JP3286549B2 - Surface flaw detection method for long steel materials - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a method for assuring the quality of a final product such as a refinement line of a rolling process and a problem such as breakage due to a flaw in a rolling line for a long steel material such as a strip steel plate or an H-shaped steel material being conveyed. The present invention relates to a method for detecting surface flaws of a long steel material, which detects surface flaws of a long steel material while being conveyed along a pass line, in order to prevent the occurrence of the flaw.

[0002]

2. Description of the Related Art As a conventional method for measuring the surface flaw of a long steel material such as a strip-shaped steel sheet being conveyed, for example, Japanese Unexamined Patent Publication No.
There is a method described in JP-A-53-122448.

[0003] In this method, each distance from a reference position is detected using a distance meter arranged opposite to the surface of a steel material along a width direction orthogonal to the pass line, and a distance change amount at each observation point is determined. This is to determine the presence or absence of a flaw.

That is, with respect to a plurality of distance detectors arranged in the width direction, a distance variation is detected by a difference signal from an adjacent distance meter, and a steel material is detected from a difference signal between the distance meters located at both ends in the width direction. By detecting a signal ignoring the average inclination between the surface and the alignment surface of each of the rangefinders, that is, the partial irregularity between the two end detection elements, and comparing the detected signal with the distance variation signal, the steel material being conveyed. Irrespective of the inclination of the detector and the surface of the steel material caused by the vibration, only the irregular shape can be accurately detected. For example, a difference signal d1 (n / 2) between two adjacent distance meters from a distance signal of each distance meter (n) fixed to a predetermined alignment surface facing in parallel with the surface of the H-shaped steel material and both ends are provided. A differential amplifier for extracting a differential signal d2 (one signal) between the located distance meters, d
It comprises a comparator for comparing 1 and d2, respectively, and an OR circuit for detecting these values via a detector to detect flaws.

[0005]

The causes of noise in the method of detecting a flaw from a change in distance include, for example, the following.

In other words, in order to suppress fluctuations in the pass line of the steel sheet due to noise equipment caused by a change in distance due to a change in the shape of the steel material (ear extension, belly extension, warpage, etc.) due to the measurement object, a roll is required. This is noise due to the eccentric component of the roll when measuring at the place where the steel sheet is wound.

In the conventional method for detecting surface flaws as described above, stable measurement can be performed if the vibration accompanying the conveyance or the shape of the steel sheet is uniform in the width direction. However, the warp and elongation of the steel material in the width direction can be measured. In some cases, the difference signal between the rangefinders located at both ends,
There is a problem that a difference is generated between a difference signal between two adjacent rangefinders and that the range signal may be erroneously determined as a flaw. The output of the flaw is from the OR circuit, and the size of the flaw cannot be measured.

Further, a change in distance (difference) along the longitudinal direction.
If an attempt is made to detect a flaw only by using a flaw, there is a possibility that a change in distance due to partial ear extension or the like is erroneously determined as a flaw. The present invention
It has been made in view of such problems, and reduces a noise component caused by a shape of a measurement target and a noise component caused by equipment, and can accurately detect a surface flaw of a long steel material. It is an object to provide a detection method.

[0009]

In order to solve the above-mentioned problems, a method for detecting surface flaws of a long steel material according to the present invention is a method for detecting surface flaws of a long steel material conveyed along a predetermined pass line. Using a plurality of distance meters arranged opposite to the surface of the steel material along the width direction orthogonal to the pass line and the longitudinal direction along the pass line, at a predetermined timing, at a predetermined interval in the width direction of the steel material. A plurality of distances from the reference position at a plurality of observation points arranged in a row are obtained at predetermined intervals in the longitudinal direction of the steel material, and each data value is obtained at each observation point, and each data value is determined in the width direction. After converting to a difference value with the data values at other observation points arranged at intervals, the difference between the converted data value and the data value separated by a predetermined interval along the longitudinal direction is obtained, and the plurality of differences are calculated. Based on data, steel surface It is characterized in that to determine whether there is allowed more defects.

[0010] In the shape change occurring in the long steel material, that is, in the warp or elongation of the long steel material, if there is no flaw, the distance changes smoothly in the width direction. Based on this, first, a data value consisting of a distance value from a reference position at an observation point arranged in the width direction is a relative value called a difference value from a data value consisting of a distance value at another observation point arranged in the width direction. By converting the data value into a value, noise corresponding to a change due to a shape change such as warpage or elongation occurring in the steel material is suppressed.

The noise caused by the equipment includes a change in distance due to, for example, eccentricity of the roll and poor roll shape. Although this occurs as undulations along the longitudinal direction, it changes relatively smoothly in the width direction as in the case of the plate shape (spread), so the undulation component is suppressed by taking a difference in the width direction in advance. .

Thereafter, by taking the difference value along the longitudinal direction of the data value at each observation point converted into the difference value in the width direction, noise components caused by equipment such as a change in steel shape and roll eccentricity are suppressed. The obtained flaw information is obtained.

As described above, according to the present invention, a flaw on the surface of a steel material is detected by making a determination based on the difference data in the longitudinal direction in which the noise component caused by the change in the shape of the steel material and the equipment is suppressed. .

Further, the following description will be made specifically focusing on adjacent channels 5 and 6 as examples of observation points. If there is an ear extension or the like, a waveform passing through the data in the longitudinal direction of each distance data in the channels 5 and 6 is represented, for example, as shown in FIG. That is, there is a distance change in the width direction between the channels 5 and 6 due to a shape change such as ear growth.
In addition, the undulation in the longitudinal direction is caused by a change in distance such as eccentricity of the roll and poor roll shape.

Then, in order to obtain a difference between each observation point and another observation point in the width direction, that is, a difference in the same cross section of the steel material, each data value in the channel 6 is calculated based on the following equation. The difference from each data value lined up in FIG.
The waveform is as shown in FIG.

Width difference value △ 1 (j) = s (j + y) -s (j) s (j): Measurement value of the same steel plate cross section of each distance meter j = 1 to n, y = 1 to k As can be seen from the waveform after the difference in the width direction in FIG. 5, when the distance data value gradually changes in the width direction due to warpage or elongation of the steel sheet, the change (undulation) can be suppressed and the relatively flat data can be suppressed. Becomes

Thereafter, when the difference between each observation point, in this example, the data value at a predetermined interval in the channel 6 arranged in the longitudinal direction is obtained based on the following equation, a waveform as shown in FIG. 6 is obtained.

The longitudinal difference value △ 2 (i) = △ 1 (i +
X)-△ 1 (i) i = 1 to P This final difference value is small and stable as a whole.
FIG. 7 shows an example in which the width direction difference processing is not performed for comparison, that is, only the longitudinal direction difference is performed.

As described above, a change in distance due to the shape of the steel sheet is not detected as a flaw signal, and more accurate detection is possible.
Further, noise components due to equipment such as roll eccentricity are also reduced, and the flaw discrimination threshold can be set small.

Next, according to a second aspect of the present invention, in the configuration according to the first aspect, a difference between the data values along the longitudinal direction at each observation point is compared with a predetermined threshold value. The above difference data is extracted, and the number and size of flaws are obtained based on the extracted difference data.

The absolute value of the extracted difference data equal to or greater than the threshold value is proportional to the depth of the flaw. If the difference data is continuous in the longitudinal direction or the width direction, the number of continuous data is It is proportional to length and width.

Therefore, the number and size of flaws are detected based on the extracted difference data.

[0023]

Next, embodiments of the present invention will be described with reference to the drawings. As shown in FIG. 1, the inspection target of the flaw, that is, the strip-shaped steel plate 1 which is a long steel material is conveyed along a predetermined pass line P, sequentially wound around a pair of rolls 2, turned, and then sent to the next process. In the state of being wound around the pair of rolls 2, a flaw inspection of the front surface and the back surface of the strip-shaped steel sheet 1 is performed.

Here, the reason why the flaw detection is performed when the sheet is wound around the roll 2 is to suppress noise at the time of flaw detection due to vibration or the like during conveyance. Since the inspection of the front and back surfaces of the steel strip 1 is the same, the inspection of the front surface will be described below.

A plurality of eddy current displacement meters 3 constituting a distance meter are provided along a reference surface (distance meter alignment surface) forming a reference position at a predetermined interval along the surface of the roll 2. It is installed facing the surface of the roll 2 (see FIG. 1).
In the figure, a part of the displacement meter 3 is schematically illustrated by ○).

The eddy current displacement gauges 3 are arranged in a zigzag pattern at intervals of 10 mm in the width direction orthogonal to the pass line P and at intervals of 75 mm in the longitudinal direction along the pass line P. Each eddy current displacement meter 3 measures the distance from the reference position to the surface of the steel strip 1 and supplies the measured value to the controller 6 via the A / D converter 5. Note that the displacement meters 3 are arranged in a staggered manner in order to avoid vortex interference of the displacement meters 3 and to dispose the displacement meters 3 as densely as possible.

A rotation speed detector 4 such as a pulse generator is connected to the rotation shaft of the roll 2, and the rotation speed detector 4 generates a signal corresponding to the rotation of the roll 2, that is, a signal corresponding to the conveying speed of the steel strip 1. Can be supplied to the controller 6.

As shown in FIG. 2, the controller 6 includes a distance data input unit 7, a width direction difference processing unit 8, a longitudinal direction difference processing unit 9, and a flaw determination unit 10. In the distance data input means 7, observation points are set in advance at a predetermined pitch along the width direction of the steel strip 1, for example, observation points of 10 channels are determined in advance in the width direction, and based on a signal from the rotation speed detector 4. By inputting a distance signal from the displacement meter 3, distance data from a reference position at each observation point at intervals of 5 mm in the longitudinal direction of the steel strip 1 is obtained and stored in a memory. This is performed in units of 500 mm in the longitudinal direction of the steel strip 1. At this time, the displacement of the longitudinal direction due to the arrangement of the displacement meter 3 is corrected.

Thus, for one flaw detection, distance data of 10 points in the width direction and 10 points in the length direction, that is, 100 points in total, are stored. In the width direction difference processing means 8,
Based on the distance data at each observation point obtained by the distance data input means 7, the distance data of the observation points located at the end in the width direction of the steel strip 1 is sequentially read from the distance data of the adjacent observation points arranged in the width direction. , The data value of each observation point is converted to the difference value and stored in the memory.

In the longitudinal difference processing means 9, on the basis of the data value of each observation point converted by the width difference processing means 8, each observation point is separated by two data, that is, 10 mm apart in the longitudinal direction. Difference from the data value at the position
Difference data at each observation point in the longitudinal direction in units of 5 mm is obtained and stored in a memory. In the above description, the difference between the data value two data before and the difference is described. However, the difference between the data value three or more data before may be used, or n data after (n ≧ n)
The difference from the data value of 1) may be used.

Thus, difference data is obtained in a matrix at intervals of 5 mm in the longitudinal direction and at intervals of 10 mm in the width direction. In addition, the flaw determining means 10 determines the presence or absence of a flaw more than allowable based on the difference data obtained by the longitudinal difference processing means 9.

For example, each difference data is compared with a predetermined threshold value α, that is, quantization is performed, and the number of flaws of a predetermined size or more or a product defect is determined based on the quantized difference data. The presence or number of extremely large flaws is determined.

That is, the current detection target range (for 500 mm)
, The number of flaws can be determined from the number of difference data equal to or larger than the threshold value. If the difference data equal to or more than the threshold value is continuous in the length direction, the length of the flaw is detected by the difference data number × the measurement pitch of 5 mm, and the difference data equal to or more than the threshold value is continuous in the width direction. If so, the number of difference data ×
The width of the flaw is detected based on the measurement pitch of 10 mm, and the depth of the flaw is detected from the absolute value of the difference data.

The length, width, and depth of each detected flaw are comprehensively determined, and flaw information on the number and size of flaws is output. And in this flaw detection device, steel strip 1 longitudinal direction 5
The flaw determination information is output in units of 00 mm. If this flaw detection is a pickling step or the like before rolling, the obtained flaw information is supplied to the controller of the rolling mill and used as feedforward information for rolling control. If the flaw detection is a refinement line, it is used for quality assurance of the final product.

In the above description, the detection of flaws on the front surface of the steel strip 1 has been described. In the flaw detection method using the flaw detection device as described above, first, the distance data, which is raw data, is converted into a width direction difference value, which is an amount of change in distance per unit in the width direction, so that the ear of the steel strip 1 is converted. Noise components due to changes in the shape of the steel strip 1 such as elongation and antinode elongation are suppressed.

Further, by converting the data into longitudinal difference data for each observation point, noise components based on equipment such as the eccentricity of the roll 2 are removed, and the threshold value for flaw discrimination is reduced. This makes it possible to more accurately detect flaws and to detect small flaws.

The distance meter is not limited to the eddy current displacement meter 3, and other known distance meters such as a laser distance meter may be used as long as the distance can be measured. Further, the detection intervals in the width direction and the longitudinal direction are not limited to the above intervals. Further, the pitch between the observation points in the width direction does not necessarily need to be the same.

[0038]

FIG. 3 shows the result of a simulation performed to confirm the effect of the flaw detection method according to the present invention.

The simulation was carried out based on the measured values of the eddy current displacement meter 3 arranged in a staggered manner. In this case, the processing is performed on data in which the edge of the steel sheet 1 has elongation. The result shows only the edge part.

Specifications of the simulation The moving speed of the steel sheet 1: 1.0 m / s Number of measuring points: 100 points in the longitudinal direction / 5 mm pitch Number of sensors: 140 (pitch in the width direction 10 mm) Width difference in the width direction: 1 point (5 mm) (Interval) Longitudinal difference interval: 2 points (10 mm interval) Flaw discrimination threshold: 0.15 mm For comparison, the result of a comparative example in which only the longitudinal difference processing was performed without performing the width difference processing is also described ( D, E in FIG. 3).

Here, in FIG. 3, the L direction is a longitudinal direction,
The C direction means the width direction. The horizontal axis of the graph is the sensor number, and the vertical axis is the magnitude of the difference value. A and B are lines connecting the maximum value and the minimum value of the difference values of the detection values based on the present invention, and D and E are lines connecting the maximum values and the minimum values of the difference values according to the comparative example. is there.

From FIG. 3, it can be seen that, in the ear extension portion, in Comparative Examples D and E, the difference value is large, and it is determined as a flaw. Therefore, in order to avoid this, the threshold value must be increased. Conversely, in the case of the present invention, since the increase in the difference value due to the shape change based on the ear growth is suppressed, the shape change based on the ear growth is not determined as a flaw.
This indicates that smaller flaws can be detected than before.

[0043]

As described above, the method for detecting surface flaws of a long steel material according to the present invention suppresses the adverse effects of the shape of the long steel material such as the ear extension and the adverse effects of the equipment for transporting the steel material, thereby reducing the flaws. There is an effect that detection can be performed stably and accurately.

In addition, since the influence of the shape of the steel material is suppressed, the size of the detectable flaw can be set small. At this time, if the invention described in claim 2 is adopted,
There is an effect that information on a flaw can be specifically obtained.

[Brief description of the drawings]

FIG. 1 is a diagram showing an outline of a device configuration according to an embodiment of the present invention.

FIG. 2 is a diagram showing a configuration of a controller according to the embodiment of the present invention.

FIG. 3 is a diagram showing an embodiment.

FIG. 4 is a diagram showing a distance data waveform for channels 5 and 6;

FIG. 5 is a diagram showing a waveform after taking a difference in a width direction.

FIG. 6 is a diagram showing waveforms after a width direction difference and a longitudinal direction difference are obtained.

FIG. 7 is a diagram showing a waveform of a comparative example.

[Explanation of symbols]

 P pass line 1 steel strip 2 roll 3 distance meter 4 rotation detector 5 A / D converter 6 controller 7 data input means 8 width direction difference processing means 9 longitudinal direction difference processing means 10 flaw determination means

Continuation of the front page (56) References JP-A-53-122448 (JP, A) JP-A-6-74755 (JP, A) JP-A-4-50608 (JP, A) JP-A-4-363045 (JP) JP-A-62-257006 (JP, A) JP-A-5-346325 (JP, A) JP-A-6-34362 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB G01B 21/30 G01N 27/90

Claims (2)

(57) [Claims] 1. A method for detecting a surface flaw of a long steel material conveyed along a predetermined pass line, wherein the steel material is disposed along a width direction orthogonal to the pass line and a longitudinal direction along the pass line. Using a plurality of distance meters arranged opposite to the surface, at predetermined timing, each distance from the reference position at a plurality of observation points arranged at predetermined intervals in the steel material width direction at predetermined intervals in the steel material longitudinal direction After obtaining a plurality of data values at each observation point, and converting each data value to a difference value from the data values at other observation points arranged at predetermined intervals in the width direction, Regarding the data value, a difference between the data value and a data value separated by a predetermined interval along the longitudinal direction is obtained, and based on the plurality of difference data, it is determined whether or not the steel material surface has an unacceptable flaw or more. Surface flaw detection method. 2. The method according to claim 1, wherein a difference between the data values along the longitudinal direction at each observation point is compared with a predetermined threshold to extract difference data equal to or larger than a predetermined threshold, and the number and size of flaws are determined based on the extracted difference data. 2. The method for detecting surface flaws of a long steel material according to claim 1, wherein