JP2000046809A - Flaw detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately detect the flaw of a roll without erroniously detecting vibration or foreign matter as a flaw. SOLUTION: Two sensors for detecting a roll profile are arranged in parallel to a flaw detecting direction in close vicinity to each other and, if the difference value between the measured values of two sensors is set to a detection value (y) (steps S1, S2), the detection value (y) becomes a value containing no fluctuation component due to the noise of vibration or the like and corresponding to a fluctuation component due to a flaw or the like. A judge value Δy is calculated on the basis of this detection value and detection values measured at the last time and the time therebefore (step S3) and, when this judge value Δy is larger than a reference value ΔR, a flaw is judged to be probably present (step S4) and a sensor attaching stand 5 is allowed to retreat to perform flaw detection a predetermined number of times Cx with respect to the same place in the same way. When the number N of times wherein the judge values Δy obtained by flaw detection of a predetermined number Cx of times are within a range of an average value La ± Y of them is larger than a predetermined value Nα that is, the same fluctuation is generated in the respective detection values (y), a flaw is judged to be present (steps S8-S17).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a flaw detection method for detecting a roll flaw or the like generated on a rolling roll.

[0002]

2. Description of the Related Art When a roll flaw is generated on the surface of a rolling roll during operation for some reason, the roll flaw is transferred to a steel sheet rolled after the roll flaw is generated. Therefore, a surface flaw is generated in the steel sheet, and this steel sheet becomes a defective product. To avoid this,
It is necessary to find the roll flaws generated on the rolling rolls at an early stage, and a method for detecting the roll flaws online is known. For example, as described in Japanese Patent Application Laid-Open No. 7-260749, a sensor for detecting a roll profile is scanned in the width direction of a rolling roll, and based on continuously detected surface wave echoes, a roll is detected. There is a method of determining the surface shape and determining that a flaw has occurred when a shape change exceeds a preset reference value.

[0003]

However, when the surface shape is measured by a roll profile sensor and the surface shape is approximated by a straight line, if the surface shape changes beyond a certain reference value, a flaw is generated. In the method of determining that there is a flaw, there is a problem that, for example, even when a foreign substance or the like is temporarily attached instead of a flaw, the flaw is erroneously determined as a flaw.

In addition, when the roll profile is measured based on the measured value of the roll profile sensor due to noise such as vibration and the like, the roll surface shape is determined. There is a problem of making a misjudgment.

Therefore, there is a problem that the line is stopped due to these erroneous judgments, and unnecessary roll change is performed, which adversely affects the productivity. Then, this invention was made paying attention to the said conventional problem, and aims at providing the flaw detection method which can detect a roll flaw accurately.

[0006]

In order to achieve the above object, a flaw detection method according to the present invention scans a flaw detection target with a flaw detection device and is concerned that there is a flaw based on a measured value of the flaw detection device. In the flaw detection method in which the corresponding portion of the flaw detection target is scanned a plurality of times when it is performed, and whether or not the flaw is determined based on a plurality of measurement values, the two flaw detection devices are separated by a predetermined distance in the scanning direction. The flaw detection is performed based on the difference between the measured values of the two flaw detection devices.

According to the present invention, the two flaw detectors are separated from each other by a predetermined distance in the scanning direction, that is, the two flaw detectors are arranged so as to have the same movement trajectory. If they are arranged close to each other, the measured values of the two flaw detectors can be regarded as substantially the same. Therefore, when vibration or the like occurs, the respective measured values change simultaneously and similarly, so that the difference value is substantially zero, and the difference value is not affected by the vibration. On the other hand, when there is a change due to a flaw or the like, since the two flaw detection devices are arranged at a predetermined interval, these measured values fluctuate with a time difference, so that a difference due to the flaw or the like appears in the difference value. Therefore, by performing flaw detection based on this difference value, it is possible to avoid making a determination as to whether or not there is a flaw based on a measurement value on which noise such as vibration is present. Will be avoided.

[0008]

Embodiments of the present invention will be described below. FIG. 1 and FIG. 2 are schematic structural views showing a main part of a rolling mill 1 to which a flaw detection method according to the present invention is applied, in which a rolled material L is rolled by a pair of work rolls 1a opposed vertically. And a backup roll 1b for backing up the work roll 1a.

As shown in FIG. 2, the rolling mill 1 is provided with a support frame 2 along the work roll 1a. The support frame 2 supports a sensor mount 5 via a guide rail 3 and a guide piece 4 in a reciprocating manner.
, Thereby reciprocating along the work roll 1a. Then, the drive device 6 moves the sensor mounting base 5 along the work roll 1a at a pitch of 1 mm per rotation of the work roll 1a by an arrangement distance of the sensor sets 7a to 7F.

A sensor set 7a to 7f composed of sensors for detecting a roll profile is attached to the sensor mounting base 5 along the work roll 1a, and each of the sensor sets 7a to 7f faces the surface of the work roll 1a. And are arranged at equal intervals. Each of these sensor sets 7a to 7a
7f is composed of two sensors respectively, the distance between the two sensors 7a ₁ and 7a ₂ constituting each sensor set,
The distance between 7b ₁ and 7b _2, ...... are respectively arranged at predetermined intervals. In addition, the sensor mounting base 5
Each sensor set 7a to 7f for the work roll 1a
Is provided with a position detection sensor 8 composed of a linear encoder or the like for detecting the position of.

Each of the sensors 7n (n = a ₁ , a ₂ ,...)
.., F ₁ , f ₂ ) are constituted by a distance sensor such as a laser sensor for detecting the distance to the surface of the work roll 1 a, and the output signal, that is, the surface wave echo is A / D.
The data is input to the data processing device 10 via the converter 9. The output signal of the position detection sensor 8 is directly input to the data processing device 10.

The data processor 10 detects the rotation position of the work roll 1a from each sensor 7 at the timing of a synchronization signal from a roll rotation synchronization signal generator 11 composed of, for example, a rotary encoder. The presence or absence of a flaw on the surface of the work roll 1a is detected on the basis of the surface wave echo. The roll rotation synchronizing signal generator 11 detects that the work roll 1a is 1 m in the circumferential direction.
A synchronization signal is generated every m rotations.

FIG. 3 is an explanatory diagram for explaining a method of detecting roll flaws in the data processing apparatus 10.
While reciprocating the arrangement interval distance of each sensor of each sensor set 7a~7f shown, each sensor 7a _1, 7a _2, ...
The surface of the work roll 1a is continuously scanned by..., 7f ₁ , 7f ₂ , and a continuous surface wave echo of the work roll 1a is taken into the data processing device 10.

In FIG. 3, R is the surface shape of the work roll 1a obtained by the surface wave echo of the work roll 1a taken into the data processing device 10. On the surface shape R of the work roll 1a, the flaws of the surface projections appear as the R1 portion, and on the surface shape R of the work roll 1a, the flaws of the chipped surface appear as the R2 portion. The tendency of the flaw appearing on the surface shape R of the work roll 1a is verified and grasped in advance. In FIG. 3, ΔR is a reference value of the size (height or depth) of the flaw that requires replacement of the work roll 1a by this verification.

Next, the determination of the size of the flaw will be described with reference to the example of the R1 part which is a flaw of the surface projection. On the surface shape R of the work roll 1a, a flaw R1 of a surface projection (hereinafter, a changing part R)
One. ) Is detected, the change part R1 is included from the approximate value of the height of the surface shape R of the plurality of other sections δ that are continuous to the area δ including the change part R1 separated by the set fixed small distance. The height of the virtual surface r of the work roll 1a in the area δ is determined, and the absolute value | Δy | of the difference between the height of the virtual surface r and the detected height of the changing portion R1 is compared with the reference value ΔR to determine the size of the flaw. Is determined.

[0016] For example, when the detection value of the sensor suite 7a to be detected based on the sensor 7a ₁ and 7a ₂ and y, the height of the surface shape R of the work rolls 1a includes a detection value ya in the change unit R1, the sensor mounting frame The detected values ya ₁ and ya ₂ of two sections δ adjacent to the changing section R1 behind the changing section R1 with respect to the moving direction 16 of FIG. 5 (areas adjacent to the changing section R1 on the right side in FIG. 3). , Changing part R
Height ya ₃ virtual surface r of area δ containing 1 is calculated by the following equation (1). The difference ± Δy between the height of the virtual surface r and the detected height of the changing portion R1 is calculated by the following equation (2).

[0017] ya_Three= (Ya₁+ Ya_Two) / 2 (1) ± Δy = ya−ya_Three ... (2) the change portion R1 in which the height ± Δy obtained here should be determined
(Flaw) height. Hereinafter, the height ± Δy is referred to as a judgment value.
You. In FIG. 3, reference numeral 15 denotes a sensor 7n (n = a ₁,
a_Two, ……, f₁, F_Two) Measurement reference point, 16 is a sensor
The direction of movement of the mounting base 5, 17 is the measurement signal of the sensor 7n.
Reference numeral 18 indicates measurement points for each section δ.

The absolute value | Δy of this determination value ± Δy
Is larger than the reference value ΔR, that is, | Δy |> Δ
When it is R, it is determined that a flaw may have occurred.

The detection value y of each of the sensor sets 7a to 7f is as follows.
It is calculated as follows. That is, the sensors forming the sensor set
Sa, for example 7a₁And 7a_TwoThe measured values for
Each S₁And S_TwoVibration
ΔS ₁And ΔS_TwoAnd This
Each sensor 7a at the time of₁And 7a_TwoAnd the difference between the measurements
Then, the difference value ΔS is expressed by the following equation (3).

ΔS = (S₁+ ΔS₁)-(S_Two+ ΔS_Two) (3) Sensor 7a forming sensor set₁And 7a_TwoTo some extent
When placed near, S ₁≒ S_Two, ΔS₁≒ ΔS_TwoAnd consider
Therefore, the difference value is ΔS ≒ 0. But
Therefore, the difference value ΔS includes a variation due to vibration or the like.
No flaw can be detected based on this difference value ΔS
If flaws are detected without being affected by vibration, etc.
Becomes possible. FIG. 4 shows the time on the horizontal axis and the sensor 7a on the vertical axis.
₁(FIG. 4 (a)) and 7a_Two(Figure 4
(B)). Time t ₁Causes vibration
Then, as shown in FIG. 4A and FIG.
The movement will take place, but as shown in FIG.
Sensor 7a₁And 7a_TwoBy taking the difference between the measured values of
The value makes it possible to eliminate the effects of vibration. And
As shown by the broken line in FIG.
Sensors are arranged at predetermined intervals,
The measured value does not fluctuate at the time point, and the difference value is shown in FIG.
(C) As shown in FIG.
Will be.

Therefore, the two sensors constituting each sensor set can regard the difference between the measured values as substantially zero when there is no roll flaw or the like. The sensors are arranged so that the variation due to the flaw appears in the difference between the measured values of the two sensors.

FIG. 5 is a flowchart showing an example of the roll flaw detection processing in the data processing apparatus 10. Note that the processing is actually performed for each of the sensor sets 7a to 7f, but the sensor set 7m will be representatively described here.

[0023] In the data processing apparatus 10, first, at step S1, to the roll surface from the sensor 7m ₁ and 7m ₂ via the A / D converter 9 at the timing of the synchronizing signal from the roll rotation synchronizing signal generator 11 The measured value obtained by detecting the distance is read, and the position detection signal from the position detection sensor 8 is read. Next, the process proceeds to step S2, in which a difference value between two measured values of each sensor set 7m is calculated, and this is set as a detected value y. At this time, the previous detection value and the detection value before the previous detection are stored.

Next, the process proceeds to step S3, where (1)
And based on the equation (2), the detection value y detected in step S2 ya, the detection value of the previous stored in advance and the detection value of the before-last as ya _1, ya ₂ respectively, and calculates a determination value [Delta] y.

Next, the process proceeds to step S4 to determine whether the absolute value | Δy | of the determination value Δy is larger than the reference value ΔR. And when | Δy |> ΔR is not satisfied,
In step S5, when the determination value Δy is | Δy | = 0, it is determined that there is no flaw or chipping on the roll surface, and the process returns to step S1. On the other hand, if the determination value Δy is not | Δy | = 0, step S6
When Δy is a negative value, the degree is small, but it is determined that there is foreign matter seizure. When Δy is a positive value, it is determined that the degree is small but there is a surface chipping flaw, and the operator is notified. After performing the processing, the process returns to step S1.

On the other hand, in step S4, when the absolute value of the determination value Δy is | Δy |> ΔR, it is determined that a flaw exists, and the determination value Δy is stored as Δy ₁ in a predetermined storage area. The process proceeds to S8, where the count variable C is updated to C = C + 1. It is assumed that this count variable is set to C = 0 in the initial state.

Next, the process proceeds to step S9, in which the driving device 6 is moved in the reverse direction based on the position detection signal of the position detection sensor 8 when the measured value is read in step S1, and the above-described determination value Δy After the sensor mounting base 5 is returned to a position obtained by adding a margin to the position where the
The sensor mounting base 5 is moved toward 6.

Next, the process proceeds to step S10, and similarly to step S1, the measured values of the sensors 7m ₁ and 7m ₂ are read in synchronization with the synchronization signal from the roll rotation synchronization signal generator 11, and the position detection sensor 8 is read. Is read, and then in step S11, each sensor set 7m
The difference value ΔS between the measured values is calculated, and this is set as a detected value y.

[0029] Then the process proceeds to step S12, wherein (1) and (2) to calculate a decision value [Delta] y on the basis of the expression, which was stored as [Delta] y ₂ in a predetermined storage area, the process proceeds to step S13. In this step S13, it is determined whether or not the count variable C has become C = Cx.
If it is not x, the process returns to step S8, and step S13
If C = Cx, the process proceeds to step S14.

In step S14, the average value L of Cx determination values Δy obtained by scanning the same location based on the determination values Δy _{1 to} Δy _Cx stored in the predetermined storage area is used.
a is calculated according to the following equation (4).

[0031] Next, the process proceeds to step S15, where each of the determination values Δy _{1 to} Δy ₁
It is determined whether or not y _Cx is within the range of La ± Y, and the number N of the determination values within the range is counted.
Then, the process proceeds to step S16 to determine whether or not the number N is larger than a preset reference value Nα.
If it is not α, it is determined that it is not a flaw, and the process returns to step S1. On the other hand, when the number N is larger than the reference value Nα, it is determined that there is a flaw, and the process proceeds to step S17. For example, predetermined processing such as notifying an operator that the work roll 1a needs to be reconfigured is performed. Step S
Return to 1. The reference value Nα is, for example, the number of measurements C
A value of about 70 to 80% of x is set.

Next, the operation of the above embodiment will be described.
The sensor mounting base 5 is moved once per rotation of the work roll 1a.
move along the work roll 1a at a rate of mm pitch,
The measurement value of each sensor 7n is read at the timing of the synchronization signal from the roll rotation synchronization signal generator 11, and the detection of the roll flaw is performed. At this time, if there is no roll flaw on the work roll 1a, the detection value y calculated in step S2 becomes substantially zero, so that the determination value Δy calculated in step S3 also becomes substantially zero, and there is no roll flaw. Then, the detection of the roll flaw is performed in the same manner. And
When the sensor mounting base 5 moves by a predetermined distance corresponding to the interval between the sensor sets 7a to 7f, the sensor sets 7a to 7f
The sensor 7n detects roll flaws on the entire surface of the work roll 1a, and the driving device 6 drives the sensor mounting base 5 in the reverse direction to detect roll flaws in the reverse direction. Is performed. By repeating this, while the sensor mounting base 5 reciprocates along the work roll 1a, roll flaws are detected on the entire surface of the work roll 1a.

In this state, if foreign matter or the like adheres to the position K of the work roll 1a, for example, this is detected by a sensor set corresponding to the area δ including the position K, for example, 7b, and FIG. since the detected value y ₁ of the first-time sensor set 7b as shown fluctuates, the absolute value of the determination value | [Delta] y | increases, the determination value | is greater than the reference value [Delta] R, | [Delta] y
The determination value | [Delta] y | is stored in a predetermined storage area as the [Delta] y _1.

Then, the sensor mounting base 5 is moved in the opposite direction by the driving device 6 and returned to a position earlier than the position K obtained by adding a margin to the position K based on the position detection signal from the position detection sensor 8. Then, the sensor mounting base 5 is moved again at a predetermined speed. As a result, each sensor 7n performs the roll flaw detection again at the same position on the roll surface (step S9).

At this time, when the adhered foreign matter or the like is removed, when the roll flaw detection is performed again, the second detection value y ₂ calculated by the sensor set 7b is as shown in FIG. substantially zero next, as shown in, the detection value y is from substantially equal to zero, also the determination value [Delta] y calculated in step S3 becomes substantially zero which is stored as the determination value [Delta] y ₂ in a predetermined storage area. Then, the sensor mounting base 5 is again moved backward to the position K.
Back to near, when similarly, because the third-time detection value y ₃ calculated by the sensor suite 7b is substantially equal to zero as shown in FIG. 6 (c), similarly to the above judgment value Δy also substantially becomes zero storing as the determination value [Delta] y _3. This process is repeated, and when the process is performed a predetermined number of times Cx, the process shifts from step S13 to step S14 to store the stored determination values Δy ₁ .
The average value La is determined based on Δy _Cx , and each of the determination values Δy _{1 to} Δy
The number N within the range of La ± Y is determined for y _Cx (steps S14 and S15).

At this time, as shown in FIG. 6, the detection value y of the sensor set 7b with respect to the position K becomes substantially zero since the adhering matter has been removed from the second time onward, so that the determination value Δy is It becomes almost zero. Therefore, the average value La is a value close to zero, and the number N in which the determination values Δy _{1 to} Δy _Cx are within the range of La ± Y is smaller than the reference value Nα, and is determined to be not a flaw.

On the other hand, a flaw is generated at the position K.
In this case, the first detection value y₁(Fig. 7 (a)), 2nd time
Detection value y_Two(FIG. 7B), the third detection value y_Three(Figure
As shown in FIG. 7 (c)), the detected value y is at the position K.
The determination value Δy for the predetermined number of times Cx₁~
Δy_CxAre almost the same value, and these determination values Δy ₁
~ Δy_CxIs within the range of La ± Y.
Since the value exceeds the value Nα, it is determined that the surface is a flaw.

Further, for example, in a state where the roll flaw is not generated, when some vibration is generated and this vibration is measured by each sensor 7n, as shown in FIG. At the same time, the two sensors have a vibration at the same time point. Therefore, if these differences are calculated, the difference value does not include the influence of the vibration. In contrast,
In the case of a flaw, since the two sensors of each sensor group are arranged at a predetermined interval, each measured value fluctuates with a time lag, and as shown by a broken line in FIG. Fluctuations appear. Therefore, by taking a difference value between the two sensors for each of the sensor sets 7a to 7f and using this as a detection value y, the detection value y becomes a detection value that does not include the influence of noise such as vibration. Erroneous detection such as detection as a flaw can be avoided.

Therefore, as described above, once a change suspected to be a flaw appears in the detection value y, roll flaw detection is performed again for the same location a plurality of times,
When the same measurement value is detected in a plurality of detections, it is determined to be a flaw. Therefore, it is possible to avoid erroneous detection such as determining an attached matter such as a foreign substance as a flaw, and to stably detect a roll flaw. It can be performed.

Each of the sensor sets 7a to 7f is composed of two sensors, and roll flaws are detected based on the difference between the measured values of these two sensors. Can be prevented from being erroneously detected as a roll flaw.

Thus, erroneous detection can be avoided and stable detection of roll flaws can be performed, so that line stoppage or unnecessary roll change based on erroneous detection can be avoided. The drop can be prevented.

In the above embodiment, the sensor mounting base 5 is moved by 1 m per rotation of the work roll 1a.
m pitch, and work roll 1
Although the case where the roll flaw is detected every 1 mm of the rotation of a has been described, the present invention is not limited to this.
What is necessary is just to set it according to measurement conditions, such as a rotation situation or operation environment of a. Similarly, the arrangement position of the sensor set,
The number and the like can be arbitrarily set according to the size of the work roll 1a and the like.

In the above embodiment, the case where a laser sensor is used as the sensor 7 has been described. However, the present invention is not limited to this, and any sensor that can measure the distance to the roll surface can be applied. . Similarly, the case where a linear encoder is applied as the position detection sensor 8 has been described.
The roll rotation synchronization signal generator 11 is not limited to the rotary encoder, and may be a sensor capable of detecting the rotation state of the work roll 1a. If so, it can be applied.

In the above-described embodiment, the case where the roll flaw of the work roll 1a is detected has been described. However, the present invention is not limited to this, and the flaw detection method in which the surface flaw is detected using the distance sensor is used. If so, it can be applied.

[0045]

As described above, according to the flaw detection method of the present invention, two flaw detection devices are arranged at predetermined intervals in the scanning direction, and flaw detection is performed based on the difference between the two measured values. Accordingly, it is possible to avoid determining whether or not there is a flaw based on a measurement value with noise such as vibration, and to avoid erroneously determining vibration or the like as a flaw.

[Brief description of the drawings]

FIG. 1 is a schematic configuration diagram of a main part of a rolling mill to which a flaw detection method according to the present invention is applied.

FIG. 2 is a schematic configuration diagram of a main part of a rolling mill to which the flaw detection method according to the present invention is applied.

FIG. 3 is an explanatory diagram illustrating a principle of detecting a roll flaw.

FIG. 4 is an explanatory diagram illustrating a principle that a difference value between measurement values of two sensors does not include a variation due to noise.

FIG. 5 is a flowchart illustrating an example of a processing procedure of roll flaw detection processing.

FIG. 6 is an explanatory diagram for explaining the operation of the present invention;

FIG. 7 is an explanatory diagram for explaining the operation of the present invention;

[Explanation of symbols]

 Reference Signs List 1a work roll 2 support frame 3 guide rail 5 sensor mounting base 6 driving device 7a to 7f sensor set 8 position detecting device 10 data processing device 11 roll rotation synchronization signal generator

Claims (1)

[Claims] 1. A flaw detection device scans a flaw detection target object,
In the flaw detection method in which the applicable portion of the flaw detection target is scanned a plurality of times when there is a concern that there is a flaw based on the measurement value of the flaw detection apparatus, and it is determined whether the flaw is a flaw based on the plurality of measurement values. A flaw detection method comprising: disposing two flaw detection devices separated by a predetermined distance in a scanning direction, and performing flaw detection based on a difference value between measured values of the two flaw detection devices.