and steel pipes.

However, in the above-mentioned electromagnetic flaw detection means, the noise may be detected as a defect due to vibration of the rolling material generated in the rolling process, noise generation due to uneven electromagnetic characteristics, etc. There is a possibility that even a defect that does not cause a problem is detected (defect overdetection).
In recent years, an image flaw detection technique for flaw detection of defects on the outer peripheral surface of a rolled material using a high-speed camera has been developed for such a problem. Image flaw detection technology, for example, irradiates a rolled material with light by an illumination means, captures irregular reflection components of surface defects (defects) caused by the irradiation of light, or detects surface defects caused by light irradiation. This is a technique for detecting the surface defects generated on the outer peripheral surface of the rolled material by capturing the shadow.

However, in the image flaw detection technology using the above-described high-speed camera, there is a risk that the surface defects generated on the outer peripheral surface of the rolled material are shallow or the fine surface defects such as the dents of the surface defects being loose may be overlooked. It becomes a very big problem in the manufacturing process of the rolled material.
In response to such problems, the outer shape of the rolled material such as wire rods, steel bars, and steel pipes is directly measured using a high-speed camera in the same manner as described above, and the dents and depths formed on the outer peripheral surface of the rolled material are measured. In addition, a surface wrinkle detection method and a detection device capable of detecting whether or not the dent is a surface wrinkle have been developed.

Examples of image processing techniques for detecting surface defects of rolled materials such as wire rods, steel bars, and steel pipes include those disclosed in Patent Literature 1 and Patent Literature 2.
Patent document 1 projects a plurality of laser slit lights on the surface of a steel pipe drilling plug from a direction perpendicular to the axial direction of the plug, captures and acquires an image of the external shape of the plug from the front in the axial direction of the plug, The acquired image is image-processed to obtain a line drawing of the outer shape, an approximate circle of the line drawing is created based on the line drawing, the line drawing of the outer shape obtained by image processing, and the approximate circle of the created line drawing A surface wrinkle inspection apparatus for determining wrinkles based on the magnitude of the deviation is disclosed.

Patent Document 2 is an image processing method for measuring the center position of a test object whose contour shape is circular in an image. A circle approximation is performed from the positions of a plurality of points in the contour portion of the test object. And calculate the center and radius of the test object, calculate the distance from the calculated center of the test object to the point of the contour part for each point of the contour part, and calculate the distance to the calculated point of the contour part and the test object. The approximate circle of the contour part is obtained by excluding points where the difference from the radius of the inspection is greater than a certain threshold, and the distance from the center of the obtained approximate circle to the point of the contour part and the residual of the approximate circle radius is the contour part. Calculate the sum of the obtained residuals, the difference between the maximum and minimum residual values, or the sum of the variances of the differences for each point of the contour, and sum the residuals and the maximum of the residuals. Circle approximation is performed from the combination of the points of the contour when the difference between the value and the minimum value and the sum of the variances of the differences are minimized. Te, an image processing method of calculating the center of the test object, and the radius is disclosed.

Japanese Patent No. 2910567 Japanese Patent No. 4919156

However, even if the technique for detecting the surface flaw of the object to be measured as shown in Patent Document 1 and Patent Document 2 is used as a method for detecting the surface flaw of the rolled material, there are the following problems.
That is, the techniques of Patent Documents 1 and 2 create an approximate circle from a plurality of position data on the contour of the test object, and detect the presence or absence of wrinkles and an abnormality in the diameter of the rolled material based on the created approximate circle. Is. However, in obtaining an approximate circle, it is often impossible to obtain a correct approximate circle unless a plurality of position data on the contour are appropriately processed.

  For example, when surface flaws exist on the outer peripheral surface of a rolled material such as a wire rod, steel bar, or steel pipe, calculating the approximate perfect circle of the outer peripheral surface of the rolled material from all the measurement points of the rolled material, the value of the surface flaw depth Since the value is smaller than the actual wrinkle depth, the surface wrinkles of the rolled material that are actually generated may be missed. The techniques of Patent Documents 1 and 2 described above do not disclose a technique corresponding to the above problem.

  Therefore, in view of the above problems, the present invention obtains an accurate approximate circle from shape profile data of a rolled material such as a wire rod, steel bar, and steel pipe, and is generated on the outer peripheral surface of the rolled material based on the obtained approximate circle. An object of the present invention is to provide a surface flaw detection method and a surface flaw detection device capable of reliably and accurately detecting a concave surface flaw.

In order to achieve the above object, the present invention takes the following technical means.
The surface flaw detection method according to the present invention is a surface flaw detection method for detecting a surface flaw of a long rolled material having a substantially circular cross-sectional shape, wherein the rolling is carried such that the longitudinal direction is the carrying direction. For a material, a two-dimensional profile having a cross-sectional shape facing the direction perpendicular to the longitudinal direction of the rolled material is extracted, and an approximate circle C of the two-dimensional profile is extracted using each plot coordinate A _i constituting the extracted two-dimensional profile. calculates _b, the calculated the approximate circle C _b on the center coordinates O and sought distance d between each plot coordinates a _i of the two-dimensional profile, obtained distance d and calculated the approximate circle C _b A depth H in the two-dimensional profile is calculated by obtaining a difference from the radius R. When the calculated depth H in the two-dimensional profile is greater than a predetermined threshold, the two-dimensional profile Determines a candidate Z _k of surface defects are present, the two-dimensional profile candidate Z _k of the surface defects is determined that there is a plurality occurs in the longitudinal direction of the rolled material, and the plurality generated two-dimensional profiles When the number is larger than a predetermined threshold value, it is determined that surface flaws Z are generated in the rolled material.

Preferably, when calculating the approximate circle C _b of the two-dimensional profile of all plots coordinates A _i indicating the two-dimensional profile, plotted coordinates excluding the plot coordinates A _z suspected of the surface flaw Z with a _(i-z), it may approximate circle C _b of the two-dimensional profile is calculated by the least squares method.
Preferably, in order to exclude the plot coordinate A _z estimated to be the surface 疵 Z from all the plot coordinates A _i indicating the two-dimensional profile, the plot coordinates for each plot coordinate A _i of the two-dimensional profile plot coordinates _a i-1 is adjacent to _{a i,} a _{i + 1}
Of using the data of at least three or more plot coordinates A _i, a diameter L _a of the approximate circle C _a in the plot coordinates A _i of the two-dimensional profile is calculated by the least square method, calculated the approximate circle C _a the diameter L _a, the absolute value of the difference between the actual diameter L of the rolled material (| L a _-L |) asking, if the absolute value is larger than a predetermined threshold value, the plot coordinates a _i, the 2 may be excluded from the plot coordinates a _(i-z) to be used when calculating the approximate circle C _b dimension profile.

Preferably, when the surface defect Z of the rolled material is detected, if the degree of occurrence of the plot coordinates A _i determined that the surface defect Z exists in the two-dimensional profile is equal to or greater than a predetermined threshold, the two-dimensional The profile may be determined as a defective profile.
Preferably, when the degree of occurrence of the two-dimensional profile determined as the defective profile in the longitudinal direction of the rolled material is equal to or greater than a predetermined threshold, the rolled material may be determined as a defective product.

Preferably, the surface flaw detection method according to the present invention is performed in real time.
The surface flaw detection apparatus according to the present invention is a surface flaw detection apparatus that detects a surface flaw of a long rolled material having a substantially circular cross-sectional shape, and the rolling is carried such that the longitudinal direction is the carrying direction. With respect to the material, the profile extracting means for extracting a two-dimensional profile having a cross-sectional shape facing the direction perpendicular to the longitudinal direction of the rolled material, and the rolled material by the surface flaw detection method according to the present invention using the extracted two-dimensional profile And a wrinkle detector for detecting surface wrinkles.

  Preferably, the profile extraction means includes at least one two-dimensional displacement laser displacement meter arranged along a circumferential direction of the rolled material, and the two-dimensional displacement laser displacement meter It may be configured to extract a two-dimensional profile.

  According to the surface flaw detection method and the surface flaw detection apparatus of the present invention, an accurate approximate circle is obtained from the shape profile data of a rolled material such as a wire, a steel bar, and a steel pipe, and based on the obtained approximate circle, It is possible to reliably and accurately detect concave surface defects generated on the outer peripheral surface.

It is the figure which showed an example of schematic structure of the surface flaw detection apparatus of this invention. It is the figure which showed an example of the installation place of the two-dimensional laser displacement meter arrange | positioned by the surface flaw detection apparatus of this invention. It is the figure which showed the other example of the installation place of the two-dimensional laser displacement meter arrange | positioned by the surface flaw detection apparatus of this invention. It is the figure which showed the result of having measured the profile shape of the normal steel bar rolled material. It is the figure which showed the result of having measured the profile shape of the rolled steel bar in which surface flaw generate | occur | produced in the outer peripheral surface. It is the figure which showed what arrange | positioned continuously the two-dimensional profile of the steel bar rolling material extracted with the surface flaw detection apparatus of this invention in the longitudinal direction of the steel bar rolling material. It is the flowchart figure which showed the method of detecting the surface flaw of a steel bar rolled material. It is the figure which compared the calculated approximate circle with the calculation method of the approximate circle in the two-dimensional profile of a rolled steel bar, and the profile shape of the rolled steel bar in which surface flaws occurred on the outer peripheral surface. It is the figure which showed the depth profile in the two-dimensional profile of a normal steel bar rolled material (corresponding to FIG. 3A). It is the figure which showed the depth profile in the two-dimensional profile of the steel bar rolling material which the surface flaw generate | occur | produced in the outer peripheral surface (corresponding to FIG. 3B). It is the figure which showed an example of schematic structure of the inspection apparatus of a steel bar rolled material.

Hereinafter, the surface wrinkle detection method and the surface wrinkle detection apparatus of the present invention will be described with reference to the drawings.
In addition, embodiment described below is an example which actualized this invention, Comprising: It is not for limiting the structure of this invention only to the specific example. Therefore, the technical scope of the present invention is not limited only to the disclosed contents of the present embodiment.

For example, in the following description, the rolled material to be inspected will be described as a rolled steel bar, but the invention is not limited thereto. Applicable.
First, before explaining the surface flaw detection apparatus 1 and the surface flaw detection method of the present invention, the inspection apparatus 10 (inspection line) for the rolled steel bar W provided in the rolling facility will be described with reference to FIG. To do.

  In FIG. 8, a rolled steel bar W is manufactured by continuously rolling a billet or the like in a steel bar rolling facility and then cutting the billet into a predetermined length. An example of the inspection apparatus 10 which inspects the outer peripheral surface and the inside of the rolled steel bar W is shown. The rolled steel bar W of the present embodiment has a substantially round cross section, and is a long bar cut to a diameter of about 30 mm and a length of about 10 m, for example.

As shown in FIG. 8, the inspection apparatus 10 corrects the bending of the long rolled steel bar W by heat treatment or the like in order from the upstream side to the downstream side, and corrects the straight rolled steel bar W. A machine 11, an outer peripheral surface flaw detector 12 for detecting a surface defect Z of the rolled steel bar W generated in the rolling process, and an internal defect flaw detector 13 for detecting an internal defect of the rolled steel bar W are provided.
Further, on the downstream side of the straightening machine 11 and the upstream side of the outer peripheral surface flaw detection device 12, the surface defect Z of the rolled steel bar W fed from the straightening machine 11 is detected, as will be described in detail later. A surface flaw detection device 1 is provided.

The straightening machine 11 is a device that continuously straightens the rolled steel bar W bent by heat treatment or the like with, for example, two hourglass rolls and forms the rolled steel bar W with high straightness.
The outer peripheral surface flaw detection apparatus 12 of the present embodiment is an apparatus that detects (flaw detection) flaws existing on the surface of a long rolled steel bar W using a leakage magnetic flux, for example, a leakage magnetic flux inspection apparatus (MLFT). Etc.

Further, the internal defect inspection device 13 of the present embodiment is a device that detects (detects) defects existing under the surface of the long steel bar W using ultrasonic waves. For example, an ultrasonic inspection device ( (UST).
As described above, the rolled steel bar W manufactured by the steel bar rolling facility is inspected on the surface and the inside by two (leakage magnetic flux, ultrasonic) flaw detectors, and the rolled steel bar on which the surface defect Z and the internal defect are not detected. Only the Wa is carried out (shipped) as a non-defective product, and the rolled steel bar Wb in which the surface defect Z is detected is discharged from the inspection apparatus 10 (inspection line).

  However, even if the above-described leakage magnetic flux flaw detector 12 is used, the ridge Z generated on the surface of the rolled steel bar W may not be detected. For example, a surface defect Z (for example, a pressing defect) is formed on the surface of the front end portion and the rear end portion of the rolled steel bar W cut with a bite in a state in which a foreign object is sandwiched, but the leakage magnetic flux inspection The device 12 has a dead zone that is not detected even if there is a surface flaw Z, and the front flaw and the rear end portion of the rolled steel bar W are not detected and may be overlooked. .

In view of the problem, the inventors of the present invention invented a surface wrinkle detection device 1 that can detect all wrinkles Z generated on the surface of the rolled steel bar W, such as wrinkles Z generated on the surfaces of the front end portion and the rear end portion. In addition, the inspection apparatus 10 is arranged between the correction machine 11 and the outer peripheral surface flaw detection apparatus 12.
Hereinafter, the surface flaw detection apparatus 1 of the present invention will be described in detail.
As shown in FIG. 1, the surface flaw detection apparatus 1 of the present invention relates to a rolled steel bar W that is transported such that the longitudinal direction is the transport direction, and a cross section that faces in a direction perpendicular to the longitudinal direction of the rolled steel bar W. A profile extracting means 2 for extracting a two-dimensional profile of the shape, and a flaw detector 3 for detecting a surface flaw Z of the rolled steel bar W using the extracted two-dimensional profile.

Flaw detector 3, a three-dimensional shape acquiring unit 4 for acquiring the three-dimensional shape of bars rolled material W, and the approximate circle calculating means 5 for calculating an approximate circle C _b of the two-dimensional profile of the steel bar rolled material W, 2-dimensional a depth calculating means 6 for calculating the depth H of the recesses present in the profiles, and the depth H of the two-dimensional profile, based on a predetermined criteria, a candidate Z _k of a surface defect in a two-dimensional profile exists determined a surface flaw candidate determining means 7 for a two-dimensional profile is determined as a candidate Z _k of surface defects are present on the basis of predetermined criteria, it determines that the surface defects Z to steel bar rolling material W is generated surface And wrinkle determination means 8.

Details of each means will be described below.
First, the profile extracting means 2 is a two-dimensional laser displacement meter 2 (laser type profile measuring instrument, for example, sampling frequency: 2 kHz), and the longitudinal length of the rolled steel bar W (for example, 150 m / min) (for example, 150 m / min) A two-dimensional profile of a cross section (measured cross section) cut in a direction perpendicular to the (axis length) direction is extracted, and at least one or more are provided along the circumferential direction of the rolled steel bar W (conveyance) Direction resolution 1.25 mm).

For example, as shown in FIG. 2A, in order to extract the contour shape of the entire circumference of the rolled steel bar W as a two-dimensional profile, the profile extracting means 2 is equidistant along the circumferential direction of the rolled steel bar W (in FIG. 2A). In addition, three of them may be arranged on an extension line having a diameter directed in the vertical direction with respect to the axial length direction of the rolled steel bar W.
If a plurality of two-dimensional laser displacement meters 2 having the same wavelength are arranged at the same position in the circumferential direction of the rolled steel bar W, the laser light S from each two-dimensional laser displacement meter 2 interferes and a two-dimensional profile is correctly extracted. As shown in FIG. 2A, the two-dimensional laser displacement meter 2 using laser light S having different wavelengths may be used, or as shown in FIG. The two-dimensional profiles are correctly extracted by shifting them by a predetermined interval in the axial length direction (vertical arrangement).

2B is arranged at equal intervals along the circumferential direction of the rolled steel bar W as in FIG. 2A (see the view taken along the line BB in FIG. 2B). ).
In this way, by providing a plurality of profile extraction means 2, the contour shape can be captured in real time over the entire circumference in the axial direction of the rolled steel bar W, and the contour shape of the entire circumference of the rolled steel bar W Can be extracted as a two-dimensional profile.

That is, by using the profile extraction means 2 described above, the image of the measured cross section can be processed in real time at a speed corresponding to the measured cross section of the rolled steel bar W, for example. Can be extracted as a two-dimensional profile.
3A and 3B show a two-dimensional profile of the cross-sectional shape of the extracted steel bar rolled material W. FIG.

The two-dimensional profile of the cross-sectional shape of the rolled steel bar W shown in FIG. 3A is a shape depicting a constant parabola, and the surface defect Z is not generated in the two-dimensional profile.
In FIG. 3A, the two-dimensional profile shape of the cross section of the rolled steel bar WW is a symmetrical shape with the position (x coordinate) being approximately 17.5 mm and the height (y coordinate) being approximately 50 mm as a vertex. Can be confirmed.

The data of the two-dimensional profile of the cross-sectional shape of the rolled steel bar W having a constant curved surface shape is stored for each size of the rolled steel bar W in advance as basic data, and approximated to the two-dimensional profile. used to calculate the circle C _b.
On the other hand, the two-dimensional profile of the cross-sectional shape of the rolled steel bar W shown in FIG. 3B is a parabolic shape in which the shape is partially changed (concave), and surface defects Z are generated in the two-dimensional profile. Is.

In FIG. 3B, the shape of the two-dimensional profile of the cross-sectional shape of the rolled steel bar W changes between a position (x coordinate) of approximately 22.5 mm to 33 mm and a height (y coordinate) of approximately 40 mm to 50 mm. Can be confirmed.
Thus, by measuring the contour shape of the rolled steel bar W being conveyed by the profile extracting means 2, the two-dimensional profile of the rolled steel bar W can be extracted in real time. The information of the two-dimensional profile of the rolled steel bar W obtained by the profile extracting means 2 is sent to the three-dimensional shape acquiring means 4 in the wrinkle detector 3 to acquire the three-dimensional shape of the rolled steel bar W. Used.

The profile extracting means 2 of the present embodiment measures the shape of the laser beam under the same conditions as irradiating a plurality of laser beams S (slit beams) in parallel at a narrow pitch to the rolled steel bar W during high-speed conveyance. Therefore, the shape of the rolled steel bar W can be measured at high speed.
Moreover, since the profile extraction means 2 is measuring with the sampling frequency sufficiently high with respect to the conveyance speed of the rolled steel bar W, the profile extracting means 2 is more than the standard value of the length of the surface flaw Z in the longitudinal direction of the rolled steel bar W. It has a function to measure a two-dimensional profile with fine resolution.

As shown in FIG. 4, the three-dimensional shape acquisition means 4 provided in the wrinkle detection unit 3 continuously samples in the longitudinal direction of the rolled steel bar W based on the information of the two-dimensional profile, and A three-dimensional shape over the entire length of the material W is acquired.
The approximate circle calculation means 5 provided in the eyelid detection unit 3 is configured with respect to each two-dimensional profile of the two-dimensional profile stored in the three-dimensional shape acquisition means 4 with plot coordinates ( using the measurement point) a _i, and calculates an approximate circle C _b of the two-dimensional profile.

Specifically, the approximate circle C _a is calculated by the least square method using all the data of the measurement data (plot coordinates A _i ) constituting the two-dimensional profile.
However, in the present embodiment, the following processing method is adopted in order to improve the accuracy of the obtained approximate circle C _a .
That is, a procedure of calculating an approximate circle C _b of the two-dimensional profile with high accuracy, and plots the coordinate A _i obtained by the profile extraction unit 2, at least the plot coordinates A _{i-1, A} i + ₁ adjacent to the A _i The diameter L _a (radius R _a ) of the approximate circle of the two-dimensional profile is calculated using data of three or more plot coordinates A _i (for example, data of 11 points). Then, based on the calculated approximate circle information, the plot coordinates A _z estimated as the surface 疵 Z are excluded from all the plot coordinates A _i indicating the two-dimensional profile.

The procedure for excluding the plot coordinate A _z that is estimated to be the surface 疵 Z is, first, for each plot coordinate A _{i and} each plot coordinate A _i of the two-dimensional profile, for example, the plot coordinate A _i adjacent to the plot coordinate A _{i. i} using the data of the left and right of the plot coordinates _{a i-1, a i + 1} for at least three or more plot coordinates _{a i,} calculated by the least squares method approximate circle _{C a} in the plot coordinates _{a i} of the two-dimensional profile, obtained The diameter L _a (radius R _a ) is obtained from the obtained approximate circle C _a .

Then, the diameter L _a of the calculated approximate circle, an absolute value of the difference between the actual diameter L of steel bar rolled material W Request (| | L a _-L). When the obtained absolute value is larger than a predetermined threshold, the plot coordinate A _i is set to the plot coordinate A _z estimated to be the surface 疵 Z, and the plot coordinate A _(i used in calculating the approximate circle C _b of the two-dimensional profile is used. Exclude from _-z) .
This exclusion procedure is performed for all plot coordinates A _{i, and} all corresponding plot coordinates A _z are excluded from the plot coordinates A _(iz) .

Thus, by using a surface flaw Z and deduced plotted coordinates A _z plotted coordinates obtained by eliminating A _(i-z), is calculated by the least squares method the approximate circle C _b of the two-dimensional profile.
The depth calculation means 6 provided in the eyelid detection unit 3 calculates the depth H of the dent existing in the two-dimensional profile.
As shown in the upper right side in FIG. 6, the procedure of calculating the depth H of the recess present in the two-dimensional profile, the center coordinates O of the approximate circle C _b of the first two-dimensional profile calculated by the approximate circle calculation means 5 The distance d from each plot coordinate A _i of the two-dimensional profile is obtained.

Then, the difference (R _b −d) between the obtained distance d and the radius R _b of the approximate circle C _b calculated by the approximate circle calculation means 5 is calculated for each plot coordinate A _i . Among the calculated differences, a value having the largest difference is set as a depth H in the two-dimensional profile.
Surface defect candidate determination unit 7 is to determine the surface defects candidate Z _k that may surface defects Z prior step of determining a surface flaw Z.

If the depth H of the two-dimensional profile calculated by the depth calculation unit 6 is greater than a predetermined threshold value, the surface defect candidate determination unit 7 adds the surface defect candidate Z to the two-dimensional profile. If it is determined that _k exists and is smaller than the predetermined threshold, it is determined that there is no surface defect candidate Z _{k in} the two-dimensional profile.
Surface flaw determination unit 8, two-dimensional profile candidate Z _k of surface defects is determined to exist, in the longitudinal direction of the steel bar rolled material W, for example, continuously multiple occurrences, and the plurality generation 2D profile When the number is larger than a predetermined threshold, it is determined that the surface defect Z is generated in the two-dimensional profile.

Procedure for detecting surface flaws Z of steel bar rolled material W is the degree of generation of plotting coordinate A _k which is determined surface flaws Z _k in the two-dimensional profile is present, if not less than the predetermined threshold value, the two-dimensional profile It is determined as a defective profile.
And when the generation | occurrence | production degree which the determined defect profile generate | occur | produces in the longitudinal direction of the steel bar rolling material W is more than a predetermined threshold value, it detects that surface flaw Z has generate | occur | produced in the outer peripheral surface of the actual steel bar rolling material Wb. Then, the steel bar rolled material Wb is determined as a defective material.

Surface flaw detection apparatus 1 of this embodiment, in order to detect surface flaws Z present in steel bar rolled material W, the distance of each measurement point from the center coordinates O of the approximate circle C _b of the two-dimensional profile Ai + p (plotted coordinates) and d, and calculates the difference between the radius R _b of the approximate circle C _b, to the exclusion of the surface flaw candidate Z _{k (the} point where the irregularities) for calculating the actual depth H of the surface defect Z, 2-dimensional since the calculated approximate circle C _b of the profile, it can be determined that the surface flaw Z accurately in a two-dimensional profile is present.

Moreover, can be readily selected by the excluded surface flaw candidate Z _k as the reference standards size bars rolled material W.
Next, a method for detecting the surface defect Z of the rolled steel bar W using the surface defect detection device 1 described above will be described with reference to FIGS.
FIG. 5 is a flowchart showing a method of detecting the surface defect Z of the rolled steel bar W.

FIG. 6 is a diagram comparing the calculated approximate circle with the profile shape of the rolled steel bar Wb having surface defects Z on the outer peripheral surface in accordance with the approximate circle calculation method in the two-dimensional profile of the rolled steel bar W.
FIG. 7A is a diagram showing a depth profile in a two-dimensional profile of a normal rolled steel bar W (corresponding to FIG. 3A). FIG. 7B is a diagram showing a depth profile in the two-dimensional profile of the rolled steel bar Wb with surface defects Z on the outer peripheral surface (corresponding to FIG. 3B).

First, in S1, the outer peripheral surface of the rolled steel bar W being conveyed at high speed is irradiated with the laser light S, the irradiated laser light S is received, and the rolled steel bar is based on the received laser light S. A two-dimensional profile of W is extracted. By sampling continuously in the longitudinal direction of the rolled steel bar W, a three-dimensional shape in the entire length of the rolled steel bar W is obtained.
From the left side of the image (eg, FIG. 3A), A ₀ , A 1,... A _i , A _{i + p} (i = 0, p = Even number, j = 0). For example, the measurement points A _{i + p} of the two-dimensional profile are 11 points.

In S2, the approximate circle C _a of each measurement point (A _{0 to} A _{i + p} ) is calculated using the measurement point (A _{0 to} A _{i + p} ) (plot coordinates) of the two-dimensional profile set in S1.
In S3, the absolute value of the difference (|) between the radius R _a of the approximate circle C _a of each measurement point (A _{0 to} A _{i + p} ) calculated in S2 and the actual radius R (diameter standard value) of the rolled steel bar W R−R _a |) is calculated. When the calculated absolute value (| R−R _a |) is smaller than the threshold value of 1 mm, the process proceeds to S4. On the other hand, when the calculated absolute value (| R−R _a |) is larger than the threshold value of 1 mm, the process proceeds to S5.

In S4, the central point of the measurement point (A _{0 to} A _{i + p} ) of the two-dimensional profile set in S1 is set as a coordinate B _j (coordinate in the circumferential direction) for calculating the overall approximate circle C _b , and then j is Increment (j = j + 1).
In S5, it is determined whether or not A _{i + p} of the measurement point of the two-dimensional profile is the final measurement point. When A _{i + p} of the measurement point of the two-dimensional profile is the final measurement point, the process proceeds to S6. On the other hand, if A _{i + p} of the measurement point of the two-dimensional profile is not the final measurement point, i is increased to i + 1, and the process returns to S2.

In S6, using the center point _{B 0, B 1, B 2} ··· B j of the measurement point of the two-dimensional profile set in S4, and calculates the approximate circle _{C b} of the entire two-dimensional profile.
Here, by using the least squares method, a procedure of calculating an approximate circle C _b of the two-dimensional profile, it will be described.
The contour shape of the two-dimensional profile is expressed by equation (1).

  The contour shape of the two-dimensional profile represented by Expression (1) is developed into Expression (2).

The center coordinates of the two-dimensional profile (a, b) and radius _{R b,} shown in equation (3).

The sum of the left side when substituting each measurement point (x ₁ , y ₁ ), (x ₂ , y ₂ ),..., (X _i , y _i ) of the two-dimensional profile into the left side of equation (1) , The difference from the sum of the right side (= 0) is minimized, that is, when the sum of the left side of Equation (1) is close to 0, the sum of the squares of the left side is also minimized, so the equation ( The values of the coefficient A, coefficient B, and coefficient C on the left side of 1) are obtained.
Where the sum of the squares of the left side,

  Since the value obtained by partial differentiation with the coefficient A, the coefficient B, and the coefficient C becomes 0, Equation (4) is established.

Then, the coefficient A, the coefficient B, and the coefficient C are calculated from the equation (4), and the approximate circle C _a of each measurement point A _i is calculated.
Using the procedure described above, the center point of the measurement point of the two-dimensional profiles _{B 0,} B _1, B 2 · · · _{B j} based, obtaining an approximate circle _{C b} fitting the two-dimensional profile. The calculation procedure of such approximate circle C _a, longitudinal all 2-dimensional profile of the steel bar rolled material W (e.g., the length steel bar rolled material W of 5 m, there is a two-dimensional profile of 4000) in Against.

In S7, as shown in FIG. 6 calculates the distance d between the center coordinate O of the approximate circle _{C b} of the entire two-dimensional profile calculated, each of the measurement points of the two-dimensional profile and _(A 0 _{~A i + p)} in S6. And the distance d determined, calculates a difference between the radius R _b of the approximate circle C _b of the entire two-dimensional profile, that the difference between the depth H in the two-dimensional profile.
As shown in FIG. 7A, when the calculated depth H is smaller than a set threshold (for example, 0.4 mm), it is determined that there is no surface defect Z in this two-dimensional profile, and the process proceeds to S8.

On the other hand, as shown in FIG. 7B, the calculated depth H is a set threshold value (for example, 0.4 m).
If it is larger than m), it is determined that there is a surface defect Z in this two-dimensional profile, and a warning (warning with surface defect Z) is issued to the outside. Then, the process proceeds to S8.
In S8, it is determined whether or not A _{i + p} of the measurement point of the two-dimensional profile is the final measurement point. If A _i of the measurement point of the two-dimensional profile is the last measuring point, and terminates the determination of the two-dimensional profile shape. On the other hand, if A _i of the measurement point of the two-dimensional profile is not the final measurement point, i is increased to i + 1, and the process returns to S7.

When the determination of the shape of the two-dimensional profile is finished, the surface defect Z of the rolled steel bar W (to be precise, the surface defect candidate Z _k ) is detected based on the two-dimensional profile.
For example, when the conveyance speed is 150 m / min, the sampling frequency is 2 kHz, and the threshold value of the length of the surface defect Z is set to 5 mm, it is determined that the surface defect Z (to be precise, the surface defect candidate Z _k ) exists. If four or more two-dimensional profiles are continuously generated in the longitudinal direction, it is determined that the surface defects Z equal to or greater than the threshold exist on the outer peripheral surface of the rolled steel bar W. And the steel bar rolling material Wb in which the surface flaw Z exists in an outer peripheral surface is detected as a defective material.

  As described above, by using the surface flaw detection method and the surface flaw detection apparatus 1 of the present invention, the contour shape of the entire length of the rolled steel bar W during high-speed conveyance is measured online, and the measured rolled steel bar W is measured. An accurate approximate circle is obtained from the shape profile data, the surface defect Z depth H is calculated based on the obtained approximate circle, and the presence or absence of the surface defect Z on the outer peripheral surface of the rolled steel bar W is reliably and accurately determined. It becomes possible.

And based on this determination result, by removing the steel bar rolled material Wb (inferior material) in which the surface wrinkle Z has occurred from the bar steel rolling line (inspection line 10), the outflow of the steel bar rolled material Wb in which the surface wrinkle Z has occurred, The effect that it can prevent being shipped as a product is brought about.
Further, by calculating the actual depth H of the surface flaw Z on the outer peripheral surface of the rolled steel bar W, it can be compared with the shape specification of the rolled steel bar W. Moreover, the difference between the size of the rolled steel bar W and the standard size (for example, standard diameter) of the rolled steel bar W can be calculated, and a diameter abnormality can be detected at all positions in the longitudinal direction. it can.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. In particular, in the embodiment disclosed this time, matters that are not explicitly disclosed, for example, operating conditions and operating conditions, various parameters, dimensions, weights, volumes, and the like of a component deviate from a range that a person skilled in the art normally performs. Instead, values that can be easily assumed by those skilled in the art are employed.

1 Surface wrinkle detection device 2 Profile extraction means (two-dimensional laser displacement meter)
3 Flaw detection unit 4 Three-dimensional shape acquisition means 5 Approximate circle calculation means 6 Depth calculation means 7 Surface wrinkle candidate determination means 8 Surface wrinkle determination means 10 Inspection apparatus (inspection line)
11 Straightening machine 12 Outer peripheral surface flaw detector (leakage magnetic flux flaw detector)
13 Internal flaw detector (ultrasonic flaw detector)
S Laser beam W Rolled steel bar Wa Rolled steel bar (no surface defects)
Wb Rolled steel bars (with surface defects)

Claims (8)

A surface flaw detection method for detecting a surface flaw of a long rolled material having a substantially circular cross-sectional shape,
For the rolled material that is conveyed so that the longitudinal direction is the conveying direction, extract a two-dimensional profile of a cross-sectional shape that faces in a direction perpendicular to the longitudinal direction of the rolled material,
Using each plot coordinate A _i constituting the extracted two-dimensional profile, an approximate circle C _b of the two-dimensional profile is calculated,
After having determined the calculated and center coordinates O of the approximate circle C _b the distance d between each plot coordinates A _i of the two-dimensional profile, the difference between the radius R of the approximate circle and the calculated distance calculated d C _b To calculate the depth H in the two-dimensional profile,
If the calculated depth H in the two-dimensional profile is larger than a predetermined threshold, it is determined that a surface flaw candidate Z _k exists in the two-dimensional profile;
When a plurality of two-dimensional profiles determined that the surface flaw candidate Z _k exists are present in the longitudinal direction of the rolled material, and the number of the generated two-dimensional profiles is greater than a predetermined threshold, the rolled material It is determined that surface wrinkles Z are generated in the surface wrinkle detection method. When calculating an approximate circle C _b of the two-dimensional profile,
Of all the plot coordinates A _i indicating the two-dimensional profile, an approximate circle C of the two-dimensional profile is obtained by using the plot coordinates A _(iz) excluding the plot coordinates A _z estimated to be the surface 疵 Z. The surface flaw detection method according to claim 1, wherein _b is calculated by a least square method. In excluding the plot coordinate A _z estimated to be the surface 疵 Z from all the plot coordinates A _i indicating the two-dimensional profile,
For each plot coordinates A _i of the two-dimensional profile, using the data of the plot coordinates adjacent to the plot coordinates A _i A _{i-1, A} i + ₁ for at least three or more plot coordinates A _i, of the two-dimensional profile The diameter L _a of the approximate circle C _{a at} the plot coordinates A _i is calculated by the least square method,
The diameter L _a of the calculated the approximate circle C _a, the absolute value of the difference between the actual diameter L of the rolled material (| L a _-L |) asking,
When the absolute value is larger than a predetermined threshold value, the plot coordinates A _i are excluded from the plot coordinates A _(iz) used when calculating the approximate circle C _b of the two-dimensional profile. The surface flaw detection method according to claim 2. When detecting the surface defect Z of the rolled material,
If the degree of occurrence of the determined plotted coordinates A _i and surface flaws Z in a two-dimensional profile is present, is not less than a predetermined threshold value, according to claim 1, characterized in that determining the two-dimensional profile as defective profile 4. The surface flaw detection method according to any one of 3 above. The rolling material is determined to be a defective product when the degree of occurrence of the two-dimensional profile determined to be the defective profile in a longitudinal direction of the rolled material is equal to or greater than a predetermined threshold value. The method for detecting surface defects as described in 1.   6. A surface wrinkle detection method according to claim 1, wherein the surface wrinkle detection method is performed in real time. A surface flaw detection device for detecting a surface flaw of a long rolled material having a substantially circular cross-sectional shape,
Profile extraction means for extracting a two-dimensional profile of a cross-sectional shape facing a direction perpendicular to the longitudinal direction of the rolled material, with respect to the rolled material conveyed so that the longitudinal direction becomes the conveying direction;
A surface wrinkle detecting device that uses the extracted two-dimensional profile to detect a surface wrinkle of a rolled material by the surface wrinkle detection method according to any one of claims 1 to 6. . The profile extraction means has at least one two-dimensional displacement laser displacement meter arranged along the circumferential direction of the rolled material,
The surface flaw detection device according to claim 7, wherein the two-dimensional displacement laser displacement meter is used to extract a two-dimensional profile of the rolled material.