JP2003247951A - Defect detection method and device for strip-shaped material - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reliably detect a defect detection signal changing in short times in a high speed line via a low-cost device by reducing a load on data processing. <P>SOLUTION: For detection of a crack as a defect in a running strip-shaped material 12, a signal from a sensor 22 is peak-held over a given length Lh, and is sampled at pitches Lp shorter than the peak hold length Lh. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a belt-shaped material defect detection method and apparatus, and more particularly to a high-speed line for a short time, which is suitable for use in detecting an edge crack which is a defect occurring in a steel plate edge. TECHNICAL FIELD The present invention relates to a belt-shaped material defect detection method and device capable of reliably detecting a defect from a defect detection signal that changes to the above.

[0002]

2. Description of the Related Art When a crack occurs at the edge of a steel plate,
The edge cracked portion may be caught by the equipment when it is threaded, and it may be broken to reduce the production capacity. As a method of avoiding this, by detecting edge cracks in a rolling process or the like where edge cracks are likely to occur and performing processing such as cutting the edge cracked portion, the edge cracks may become a factor in the process after the rolling process. A possible method is to avoid the occurrence of such troubles.

As a method for detecting such edge cracks in a steel sheet, the applicant has disclosed in Japanese Patent Laid-Open No. 2000-230924 that a plurality of eddy current sensors having an E-shaped core are arranged and the relationship between the area and depth of cracks and the detection signal is used. It proposes to detect only cracks and determine the depth and width of cracks.

[0004]

However, in this technique, it is necessary to fetch the signal from the sensor into the crack determination device and process it. Especially, the line speed is fast (2
(0 m / s or more) When applied to a rolling line, if the width of the crack is small, in order to reliably capture the signal from the sensor,
Very fast sampling is required.

For example, when the line speed is 25 m / s and the crack width is 3 mm, the signal change time is 0.12 ms.
Becomes Therefore, to catch the peak,
It is necessary to perform sampling at about 0.012 ms of / 10 or less.

In addition to this, in order to carry out a plurality of processes in parallel, such as a crack determination, a higher speed process is required.

In particular, in the case of crack detection on a high-speed line such as the exit side of a rolling mill, the change in signal is very short and the load of data processing becomes large.

Therefore, not only the apparatus becomes expensive, but also sampling may not be performed in time, so that a so-called sampling omission may occur in which a peak falls between sampling points.

The present invention has been made to solve the above-mentioned conventional problems, and it is possible to reduce the load of data processing and to reliably detect a crack detection signal that changes in a short time on a high-speed line. It is an issue.

[0010]

According to the present invention, when detecting a defect in a running strip, a signal from a sensor is peak-held for a predetermined length and sampling is performed at a pitch shorter than the length of the peak-hold. In this way, the above problems are solved.

Further, the length of the peak hold is set to 1.5 times the sampling pitch, so that defect positions can be specified at intervals of half the sampling cycle.

The present invention also relates to a belt-shaped material defect detecting device for detecting a defect of a belt-shaped material during traveling, and means for peak-holding a signal from the sensor for a predetermined length, and the peak-holding length. The problem is solved by providing a means for sampling at a short pitch.

In the present invention, as shown in FIG. 1, the output signal A from the sensor is first peak-held by a peak-hold circuit for a predetermined length (referred to as peak-hold length) Lh. Specifically, (1) When the output signal A of the sensor exceeds the set threshold value Vt, peak hold is started, and this is set as the output signal Vp (= A) of the peak hold circuit (FIG. 1, left).

(2) Current peak value Vp during peak hold
When a signal A2 larger than 1 (= A1) is generated, V
Hold p2 (= A2) as a new peak value (FIG. 1, right).

(3) The counting of the hold length is started from the time when the peak value changes, and after the peak hold length is held, the hold length is reset.

(4) It is necessary to determine the peak hold length Lh so as to satisfy the positional accuracy required from the workability in the operation when confirming the position of the crack in the subsequent process. Then, it is set longer than the sampling pitch Lp described below,
It is preferably 1.5 times.

Subsequently, the output signal of the peak hold circuit is sampled at a constant pitch by the sampling circuit.
The output signal B from the peak hold circuit is sampled at a constant pitch Lp asynchronously with the peak hold circuit. The sampling pitch Lp is determined based on the processing speed of the data processing device.

For example, when the confirmation of the post-process is performed within the range of 20 m in length by reducing the line speed, the crack position is
For example, it is sufficient if the determination can be made in units of 1/20 of 1/20. If the sampling pitch Lp based on the processing speed of the data processing device is 0.1 m, the condition of the present invention can be satisfied by making the peak hold length Lh longer than this sampling pitch Lp.

For example, the sampling pitch Lp = 0.1
In the case of m, the peak hold length Lh can be set to 0.15 m. When the line speed is 1500 mpm, the sampling period of the data processing device is 4 ms, which is a speed that can be sufficiently processed even by using a general-purpose personal computer.

On the other hand, the peak length Ls shown in FIG.
When the sensor output signal A of 1 is directly sampled, it is impossible to reliably detect the peak point at the sampling pitch Lp.

Further, it is possible to determine the position of the defect based on the sampling signal during the peak hold. Here, a case where the peak hold length Lh is 1.5 times the sampling pitch Lp will be described.

The position of the defect is determined based on the sampling number Ns of the sampling signal Sn existing in one peak hold. Here, the sampling number Ns is the sampling number of the sampling signal Sn satisfying the following conditions.

Sampling signal Sn exceeds threshold value Vt (Sn = Vp> Vt) Difference ΔS = Sn-Sn- from the previous signal of sampling signal Sn
The value of 1 is positive and changes to negative. It is the maximum within that range. The number of sampling times Ns of the sampling signal Sn satisfying this condition is 1 or 2 times. The position of the defect is determined.

(A) When the number of sampling times Ns = 1 (FIG. 2 (a)) From the relationship between the peak hold length Lh and the sampling pitch Lp, when the number of sampling times Ns = 1,
The sampling position at which the sampling signal Sn is maximum is Pn, and the intermediate position between the previous sampling positions Pn-1 and Pn is Pn.
′, The defect position is in the range of Pn−1 to Pn ′ (length Lp
/ 2) is determined.

(B) When the number of sampling times Ns = 2 (FIG. 2 (b)) Similarly, from the relationship between the peak hold length Lh and the sampling pitch Lp, when the number of sampling times Ns = 2, the sampling signal The sampling position at which Sn is maximum is Pn and Pn + 1, and the intermediate position between Pn-1 and Pn is Pn.
′, The defect position is in the range of Pn ′ to Pn (length Lp /
It is determined to be 2).

FIG. 2C shows an example of a case where defects are continuous. In this case, the sampling signal Sn at the sampling position Pn satisfies the condition (in particular). Therefore, the number of sampling times Ns is 1, and the defect position is determined by the method (a).

The present invention has been made based on the above findings.

[0028]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail below with reference to the drawings.

In this embodiment, as shown in FIG.
When winding the steel plate 12 rolled at 0 with the coiler 14,
An apparatus for detecting the crack, which includes a pulse generator (PLG) 20 for detecting the feed length of the steel sheet from the rotation amount of a roll for feeding the steel sheet 12, a sensor 22 for detecting the crack, and An amplifier 24 for amplifying the output of the sensor 22, a data processing device 26, and a host computer 28 for giving coil information to the data processing device 26.
And a printer 30 for outputting the inspection result, the output A of the amplifier 24 is set to a predetermined length Lh.
A peak hold circuit 40 for performing peak hold only, a counter circuit 42 of, for example, a subtraction method for counting the output of the PLG 20 and defining the peak hold length Lh of the peak hold circuit 40, and the peak hold length. A setter 44 for setting a reset value of the counter circuit 42 corresponding to Lh, a sampling circuit 50 for sampling the output Vp of the peak hold circuit 40 at a predetermined pitch Lp, and an output of the PLG 20 are counted. Then, for example, a subtraction type counter circuit 52 for defining the sampling pitch Lp of the sampling circuit 50 and a setter 54 for setting a reset value of the counter circuit 52 corresponding to the sampling pitch Lp are provided. Be prepared.

The peak hold circuit 40, as shown in detail in FIG. 4, holds the peak when the output signal A from the amplifier 24 exceeds a certain threshold value T. The peak hold length Lh is set by the setter 44. The peak hold circuit 40 releases a hold after the counter circuit 42 counts a predetermined count number based on the output pulse (referred to as a PLG signal) of the PLG 20. When the pulse rate is x (mm / P), the count number Ch is determined by the following equation.

Ch = Lh / x (1)

The sampling circuit 50 includes the PLG.
The signal from the peak hold circuit 40 is sampled at a predetermined pitch based on the output pulse of 20. The sampling pitch Lp is set by the setter 54. The counter circuit 52 counts and samples the sampling pitch Lp, and outputs it to the data processing device 26. The count number Cp is determined by the following equation.

Cp = Lp / x (2)

Here, the sampling pitch Lp is set shorter than the peak hold length Lh.

The settings in FIG. 4 are as follows: peak hold length Lh = 100 mm, sampling pitch Lp = 90 m.
m and x = 10 mm / P.

The totaling result by the data processing device 26 can be output from the printer 30 in association with the coil information received from the upper computer 28.

As the sensor 22, Japanese Patent Laid-Open No. 2000-
In addition to the E-type eddy current sensor similar to 230924, other types of sensors such as an optical sensor and a magnetic sensor can be used.

The position of the sensor 22 is set at the steel plate 12
One unit can be arranged on each of the two edges in the width direction. It should be noted that, with respect to variations in the steel plate edge position, a plurality of sensors 22 may be arranged in the width direction or may be scanned in the width direction.

Although the subtraction counter is used in the above embodiment, an addition counter or a distance measuring means of another method may be used.

[0040]

EXAMPLE An example in which the peak hold length Lh is 1.5 times (Example 1), 1.2 times (Example 2) and 1.8 times (Example 3) of the sampling pitch Lp, respectively. To
It shows in FIG. 5 with a comparative example.

In the comparative example shown in FIG. 5A, the amplifier 24
After outputting the sensor signal A, the peak hold is performed until the next sampling, and the sampling circuit outputs the sampling output signal C at the next sampling. In this case, the predicted position of the defect is any position within the one sampling period D before the sampling position at which the sampling output signal C is output.

On the other hand, in the present invention, the peak hold length Lh is made longer than the sampling pitch Lp. As a result, the signal C output from the sampling circuit
Is continuously detected at two consecutive sampling positions (FIG. 5 (d) of the first embodiment, FIG. 5 (g) of the second embodiment,
FIG. 5 (j) of the third embodiment, or is it detected independently at only one sampling position (FIG. 5 of the first embodiment)?
(C), FIG. 5 (f) of the second embodiment, FIG. 5 of the third embodiment
By any of (i), the position of the defect can be specified at intervals E1, E2, E3 shorter than the sampling period. In the figure, A1 is the earliest sensor signal,
A2 is the sensor signal in the slowest case, which means that there is a defect between A1 and A2.

In particular, in the first embodiment in which the peak hold length Lh is 1.5 times the sampling pitch Lp, FIG.
As shown in (b) to (d), the defect position can be specified at the interval E1 that is half the sampling period D.

[0044]

According to the present invention, it is possible to reduce the data processing load and reliably detect a defect with an inexpensive device such as a general-purpose personal computer. Further, stable detection is possible without missing the peak value of the signal. Furthermore,
The defect position can be specified in a narrower range as compared with the case where the defect detection signal is peak-held until the next sampling.

[Brief description of drawings]

FIG. 1 is a time chart showing the concept of peak hold for explaining the principle of the present invention.

[FIG. 2] Similarly, a time chart showing the concept of defect position measurement

FIG. 3 is a configuration diagram showing an entire embodiment of the present invention.

FIG. 4 is a diagram showing an example of a signal waveform of each part in the embodiment.

FIG. 5 is a diagram showing a relationship between a sampling position and a defect position specifying range in an example of the present invention as compared with a comparative example.

[Explanation of symbols]

10 ... rolling mill 12 ... Steel plate 14 ... Koira 20 ... PLG 22 ... Sensor 24 ... Amplifier 26 ... Data processing device 40 ... Peak hold circuit 42, 52 ... Counter circuit 44, 54 ... Setting device 50 ... Sampling circuit

Continued front page    F term (reference) 2F063 AA41 AA50 BA23 BB02 BC05                       BC09 BD07 BD11 BD18 CA04                       CA11 DA01 DA05 DD06 GA08                       LA17 LA29 NA06                 2G051 AA37 DA06 EA02 EA03                 2G053 AA11 AB21 BA02 BA15 BB03                       DA01 DB02 DB06 DB19

Claims (3)

[Claims] 1. When detecting a defect in a running strip, a signal from a sensor is peak-held for a predetermined length, and sampling is performed at a pitch shorter than the length of the peak hold. Defect detection method. 2. The method for detecting defects in a strip-shaped material according to claim 1, wherein the length of the peak hold is 1.5 times the pitch of the sampling. 3. A belt-shaped material defect detecting device for detecting a defect of a belt-shaped material during running, wherein a signal from the sensor is peak-held for a predetermined length, and sampling is performed at a pitch shorter than the peak-holding length. An apparatus for detecting defects in a strip-shaped material, comprising: