JPS608737A - Surface flaw detecting method of metallic material - Google Patents

Abstract

PURPOSE:To detect flaw with high precision not only a longitudinal crack but also a transverse crack or a crack of complex shape such as star crack by irradiating slantly outer light in band shape on a surface of a body to be inspected so that the light crosses a running direction of said body at a prescribed angle. CONSTITUTION:A light emitting source 12 and a projecting part 14 are connected by an optical fiber 16, the part 14 are connected by an optical fiber 16, the part 14 is arranged slantly above the inspecting body 18, and irradiates band- shaped light which forms an angle theta by the intersection of the running direction Y of the body 18. A photodetecting part 22 containing a linear image sensor for photodetecting reflected light forms a projecting and detecting part 24 combined with the part 14. The transverse cracking flaw perpendicular to the direction Y can be detected at >=90% rate in theta having 0-85 deg. range, and the longitudinal cracking flaw parallel to the direction Y can be detected at >=90% rate in theta having 20-90 deg. range. As the practical theta value, angular ranges having 30-85 deg. are effective.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for detecting defects on the surface of a metal object, and in particular, the surface of a moving object is irradiated with light from the outside, and the reflected light from the object is received. This article relates to improvements in surface detection methods for metal objects that detect defects on the surface.

Optical methods are widely used as a technique for online detection of surface flaws on metal objects, particularly on plate-shaped metal objects that are in a high temperature state such as continuous casting slabs during running.

The conventional optical surface flaw detection method is a flying method that scans small diameter dots and/or light and receives the reflected light.
The spot method and the flying image method, which detects defects by scanning a light receiving system with a small field of view with reflected light from an irradiation field with a certain area, have been devised.

In addition, as conventional techniques related to light projection, there are those that condense white light and project the light into a band shape, or the methods that spread laser light into a band shape and project the light.

As a known technology for light reception, information in a microscopic field of view is photoelectrically converted and then electrically scanned at a constant cycle and read out.
l', using a linear image sensor such as a CCD
There are some known examples. This method of using a linear image sensor as a light receiving system is a technology that eliminates the need for a rotating polygon mirror as in the conventional flying image method, regardless of whether white light or laser light is used as the light source. excellent in terms of Furthermore, when using the light receiving system of this method, only a band-shaped field of view is sufficient, and therefore a band-shaped field of view is sufficient for irradiating the object surface. Therefore, when a light source with the same power is used as compared to a planar light projection system, 1q of reflected light with a higher irradiation light amount density can be obtained, which is advantageous.

In addition, conventional reflection technology involves applying a band-shaped irradiation field that spreads in the width direction (direction perpendicular to the running direction) to a moving subject, and detecting changes in the amount of reflected light using a COD sensor, etc., to detect defects. By the way, defects on metal surfaces such as continuous cast slabs are mainly cracks.The shapes are vertical cracks parallel to the running direction, and transverse cracks perpendicular to the running direction. (This includes cracks in keys at corners, etc.).In addition to these, there are also complex-shaped flaws such as star cracks.They are also called tsutomi, vertical cracks, and horizontal cracks. When the flaws are closely observed, they have a somewhat zigzag shape, with a component perpendicular to the running direction and a component parallel to the running direction.

However, the conventional method of applying a band-shaped irradiation field that spreads in the width direction is relatively advantageous in detecting vertical cracks parallel to the direction of travel of the object; Also, it was extremely disadvantageous for detecting flaws in near orthogonal directions. And even if it is detected, that j
The flaw signal is not completely continuous,
The reality was that many of them were disjointed.

Invention A was made in view of these conventional problems, and while taking advantage of the advantages of using a linear image sensor, it can be used to prevent not only vertical cracks but also horizontal cracks and complex shapes such as star cracks. The object of the present invention is to provide a surface flaw detection device for metal objects that enables highly accurate flaw detection even for flaws, etc., and that is easy to detect.

The present invention provides a surface flaw detection 75F1 for a metal object in which the surface of a moving object is irradiated with external light and the surface of the object is detected by receiving reflected light from the object.
From the light emitter placed diagonally above the subject, onto the surface of the subject.
External light is irradiated obliquely and in a strip shape so as to form a predetermined angle with the running direction of the subject, and this reflected light is received by a linear image sensor with the strip-shaped irradiation field as a field of view. The above objective was achieved by detecting flaws on the surface of the object based on the signals.

The present invention will be described in detail below with reference to the drawings.

FIG. 1 shows an example of a surface flaw detection device for metal objects, in which an embodiment of the method of the present invention is adopted. In the figure, reference numeral 10 denotes a light projecting system, which is separated into a light emitting source 12 and a light projecting section 14, which are coupled by an optical fiber 16. Light projecting section 1
4 is disposed obliquely above the subject 18 and can irradiate the surface of the subject 18 with a band-shaped light at a predetermined angle θ with respect to the running direction Y of the subject 18 . A light receiving section 22 including a linear A7 image sensor that receives reflected light from the strip-shaped irradiation field 20 is integrally constructed with the light projecting section 14 to form a single light projecting/receiving section 24 .

When C-carrying the band-shaped irradiation field 20 using a linear image sensor, it is easier to adjust the field of view when the light emitting/receiving section 24 has an integral structure, which is advantageous in terms of maintainability and workability. When a laser light source is used as the light emitting source 12 of the light projecting system 10, the laser light has extremely good power convergence, little attenuation due to distance, and uniform wavelengths, so it can be effectively illuminated using an interference filter. It is possible to cut background noise such as natural light components. Since the laser beam output from the light source 12 is an extremely small spot, if this spot is widened with a cylindrical lens, an extremely small strip-shaped irradiation field 20 with a width of q is obtained. In addition, C126 in FIG. 1 is a conveyance roll,
28 is a signal processing circuit, and 30 is a recorder.

FIG. 2 shows the detection ability of the flaw detection device based on the present invention for vertical cracks parallel to the running direction Y and horizontal cracks perpendicular to the running direction Y of the test object 18, depending on the running direction Y of the test object and the strip-shaped irradiation field 20. The evaluation was made by changing the angle θ formed by the angle θ. For horizontal cracks perpendicular to the running direction Y, a detection rate of 90% or more was obtained in the range of θ from 0 degrees to 85 degrees. Regarding vertical cracks parallel to the running direction Y, a detection rate of 90% or more was obtained in the range of θ from 20 degrees to 90 degrees. In this way, the detection rate decreases in the range where θ is 85 degrees or more for horizontal scratches and less than 20 degrees for vertical cracks, depending on the direction of the scratch and the band-shaped irradiation field 20. This is thought to be due to the fact that the angle becomes close to zero. By the way, if θ is set too small, the length of the band-shaped irradiation field 20 covering the subject width direction X becomes short, resulting in a disadvantage that the number of devices for scanning the entire width direction X increases. Therefore, in practice, the value of θ is 3
An angular range from 0 degrees to 85 degrees is valid. Accordingly, C1 irradiates a strip of light obliquely from the running direction Y of the subject by a certain angle θ in this range, and detects abnormalities (flaws) on the surface of the subject from changes in the amount of reflected light.

The shape of the flaw is determined by the following procedure.

(Also, set the origin at an appropriate reference temperature on the inspection surface of the subject, set virtual orthogonal coordinates, and set the width direction X as the x axis and the traveling direction Y as the y axis. When the value on the coordinates of the obtained flaw signal is as shown in equation (1), it is determined that the flaw is a vertical crack.

Here, A and B are constants, ■ is the traveling speed of the object to be inspected, and [ is the elapsed time from 1 when the flaw signal was generated.

On the other hand, if 10 on the coordinates is 1 as shown in equation (2), it is determined that there is a transverse crack.

x = (C-vt) D, V = C... (2) Here, C and D are constants.

Furthermore, when the coordinate position changes as shown in equation (3), it is generally determined that the linear flaw is a line-shaped flaw that is connected to the traveling direction Y of the subject and a certain inclination α.

X = (vL-E)/(F-tan (π/2-θ
)) y = (F-vt-E ・ shian(yr/2
- θ ) )7('i nee tan (π/2-〇>)
---(3) Here, E and F are constants. Then, F is calculated from the coordinate value x at an arbitrary time and the value of ■, and the inclination angle α of the flaw is calculated from equation (4).

tan −iF −(vt−[)′/x)=π/2−′
cy<4) Furthermore, in each case, the length of the flaw can be calculated based on the elapsed time from the time when the flaw signal is generated.

Next, an example will be described in which an embodiment of the present invention is applied to detect surface flaws in a hot continuously cast slab.

Most of the surface flaws on continuous cast slabs are full-length or partially linear J-cracks. With this flaw detection method, uninterrupted information on cracks can be obtained, even for cracks with a zigzag shape that is partially straight, or cracks with complex shapes such as star cracks.

Embodiment C shows a case in which the surface of a continuous slab is probed on the roller apron exit side (pinch roll exit side) before being cut with a °C1 torch. The characteristics of flaw detection at this position, compared to the case after torch cutting, are (1) since it is further upstream, it is easier to feed back surface information to optimize casting operations; (2) it is easier to feed back surface information for optimization of casting operations; It is possible to control the cutting length of the slab.

Laser light was used as the projection light. In order to more effectively receive the reflected component from the projected light, the wavelength of the laser light is set in a spectral region where the energy of the subject's self-luminescence is low, i.e.
It is desirable to select a short visible wavelength outside the infrared region as much as possible, and in this example, an argon laser that emits laser light with a wavelength of around 500 nm was selected as the optimal light emitting source. The light emitting source and the light projecting unit were coupled through an optical fiber, and the light was transmitted over a distance of nearly 100 meters with high transmission efficiency. This made it possible to easily install the device even on sites with strict installation conditions. Note that the optical fiber between the laser emission source and the light projecting unit is capable of transmitting a laser beam with a wavelength of 400 mm or more and an output of 9 W or more, at a heat receiving temperature of 200°C.
Under the following conditions, the transmission power has a high transmission efficiency of less than 5 dB per 100 m length.

Therefore, even including the lens systems of the light emitting and light receiving sections, the power attenuation was only 50% or less in the case of an optical fiber length of about 90111 mm. The laser beam output from the light projecting section is irradiated onto the subject by a cylindrical lens in the form of a belt approximately 3 meters wide and approximately 1 meter long. The predetermined angle θ between this belt-shaped irradiation field and the slab running direction Y was set to 45°. The reflected light from the band-shaped irradiation field is 1024 elements or 2
The O48 element was used to receive light using an image sensor, and it was decided that normal parts (flaws) could be determined with high resolution from changes in the amount of light.

Now, on the slab surface, the left end of the band-shaped irradiation field at the start of the inspection is set as the origin, the x axis is set in the axial direction X, and y@Il is set virtually in the running direction Y. The X and Y coordinates at each time of the flaw signal obtained within the scanning range irradiated with laser light are detected, and on the other hand,
Determine the flaw threshold under appropriate conditions. Regarding the flaws that are determined to be connected, the flaw length is excluded from the elapsed time from the time when the flaw signal was generated.

Figure 3 shows the specific flow. First, it is determined whether the y-coordinate of the flaw is constant. If it is constant, it is
It's either an oscillation mark or a horizontal crack. To distinguish between oscillation marks and horizontal cracks, follow the steps below. First, since the oscillation marks appear at approximately equal intervals, it is determined whether the flaw signals appear at equal intervals or not.

Next, the frequency components of the signals are different between the A-silation pattern and the flaw. FIG. 4 shows an example of the waveform. Figure (a)
is the waveform of the oscillator gate obtained by frequency analysis. (1) in the same figure is a waveform that has been multiplied by 1q by the light receiving system. By subtracting the waveform component of the oscillation mark from this waveform, a true flaw waveform such as the waveform shown in FIG. 3(C) is obtained. The horizontal crack waveform at the bottom of the J and uni oscillation marks seen in this figure can also be separated. The two operations on the surface accurately distinguish each crack from the subsilation center.

Since the information on the oscillation mark is unnecessary, it is not outputted, but the xy coordinate position of the horizontal crack on the specimen and its length are outputted.

Next, among the signals in which the y-coordinate of the flaw is not constant, it is determined whether the Output the position and its length -4.

Next, the above four
) Calculate the inclination angle α using the calculation method of the formula.

For a flaw signal whose inclination angle σ changes for each length, the inclination angle α that changes sequentially is calculated, and the connection of the squares is determined. If it is determined that there is a connection, one zigzag-shaped It is judged as a crack and output.

For these flaws having an angle of inclination with respect to the running direction Y, the xy coordinate positions and lengths thereof on the object to be inspected are also output.

In these flaw determination flows, the computer uses an optimal blocking parameter that assumes that flaw signals whose addresses are very slightly apart in the X-axis direction and the Y-axis direction are consecutive signals. Optimal filtering parameters are set such that small values are ignored as noise.

As a result of hot online inspection of surface flaws on continuous cast slabs using this method, it was confirmed that various types of paper that are considered harmful to downstream processes and products can be detected with much higher accuracy than conventional methods.

The application of the fugitive steel is not limited to this example, but can be effectively used in fields such as hot-rolled steel sheets, pickled steel sheets, cold-rolled steel sheets, and surface-treated steel sheets. j; The present invention also provides these steel materials,
Very good results can be obtained even when applied to high-speed A-line inspection of sheet-shaped nonmetallic objects such as sheet metal plates, paper, etc., rather than steel plates.

As explained above, according to the present invention, not only vertical cracks 9 but also horizontal cracks and surface flaws with complex shapes such as star cracks can be detected with high precision. Effects can be obtained.

Moreover, since a linear image sensor is used to receive light, there is no need for a rotating polygon mirror as in the conventional flying image method, even when using white light or laser light as a light source. figure,
Furthermore, since detection can be performed using a strip-shaped field of view, compared to the case where the projected light is spread out over a planar surface, the effect of using a linear image sensor, which is that the density of the irradiated light can be maintained at a high level, is synergistically obtained. It is something that can be done.

[Brief explanation of the drawing]

FIG. 1 is a perspective view of a surface flaw detection field for a metal object in which an embodiment of the method of the present invention is adopted, and FIG. Fig. 3 is a flowchart showing the procedure for discriminating various types of paper used in the above-mentioned device. Figures a), (b), and (c) are diagrams illustrating how to distinguish between A sillage marks and horizontal cracks. DESCRIPTION OF SYMBOLS 12... Light emitting source, 14... Light projecting part, 18... Subject, 20... Band-shaped irradiation field, 22... Lee-17.
Image sensor, X...width direction, Y...travel direction, θ...predetermined angle. Agent Takaya Ron (and 1 other person) Figure 1 Figure 2"--1, Line split eave □ Product separation escape OO30a 600 90" □θ Figure 3

Claims (2)

[Claims] (1) A surface flaw detection method for metal objects in which the surface of a moving object is irradiated with light from the outside and the reflected light from the object is received to detect flaws on the surface of the object. External light is irradiated obliquely and in a strip shape onto the surface of the subject from a light projector placed diagonally above so as to form a predetermined angle with the running direction of the subject, and this reflected light is used to illuminate the strip-shaped irradiation field. A method for detecting defects on the surface of a metal object, in which light is received by a linear image sensor as a field of view, and the surface of the object is detected based on the received light signal. (2) The method for detecting defects on the surface of a metal object according to claim 1, wherein the predetermined angle is from 30 degrees to 85 degrees.