JP2024035424A - Method for inspecting defect of fiber bundle and method for manufacturing fiber bundle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an inspection method for capable of quickly detecting defects occurring in a process of manufacturing a fiber bundle.

SOLUTION: A method for inspecting defects in a plurality of fiber bundles running in parallel, comprises: (A) irradiating a pre-assigned region of a running path of the plurality of fiber bundles with measurement light; (B) imaging the region with imaging means arranged at a position of being able to receive reflected light of the measurement light from the plurality of fiber bundles; and (C) processing image data obtained in the step (B) to obtain defect data. The step (C) includes (a) a process of detecting a defect based on difference in brightness between a fiber bundle portion and a defective portion from the image data obtained in the step (B); (b) a process of classifying the detected defect using a threshold value depending on one or more preset shape characteristics; and (c) a process of linking the coordinates of the detected defect with the coordinates in the region, and identifying a fiber bundle in which the defect has occurred.

SELECTED DRAWING: Figure 4

COPYRIGHT: (C)2024,JPO&INPIT

Description

The present invention relates to a method for inspecting defects in fiber bundles and a method for manufacturing fiber bundles.

Synthetic fibers are widely used in clothing and industrial materials. For example, nylon fibers and polyester fibers are used for clothing materials, aramid fibers are used for high-strength belts and tire cords, and polyacrylonitrile fibers are used as precursors for carbon fibers. Further, carbon fiber obtained by firing this precursor is used as a high-performance fiber in various fields such as sports, automobiles, ships, and civil engineering and construction.
While these fibers are required to have high performance, they are also required to reduce costs in order to meet the demands of users. In order to reduce costs, efforts are being made to increase the size of production equipment, increase production speed, and increase the total number of fibers per unit production machine.

However, as production equipment becomes larger and production speed increases, it becomes more difficult to uniformly process fiber bundles, and minor defects in fiber bundles become a serious problem. For example, during the precursor manufacturing process, defects such as fuzz, pilling, adhesion of foreign matter, uneven thickness, etc. due to breakage of single fibers may occur. If such a defect exists, when the fiber is fired to produce carbon fiber, the processability may be deteriorated due to wrapping around a roll, etc., or it may lead to abnormal cutting during firing. Furthermore, when producing a prepreg using the obtained carbon fiber, there are cases where the spreadability is reduced and process passability is deteriorated, or gaps are formed in the prepreg and the appearance quality is deteriorated. Therefore, defects in fiber bundles have become an important item to be managed.

Patent Document 1 proposes a method in which a measurement light is irradiated to a range exceeding the width of a fiber bundle, the reflected light is received, and defects are detected from the light intensity of the reflected light.

Patent Document 2 describes a method of detecting defects in fiber bundles by illuminating the fiber bundle, capturing an image of the fiber bundle together with the background using a camera, and performing processing such as binarization and thinning on the obtained image. Proposed.

JP2008-203251A Japanese Patent Application Publication No. 2011-053173

However, with the method described in Patent Document 1, it is possible to detect large defects such as foreign matter adhering to a wide range of fiber bundles and entangled growth of fluff of single fibers, but small defects such as fluff of single fibers can be detected. Difficult to detect.
The method described in Patent Document 2 can detect fuzz, fluff, etc. that appear to be protruding from the fiber bundle when viewed from the camera, but it is difficult to detect defects that fit within the width of the fiber bundle.

An object of the present invention is to provide a defect inspection method that can quickly detect defects that occur during the manufacturing process of fiber bundles, and a method of manufacturing fiber bundles using this method.

The present invention has the following aspects.
[1] A method for inspecting defects in a plurality of fiber bundles running in parallel,
(A) irradiating measurement light onto a pre-assigned region of the travel path of the plurality of fiber bundles;
(B) imaging the area under irradiation with the measurement light using an imaging means arranged at a position capable of receiving reflected light from the plurality of fiber bundles of the measurement light;
(C) processing the image data obtained in the step (B) to obtain defect data;
The step (C) above is
(a) A process of detecting a defect based on the difference in brightness between the fiber bundle portion and the defective portion from the image data obtained in the step (B);
(b) a process of classifying the detected defects using a threshold value according to one or more preset shape characteristics;
(c) A method for inspecting defects in fiber bundles, including the steps of: associating the coordinates of the detected defect with the coordinates in the area, and identifying the fiber bundle in which the defect has occurred.
[2] (D) The fiber bundle defect inspection method according to [1], further comprising the step of recording the defect data obtained in the step (C).
[3] The fiber bundle according to [1] or [2], wherein the threshold value depending on the one or more shape characteristics includes a threshold value depending on the shape feature in which some single fibers of the fiber bundle are cut or separated. defect inspection method.
[4] Any one of [1] to [3], wherein the threshold according to the one or more shape features includes a threshold according to the shape feature in which some single fibers of the fiber bundle are cut and become a lump. The described method for inspecting defects in fiber bundles.
[5] The threshold according to the one or more shape features includes the threshold according to the shape feature having a length of 15 mm or more where some single fibers of the fiber bundle are cut. A method for inspecting defects in a fiber bundle according to any one of the above.
[6] The fiber bundle defect inspection method according to any one of [1] to [5], wherein the number of single fibers constituting each of the plurality of fiber bundles is 1,000 to 100,000.
[7] The fiber bundle defect inspection method according to any one of [1] to [6], wherein the number of the plurality of fiber bundles is 10 to 1000.
[8] A step of manufacturing a plurality of fiber bundles;
A method for producing a fiber bundle, comprising the step of running the plurality of fiber bundles in parallel and inspecting for defects using the fiber bundle defect inspection method according to any one of [1] to [7].
[9] The method for manufacturing a fiber bundle according to [8], wherein when a defect is detected in the inspecting step, the conditions of the manufacturing step are changed according to the data of the detected defect.
[10] The method for manufacturing a fiber bundle according to [8] or [9], wherein an alarm is issued when a defect is detected in the inspection step.

According to the present invention, it is possible to provide a defect inspection method that can quickly detect defects that occur during the manufacturing process of fiber bundles, and a method for manufacturing fiber bundles using this method.

FIG. 2 is a schematic diagram showing an example of a defect in a fiber bundle to be detected in the present invention. FIG. 2 is a schematic diagram showing an example of a defect in a fiber bundle to be detected in the present invention. FIG. 2 is a schematic diagram showing an example of a defect in a fiber bundle to be detected in the present invention. It is a schematic side view showing an example of a defect inspection device.

The fiber bundle defect inspection method of the present invention is a method for inspecting defects in a plurality of fiber bundles running in parallel.
Hereinafter, the present invention will be described in detail by showing embodiments.

(fiber bundle)
A fiber bundle is composed of a plurality of single fibers.
The type of single fiber is not particularly limited, but preferable examples include synthetic fibers such as polyacrylic fibers, polyester fibers, nylon fibers, aramid fibers, and arylate fibers, and carbon fibers.

The number of single fibers constituting one fiber bundle is preferably 1,000 to 100,000. If the number of single fibers is 1000 or more, productivity can be improved, and if the number is 100000 or less, defects can be sufficiently detected.
From these viewpoints, the number of single fibers constituting the fiber bundle is more preferably 3,000 to 80,000, and even more preferably 5,000 to 70,000.

The number of fiber bundles running in parallel is preferably 10 to 1000. If the number of fiber bundles is 10 or more, defects can be detected efficiently, and if the number is 1000 or less, defects can be detected sufficiently.
From these viewpoints, the number of fiber bundles is more preferably from 15 to 800, and even more preferably from 30 to 500.

By running the fiber bundle at a predetermined position on the travel path, it becomes easier to identify the fiber bundle in which a defect has occurred. Therefore, it is preferable to apply a tension of 0.01 cN/dtex or more to each of the traveling fiber bundles. By applying a tension of 0.01 cN/dtex or more, it is possible to prevent the fiber bundle from meandering or to prevent several single fibers from coming off the fiber bundle.
More preferably, the tension is 0.05 cN/dtex or more.
The upper limit of the tension may be within a range in which the fiber bundle does not break.

(defect)
Defects to be detected typically originate from single fibers that constitute a fiber bundle.
Conventionally, it has been difficult to detect small defects or defects that fit within the width of one fiber bundle at most. According to the invention, these defects can also be detected. Therefore, the present invention is suitable for detecting these defects.
Specific examples of the above-mentioned defects include a defect 11 in which some of the single fibers of the fiber bundle 1 are cut or separated, as shown in FIG. Examples include a defect 12 in which single fibers are cut and become a lump, and a defect 13 having a length of 15 mm or more in which some single fibers of a fiber bundle are cut, as shown in FIG.
The size of the defect 11 is, for example, 1 mm or more and less than 15 mm in terms of the main axis length. The size of the defect 12 is, for example, 1 mm ² or more in terms of area.

(Defect inspection equipment)
FIG. 4 is a schematic side view showing an example of a defect inspection apparatus used in the defect inspection method of the present invention.
The fiber bundle defect inspection device in this example is
illumination means 2 that irradiates a predetermined area (hereinafter also referred to as area X) with measurement light L1;
an imaging means 3 arranged at a position capable of receiving reflected light L2 of the measurement light L1 from the plurality of fiber bundles 1, and configured to image the region X;
a data processing means 4 for processing the image data obtained by the imaging means 3 and obtaining defect data;
It has a recording means 5 for recording defect data obtained by the data processing means 4.
The plurality of fiber bundles 1 are configured to run in parallel by a plurality of transport rolls 6.

<Area X>
The region X is a region assigned in advance to the traveling route of a plurality of fiber bundles 1 running in parallel in order to detect defects. The location to which the region
Regarding the width of the region X, that is, the length of the region X in the width direction, it is sufficient that all the fiber bundles 1 running in the region X are included in the region X.

<Lighting means>
The illumination means 2 is arranged on one side of a plane including the region X. Like the illumination means 2, the imaging means 3 is also arranged on one side of the plane containing the region X.
The plane including the region (the depth direction in FIG. 4; hereinafter also referred to as the width direction).

The angle θ1 between the plane containing the plurality of fiber bundles 1 and the measurement light L1 is preferably 10 to 90 degrees, more preferably 20 to 65 degrees. If the angle θ1 is within the above range, it will be less susceptible to vibrations of the object or changes in the way it appears due to unevenness of the object itself.

The illumination means 2 is not limited to the intensity or wavelength of the measurement light to be irradiated, as long as sufficient reflected light from the fiber bundle 1 can be obtained. It can be selected as appropriate depending on the installation location of the illumination means 2, the number of fiber bundles 1, the distance from the illumination means 2 to the fiber bundles 1, etc.
When simultaneously inspecting each of the plurality of fiber bundles 1 for defects in the region X where the plurality of fiber bundles 1 are running in parallel, it is preferable that each fiber bundle 1 can be uniformly illuminated in the width direction. When the number of fiber bundles is large, a plurality of illumination means 2 can be used.

Specific examples of the illumination means 2 include high-frequency lighting fluorescent lamps, metal halide lamps, LED lighting, X-rays, infrared light, ultraviolet light, and the like. Alternatively, light from a light source such as a halogen or LED may be guided through an optical fiber and irradiated.
When simultaneously inspecting the defects present in each of the plurality of fiber bundles 1 in the region Preferably, lighting is used.

<Imaging means>
The imaging means 3 may be any device that can obtain data regarding the brightness of the reflected light L2 from the region Sensor cameras arranged in a targeted or two-dimensional manner can be employed. In particular, when simultaneously inspecting each of a large number of fiber bundles running in parallel over a wide area for defects, the light-receiving elements are arranged linearly because they have excellent width resolution and can inspect a wide range. A line sensor camera is preferred.
The reflected light L2 may be diffusely reflected light or specularly reflected light.

The angle θ2 formed by the plane containing the plurality of fiber bundles 1 and the plane extending from the center of the region X to the imaging means 3 in the running direction of the plurality of fiber bundles 1 is preferably 10 to 90 degrees, It is more preferable that the If the angle θ2 is within the above range, it will be less susceptible to vibrations of the object or changes in the way it appears due to unevenness of the object itself.

<Data processing means>
The data processing means 4 processes the image data obtained by the imaging means 3 to obtain defect data. The processing in the data processing means 4 will be explained in detail later.
As the data processing means 4, a CPU etc. can be mentioned.

<Recording means>
The recording means 5 records the defect data obtained by the data processing means 4.
The recording means 5 may further record image data obtained by the imaging means 3.
Examples of the recording means 5 include HDD, NAS, and the like.

(Defect inspection method)
An embodiment of a defect inspection method using the defect inspection apparatus shown in FIG. 4 will be described.
The defect inspection method of this embodiment includes the following steps (A) to (D).
(A) A step of irradiating the area X with the measurement light L1 using the illumination means 2.
(B) Step of imaging the region X by the imaging means 3 under irradiation with the measurement light L1.
(C) Processing the image data obtained in step (B) by the data processing means 4 to obtain defect data.
(D) A step of recording the defect data obtained in step (C), if necessary.

<(B) process>
In order to detect defects with high accuracy, it is preferable to image the region X with a resolution that is five times or more than the defect size to be detected. For example, when detecting a defect with a size of 1 mm in both the width direction and the longitudinal direction, the imaging resolution is preferably 0.2 mm or less, and preferably 0.1 mm or less.
On the other hand, if the imaging resolution is set too fine, the amount of data becomes too large and the data processing speed becomes slow. Therefore, the imaging resolution is preferably 0.01 mm or more. If it is 0.01 mm or more, it can be used without slowing down the data processing speed too much.

The imaging resolution in the running direction of the fiber bundle 1 is related to the number of imaging times per time. Therefore, the number of times of imaging per hour is adjusted according to the size of the defect to be detected, the traveling speed of the fiber bundle 1, etc.
For example, if the size of the defect to be detected in the running direction of the fiber bundle 1 is 1 mm and the running speed of the fiber bundle 1 is 300 m/min (=5000 mm/sec), the number of imaging times per hour is 25000 times/ It is preferable that the time is longer than seconds. If the number of times of imaging is 25,000 times/second or more, the imaging resolution will be 0.2 mm or less, and defects of the above size can be detected with high accuracy.

The image data obtained in step (B) is guided to data processing means 4.
Image data is a collection of pixel data arranged in a grid. The pixel data is preferably represented by 256-level gray scale data, for example, but is not particularly limited.
The image data obtained in step (B) includes at least one portion of the fiber bundle and a background portion. While the irradiated light L1 is reflected by the fiber bundle 1, the irradiated light L1 is not reflected by the background portion, so in the obtained image data, the fiber bundle 1 portion has higher brightness than the background portion. Further, if a defect exists on the fiber bundle 1, the irradiation light L1 is scattered by the defect, so that a difference in brightness (difference in brightness) occurs between the fiber bundle 1 portion and the defective portion in the obtained image data.

<(C) process>
In the step (C), the data processing means 4 performs the following processes (a) to (c).
(a) A process of detecting a defect from the image data obtained in the step (B) based on the difference in brightness between one portion of the fiber bundle and the defective portion.
(b) A process of classifying the defects detected in the process of (a) using a threshold value according to one or more preset shape characteristics.
(c) A process of linking the coordinates of the defect detected in the process of (a) with the coordinates in the region X to identify the fiber bundle in which the defect has occurred.

[Processing of (a)]
As mentioned above, when the fiber bundle 1 includes a defect, there will be a difference in brightness (brightness difference) between the defective part and the part of the fiber bundle 1 in the image data, so the defective part can be identified from the image data based on this difference in brightness. can be extracted.
Specifically, in the process (a), the following first data processing procedure, second data processing procedure, third data processing procedure, and fourth data processing procedure are performed. However, the third data processing procedure can also be omitted.

The first data processing procedure is a data processing procedure for obtaining extracted data from the brightness difference between the fiber bundle 1 portion and the defective portion in the image data obtained in the step (B).
The image data obtained in the step (B) includes one part of the fiber bundle, a defective part included in the first part of the fiber bundle, and a background part. In the first data processing procedure, the difference in brightness between the fiber bundle portion and the defective portion is sharpened from this image data to obtain extraction portion data.
Specifically, the difference in brightness and darkness can be made clearer by contrast expansion.

The second data processing procedure is a data processing procedure that removes data having a predetermined number of pixels or less from the extraction part data obtained by the first data processing procedure to obtain first residual part data.
By removing data below a certain number of pixels, noise components (dust, fine fluff, etc.) can be removed. The upper limit of the number of pixels of data to be removed can be set depending on the size of the defect to be detected.

The third data processing procedure is a data processing procedure that performs a blurring process on the first residual portion data obtained by the second data processing procedure to obtain second residual portion data.
Here, the blurring process refers to an averaging filter, and is a process that converts the density of the center pixel of a certain square range (for example, 3 x 3) to the average density within the square range that includes the center pixel. . By performing the blurring process on the first residual portion data, it is possible to determine a plurality of small defects that are close to each other as one defect.

The fourth data processing procedure uses the first residual data obtained by the second data processing procedure or the second residual data obtained by the third data processing procedure (hereinafter, these are collectively referred to as residual data). This is a data processing procedure in which defect data is obtained by comparing the data (also referred to as partial data) with a preset threshold value to determine the presence or absence of a defect.
In the first remaining portion data and the second remaining portion data, portions that are likely to be defects included in the fiber bundle are extracted. A threshold value and judgment condition (if it exceeds the threshold value, it is a defect, or if it is below the threshold value, it is a defect) are set in advance according to the shape characteristics of the defect to be detected, and the remaining part data and the threshold value are set in advance. By comparing, it is possible to determine whether there is a defect.
As the threshold value according to the shape feature, for example, the length of the principal axis of the remaining portion data can be cited. The principal axis length is the length of the long side when a circumscribed rectangle containing the largest number of remaining parts is drawn. If the main axis length of the remaining portion data exceeds a preset threshold, the remaining portion data can be determined to be defective.
The remaining portion data determined to be defective is defined as defective data.

[Processing in (b)]
In the process (b), the defects detected in the process (a) are classified using a threshold value according to one or more preset shape characteristics. Depending on the conditions of the manufacturing process of the fiber bundle 1, the types (shape characteristics) of defects that occur may vary. By classifying the defects, the conditions of the manufacturing process of the fiber bundle 1 can be adjusted and the quality of the fiber bundle 1 can be improved.
One or more shape characteristics for classifying defects include, for example, the brightness of defect data, the length of the main axis, and the number of pixels. The brightness of the defect data indicates the unevenness of the defect. The principal axis length of the defect data indicates the maximum length of the defect. The number of pixels in the defect data indicates the area of the defect. By comparing the brightness, principal axis length, and number of pixels of the defect data obtained by the fourth data processing procedure with preset threshold values, the defect data can be classified by defect type.

The threshold value according to one or more shape characteristics preferably includes a threshold value according to the shape characteristic of a defect 11 in which some single fibers of the fiber bundle 1 are cut or separated, as shown in FIG.
These shaped defects 11 rarely directly affect the quality of the fiber bundle, but they may subsequently come into contact with conveyor rolls or part of the equipment while traveling in the manufacturing process, causing friction and degrading the quality. In some cases, the defect may transform into a lumpy defect. Therefore, from the viewpoint of quality control of the fiber bundle 1, it is preferable to detect the defect 11.

It is also preferable that the threshold value according to one or more shape characteristics includes a threshold value according to the shape characteristic of a defect 12 in which some single fibers of the fiber bundle 1 are cut and become a lump, as shown in FIG.
If the fiber bundle 1 includes a defect 12, surrounding fiber bundles 1 may become entangled starting from the defect 12 when the yarn is unwound from the package in which the fiber bundle 1 is wound. In this case, the shape of the unwound fiber bundle 1 may collapse, or defects 12 may remain on the unwound fiber bundle 1, which may cause the fiber bundle 1 to be caught in equipment during additional machining. Therefore, among the various types of defects, it is preferable that the bulk defects 12 be detected before packaging and not included in the fiber bundle package.

The threshold value according to one or more shape characteristics may include the threshold value according to the shape characteristic of a defect 13 having a length of 15 mm or more, in which some single fibers of the fiber bundle 1 are cut, as shown in FIG. preferable.
The presence or absence of defects 13 can serve as auxiliary data for determining whether the breakage of the single fibers is due to excessive contact with equipment or a process in which the fiber bundle is stretched. Therefore, when the number of defects 13 increases rapidly, it is possible to narrow down the steps in which countermeasures are taken.

An example of setting the threshold value according to the shape characteristics of the defect 11 is an example in which the value of the main axis length is set to 1 mm. In this case, if the value of the main axis length of the defect data exceeds the above threshold value, the defect data is classified as defect 11. Note that in this case, the detected defect 11 includes the defect 13. Further defects 13 can be extracted as needed.
An example of setting the threshold value according to the shape characteristics of the defect 12 is an example in which the area value is set to 1 mm ² . In this case, if the area value of the defect data exceeds the threshold, the defect data is classified as defect 12.
An example of setting the threshold value according to the shape characteristics of the defect 13 is an example in which the value of the main axis length is set to 15 mm. In this case, if the value of the main axis length of the defect data exceeds the above threshold value, the defect data is classified as defect 13.
However, the threshold value may vary depending on the measurement conditions. It is possible to compare data obtained under conditions matching the inspection measurement conditions for a sample with a known defect and a sample without a defect, and set a threshold value that can determine the presence or absence of the defect.

[Processing in (c)]
In the process (c), the coordinates of the defect detected in the process (a) are linked with the coordinates in the region X, and the fiber bundle in which the defect has occurred is identified.
In the width direction of the region X, the position at which each of the plurality of fiber bundles 1 runs is approximately constant. Depending on where the center of gravity coordinates of the defect are located in the width direction of the region X, the fiber bundle containing the defect can be specified.
By identifying the fiber bundle in which a defect has occurred, the quality of the fiber bundle 1 can be controlled, and the quality of the fiber bundle can be improved by removing the defective fiber bundle.

<(D) process>
Although step (D) is not essential, since the traveling speed is high in the fiber bundle manufacturing process, it is difficult to visually check the traveling fiber bundles to understand the form of defects and the shape of the fiber bundles. difficult. Therefore, it is preferable to perform the step (D). This makes it possible to identify defects that cannot be detected visually, which can be used to improve manufacturing processes and product quality.
Image data captured by the imaging means 3 may be recorded together with the defect data.

After the fiber bundle is produced in a fiber bundle manufacturing process, it may be wound into a fiber bundle package.
The fiber bundle inspected by the defect inspection method according to the present embodiment may be a fiber bundle manufactured in the fiber bundle manufacturing process but before being made into a fiber bundle package, or may be a fiber bundle that is unwound from the fiber bundle package. It may also be a fiber bundle.

By performing an inspection process using the defect inspection method of the present invention when manufacturing a fiber bundle, it is possible to easily identify the presence or absence of a defect, the type of defect, the fiber bundle in which a defect has been detected, and the like. Therefore, it is possible to prevent defective fiber bundles from flowing out. Furthermore, by classifying defects, it becomes easier to identify the location where an abnormality occurs in the process. Therefore, abnormalities that occur during the process of manufacturing the fiber bundle can be quickly detected, and thereby the quality of the fiber bundle can be well controlled, making it easier to obtain high quality that is not limited to visual inspection.

(Method for manufacturing fiber bundle)
The method for manufacturing a fiber bundle of the present invention includes a step of manufacturing a fiber bundle (manufacturing step);
The method includes a step (inspection step) of running a plurality of fiber bundles produced in the manufacturing process in parallel and inspecting for defects using the fiber bundle defect inspection method of the present invention.
The manufacturing process can be carried out by a known method.

In the fiber bundle manufacturing method of the present invention, when a defect is detected in the inspection process, it is preferable to change the conditions of the manufacturing process according to the data of the detected defect.
The number, location, size, length, etc. of defects that occur vary depending on the conditions of the manufacturing process. By carrying out the detection process and changing the manufacturing process conditions according to the defect data detected in the detection process, defects in the manufacturing process can be discovered early, and the process can be improved and conditions changed to improve quality. It becomes easier to obtain good fiber bundles.

In the fiber bundle manufacturing method of the present invention, it is preferable to issue an alarm when a defect is detected in the inspection process.
By issuing an alarm, it is possible to promptly take measures such as process improvement as described above.
From the viewpoint of responding quickly to the alarm, it is preferable that the alarm be notified by a lamp or sound near the manufacturing equipment or in the control room, or that the details be displayed on a monitor.
When issuing an alarm, the type of alarm may be changed depending on the type and frequency of the defect.

When a defect is detected in the inspection process, an alarm may be issued and the conditions of the manufacturing process may be changed according to the data of the detected defect.
In this case, the fiber bundle manufacturing method of the present invention identifies the abnormality in the manufacturing process from the data of the detected defects (number, location, size, length, etc.), and adjusts the manufacturing process based on the identified abnormality. It may also include operating procedures for changing conditions.
In the inspection process, when there is a sudden change in the type of defects detected or a sudden increase in the number of defects, an alarm is issued, identifying abnormalities in the manufacturing process, and inspecting the manufacturing process according to the abnormality. By taking appropriate measures, it is possible to shorten the time during which a running fiber bundle contains defects. As a result, many fiber bundles can be manufactured with few or no defects in the operating time of the manufacturing process.

In the fiber bundle package obtained using the fiber bundle manufactured by the fiber bundle manufacturing method of the present invention, the type and number of defects contained in the fiber bundle in the package and the position information of the defects can be ascertained through the inspection process. There is. Therefore, for example, when unwinding a fiber bundle from a fiber bundle package and using it in a later process, only fiber bundles with no defects can be used, and defects in the later process, which were previously caused by defects, can be used. Troubles can be prevented and yields can be improved.

EXAMPLES Hereinafter, the present invention will be explained in more detail with reference to Examples, but the present invention is not limited by the Examples.

(Example 1)
KDPL2-LK1000W manufactured by Kyoto Denkiki Co., Ltd. was used as the illumination means for the measurement light. As an imaging means, a line camera XG-HL08M (8192 pixels) manufactured by Keyence Corporation was used. An image sensor XG-8700L manufactured by Keyence Corporation was used as a data processing means and a recording means. I connected the XG-HL08M to the XG-8700L and saved all captured images to the storage inside the XG-8700L.

As shown in FIG. 4, the illumination means was installed on one side of the travel path of the fiber bundle so that the distance from the irradiation part of the illumination means to the area X was 200 mm, and the illumination means was inclined at 45 degrees with respect to the traveling direction. The imaging means was installed on one side of the running path of the fiber bundle so that the distance from the region X to the tip of the lens of the imaging means was 1500 mm and was inclined at 30 degrees with respect to the running direction, and images were acquired by diffuse reflection.
The principal axis length and area were set as the threshold values according to the shape characteristics in the fourth data processing procedure of the process (a). In the process of (b), the principal axis length and area were set as threshold values according to the shape characteristics.

Defect inspection was conducted for 3 days during the manufacturing process of a polyacrylonitrile fiber bundle (single fiber fineness 1.0 dtex) consisting of 60,000 single fibers.
During the period of defect inspection, 12 fiber bundles were run in parallel at 60 m/min, the area X was irradiated with the illumination means, and the image was taken with the imaging means. At this time, the number of imaging times per second was set to 10,000 so that the imaging resolution in the direction perpendicular to the traveling direction of the fiber bundle was 0.1 mm.

Based on the image data acquired by imaging, the data processing means performs the above-mentioned processes (a) to (c) to detect defects in each fiber bundle, compare fiber bundles, and compare them during the day and at night. We monitored the changes in and the changes over time.
In parallel, a visual defect inspection was conducted, and the results were consistent with the above defect inspection results.

Defect detection for each fiber bundle was performed in-line using data processing means. The acquired defect information, including the defect image, location, and time of occurrence, was displayed on a monitor at the site.
The in-line results also matched the visual defect inspection results.

During the manufacturing process of fiber bundles, if the surface of, for example, a roll for running the fiber bundle or a guide roll that determines the running position of the fiber bundle is damaged, defects will occur in the fiber bundle at locations that come into contact with the damaged area.
By detecting these defects inline, we were able to detect defects early. By appropriately treating abnormal locations, process abnormalities could be avoided and yields improved.

1 Fiber bundle 2 Illumination means 3 Imaging means 4 Data processing means 5 Recording means 6 Conveyance roll 11 Defects (cuts/separations)
12 Defect (lump)
13 Defect (long single fiber)

Claims (10)

A method for inspecting defects in a plurality of fiber bundles running in parallel, the method comprising:
(A) irradiating measurement light onto a pre-assigned region of the travel path of the plurality of fiber bundles;
(B) imaging the area under irradiation with the measurement light using an imaging means arranged at a position capable of receiving reflected light from the plurality of fiber bundles of the measurement light;
(C) processing the image data obtained in the step (B) to obtain defect data;
The step (C) above is
(a) A process of detecting defects from the image data obtained in the step (B) above based on the difference in brightness between the fiber bundle portion and the defective portion;
(b) a process of classifying the detected defects using a threshold value according to one or more preset shape characteristics;
(c) A method for inspecting defects in fiber bundles, including the steps of: associating the coordinates of the detected defect with the coordinates in the area, and identifying the fiber bundle in which the defect has occurred. The fiber bundle defect inspection method according to claim 1, further comprising the step of (D) recording data on the defects obtained in the step (C). 2. The fiber bundle defect inspection method according to claim 1, wherein the threshold value depending on the one or more shape characteristics includes a threshold value depending on a shape feature in which some single fibers of the fiber bundle are cut or separated. 2. The fiber bundle defect inspection method according to claim 1, wherein the threshold value according to the one or more shape characteristics includes a threshold value according to a shape feature in which some single fibers of the fiber bundle are broken and become a lump. 2. The defect inspection of a fiber bundle according to claim 1, wherein the threshold value according to the one or more shape features includes a threshold value according to a shape feature having a length of 15 mm or more where some single fibers of the fiber bundle are cut. Method. The fiber bundle defect inspection method according to claim 1, wherein the number of single fibers constituting each of the plurality of fiber bundles is 1,000 to 100,000. The fiber bundle defect inspection method according to claim 1, wherein the number of the plurality of fiber bundles is 10 to 1000. a step of manufacturing a plurality of fiber bundles;
A method for manufacturing a fiber bundle, comprising the step of running the plurality of fiber bundles in parallel and inspecting for defects using the fiber bundle defect inspection method according to any one of claims 1 to 7. 9. The method for manufacturing a fiber bundle according to claim 8, wherein when a defect is detected in the inspecting step, conditions for the manufacturing step are changed according to data of the detected defect. The method for manufacturing a fiber bundle according to claim 8, wherein an alarm is issued when a defect is detected in the inspecting step.

Images