Method and apparatus for analyzing reinforcement fiber bundles

Abstract

To provide a reinforcing fiber bundle analysis method capable of accurately separating a region of reinforcing fibers from a region of convergence agent in a tomographic image and analyzing a distribution state of a process agent inside of the reinforcing fiber bundle in a short time and with good reproducibility.SOLUTION: The reinforcing fiber bundle analysis method includes the steps of obtaining a tomographic image of a reinforcing fiber bundle parallel to the length direction of the reinforcing fibers, calculating a differential value of the gray value for a location in the length direction of the reinforcing fibers of the tomographic image to obtain a differential image, obtaining a mask image for extracting an analysis region including a process agent by setting a region expanded in the length direction of the reinforcing fiber starting from a location on the differential image corresponding to a differential value of a certain threshold value or more, and clustering the analysis region extracted by the mask image in the tomographic image.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a method and an apparatus for analyzing a reinforcing fiber bundle. [Background technology]

[0002] In recent years, reinforced resin compositions containing reinforcing fibers such as carbon fibers have been attracting attention as industrially important materials because of their excellent mechanical properties such as strength and rigidity despite their low specific gravity. In particular, reinforced resin compositions are attracting attention as alternative materials to metal materials in automobile and aircraft parts.

[0003] The reinforcing fibers used in such reinforced resin compositions are bundles of multiple reinforcing fibers (reinforcing fiber bundles) impregnated with a sizing agent for the purposes of preventing breakage during molding or for the purposes of modifying the surface to increase compatibility with the resin.

[0004] However, since the impregnation state of the sizing agent in the reinforcing fiber bundle affects the fiber opening (dispersibility) during molding of the reinforced resin composition, a method for analyzing the impregnation state of the sizing agent in the reinforcing fiber bundle is required. For example, a method is known in which the coating state of the sizing agent on the surface of the reinforcing fiber bundle is analyzed by observing the surface of the reinforcing fiber bundle with a SEM (for example, Patent Document 1).

[0005] In addition, the strength of a molded product of a reinforced resin composition is likely to vary depending on the orientation direction (meandering state) of the reinforcing fibers in the molded product. Therefore, a method for analyzing the meandering state of the reinforcing fibers in the molded product has also been considered. For example, an analysis method is known that includes the steps of 1) acquiring an X-ray CT image of a molded product of a reinforced resin composition containing carbon fibers and a matrix resin, 2) binarizing the obtained image to separate (separate) the carbon fiber region and the matrix resin region, and 3) quantifying the meandering state of the carbon fibers from the fiber information obtained by the binarization (for example, Patent Document 2). In addition, this document describes that, when performing the binarization process in 2), a filter process such as a median process is performed to make the boundary between the carbon fibers and the matrix resin clear. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Special Publication No. 2017-504734 [Patent Document 2] JP 2018-159691 A Summary of the Invention [Problem to be solved by the invention]

[0007] However, the analysis method shown in Patent Document 1 was only capable of observing the adhesion state of the sizing agent on the surface of the reinforcing fiber bundle, and was not able to observe the distribution state or amount of the sizing agent inside the reinforcing fiber bundle.

[0008] In addition, Patent Document 2 analyzes a molded body of a reinforced resin composition. In a tomographic image of a molded body of a reinforced resin composition, the boundary between the reinforcing fiber region and the matrix resin region is relatively clear, and the pixel values ​​(e.g., gray values) of the pixels constituting the reinforcing fiber region and the pixel values ​​of the pixels constituting the matrix resin region are unlikely to overlap (are different). Therefore, the reinforcing fiber region and the sizing agent region can be separated depending on whether the pixel value is equal to or greater than a threshold, and region division by binarization processing is possible. However, in a tomographic image of a reinforcing fiber bundle, for example, when the densities of the reinforcing fiber and the sizing agent are close, or when the energy of the X-rays used in the X-ray CT measurement is low and X-rays are likely to refract, the boundary between the reinforcing fiber region and the sizing agent region is relatively unclear, and the pixel values ​​of the pixels constituting the reinforcing fiber region and the pixel values ​​of the pixels constituting the sizing agent region may partially overlap (be the same) with each other. In such a case, it is not possible to distinguish between the reinforcing fiber region and the sizing agent region depending on whether the pixel value is equal to or greater than a threshold value, making it impossible to separate the regions using a binarization process as in Patent Document 2.

[0009] Furthermore, when attempting to analyze the sizing agent in a reinforcing fiber bundle in an SEM image, etc., there was no effective method for selecting (extracting) the area where the sizing agent exists to be analyzed. In other words, even with image processing software, there was no effective method for selecting the area where the sizing agent exists in the image, and it was necessary to extract it manually. Therefore, not only was the analysis work time-consuming, but the selection criteria were not strict, so there was also the problem that it was not possible to perform a precise and reproducible analysis.

[0010] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a method and apparatus for analyzing reinforcing fiber bundles that can accurately separate reinforcing fiber regions and bundling agent regions even when the boundary between the reinforcing fiber regions and the bundling agent regions is unclear in a tomographic image, and that can analyze the distribution state of a treatment agent inside a reinforcing fiber bundle in a short period of time with good reproducibility. [Means for solving the problem]

[0011] The present invention relates to the following reinforcing fiber bundle analysis method and analysis device.

[0012] The method for analyzing a reinforcing fiber bundle of the present invention is a method for analyzing a reinforcing fiber bundle containing a plurality of reinforcing fibers and a treatment agent, and includes the steps of: obtaining a tomographic image parallel to the longitudinal direction of the reinforcing fibers of the reinforcing fiber bundle; calculating a differential value of a gray value with respect to a position in the longitudinal direction of the reinforcing fibers for the tomographic image, and obtaining a differential image from the calculated differential value and the corresponding position; setting a region expanded in the longitudinal direction of the reinforcing fibers from a position in the differential image corresponding to a differential value equal to or greater than a certain threshold value as a starting point, thereby obtaining a mask image for extracting an analysis region including the treatment agent; and clustering the analysis region extracted by the mask image from the tomographic image.

[0013] The reinforcing fiber bundle analysis device of the present invention is an analysis device for a reinforcing fiber bundle containing a plurality of reinforcing fibers and a treatment agent, and includes an image acquisition unit that obtains a tomographic image parallel to the longitudinal direction of the reinforcing fibers of the reinforcing fiber bundle, a differential processing unit that calculates a differential value of a gray value with respect to a position in the longitudinal direction of the reinforcing fibers for the tomographic image and obtains a differential image from the calculated differential value and the corresponding position, a mask image generation unit that sets a region expanded in the longitudinal direction of the reinforcing fibers from a position in the differential image corresponding to a differential value equal to or greater than a certain threshold value as a starting point, and obtains a mask image for extracting an analysis region including the treatment agent, and a clustering unit that clusters the analysis region extracted by the mask image from the tomographic image. Effect of the Invention

[0014] According to the present invention, even if the boundary between the reinforcing fiber area and the bundling agent area is unclear in a tomographic image, the reinforcing fiber area and the bundling agent area can be accurately separated, and a method and apparatus for analyzing a reinforcing fiber bundle can be provided that can analyze the distribution state of the treatment agent inside the reinforcing fiber bundle in a short time with good reproducibility. [Brief description of the drawings]

[0015] [Figure 1] FIG. 1A is a partially enlarged view of a tomographic image, and FIG. 1B is a view showing a differential image of the tomographic image of FIG. 1A. [Diagram 2] FIG. 2A is a diagram showing a mask image obtained from the differential image of FIG. 1B, and FIG. 2B is a diagram showing a composite image obtained by superimposing the differential image and the mask image. [Diagram 3] 3A to 3C are diagrams showing images obtained by clustering the analysis region extracted by the mask image. [Figure 4] FIG. 4A is a graph showing the gray value of each pixel on line BB of the tomographic image on the left, FIG. 4B is a diagram showing a differential image based on the differential value data of the tomographic image of FIG. 4A, and FIG. 4C is a diagram showing a differential image based on the absolute value of the differential value data of the tomographic image. [Diagram 5] FIG. 5A is a diagram showing the region of the converging agent obtained by clustering, and FIG. 5B is a diagram showing the region of the converging agent of FIG. 5A superimposed on a tomographic image. [Figure 6] FIG. 6 is a block diagram of a control system of the reinforcing fiber bundle analysis device according to this embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In this embodiment, an example will be described in which the distribution state of a sizing agent is analyzed in a reinforcing fiber bundle containing a sizing agent as a treatment agent.

[0017] In the past studies, the inventors discovered a method for separating the reinforcing fiber region from the sizing agent region by converting a tomographic image into a differential image showing the gradient of the gray value and identifying the boundary based on the maximum or minimum value of the differential value (of the gray value) in the differential image. However, when identifying the sizing agent region based on the boundary, it was necessary to manually set the starting point of the region (the starting point for extracting the location of the sizing agent to be analyzed), which required time for analysis.

[0018] In contrast, in the present invention, a mask image is generated based on a differential image of a tomographic image parallel to the length direction of the reinforcing fiber, and the mask image and the tomographic image are superimposed to automatically extract the region to be analyzed (the region including the sizing agent). Therefore, it is possible to perform an analysis with good reproducibility in a short time while accurately separating the reinforcing fiber region from the sizing agent region. The present invention will be described below.

[0019] 1. Analysis method for reinforcing fiber bundles 1 to 3 are diagrams showing the flow of the analysis method of the reinforcing fiber bundle according to the present embodiment. Among them, FIG. 1A is a partially enlarged view of a tomographic image of the reinforcing fiber bundle, FIG. 1B is a diagram showing a differential image of the tomographic image of FIG. 1A, FIG. 2A is a diagram showing a mask image obtained from the differential image of FIG. 1B, and FIG. 2B is a diagram showing a composite image obtained by superimposing the differential image and the mask image. FIG. 3 is an image obtained by clustering the analysis region extracted by the mask image, FIG. 3A is a diagram showing the region (distribution) of the sizing agent, FIG. 3B is a diagram showing the region (distribution) of the reinforcing fiber, and FIG. 3C is a diagram showing the background. In addition, in FIGS. 1A to 3A, a partially enlarged view of a part of the region of the image on the left (a region corresponding to region A in FIG. 1A) is also shown on the right side.

[0020] That is, the method for analyzing a reinforcing fiber bundle according to this embodiment includes: 1) obtaining a tomographic image of the reinforcing fiber bundle (see FIG. 1A); 2) obtaining a differential image of the tomographic image (see FIG. 1B); 3) obtaining a mask image for extracting an analysis region based on the differential image (see FIG. 2A); and 4) clustering the analysis region extracted by the mask image from the tomographic image (see FIGS. 2B and 3A-C).

[0021] 1) About the process of obtaining tomographic images First, a cross-sectional image of the reinforcing fiber bundle is obtained (see FIG. 1A). Specifically, a cross-sectional image along the length direction of the reinforcing fiber (or reinforcing fiber bundle), preferably a cross-sectional image parallel to the length direction of the reinforcing fiber, is obtained. The length direction of the reinforcing fiber is the Y direction in FIG. 1A (see FIG. 1A).

[0022] A tomographic image is displayed as a pixel value (shade, brightness). Since a tomographic image is usually a grayscale image, it is displayed as an 8-bit 256-level gray value (pixel value) from pure black (0) to pure white (255).

[0023] In such a tomographic image of a reinforcing fiber bundle, although it depends on the type of reinforcing fiber, for example, in the case of carbon fiber, the following relationship is established: gray value of pixels constituting the reinforcing fiber region (white region)> gray value of pixels constituting the sizing agent region (gray region)> gray value of pixels constituting the void (air void) region and background (dark gray region to black region). Note that the gray value of pixels constituting the sizing agent region and the gray value of pixels constituting the reinforcing fiber region may partially overlap each other.

[0024] The tomographic image can be obtained by any method, for example, 3D-TEM observation, FIB-SEM observation (Focused Ion Beam), X-ray CT observation, etc. In this embodiment, it is preferable to obtain it by X-ray CT observation from the viewpoint of relatively easily grasping the internal state of the reinforcing fiber bundle, and in particular, compared with 3D-TEM, etc., from the viewpoint of non-destructive observation of the sample and observation over a wide field of view (mm order compared with μm order such as TEM). X-ray CT observation can be performed using, for example, a 3D micro X-ray CT imaging device.

[0025] Specifically, the X-ray CT observation can be performed by the following procedure. Although it depends on the type of reinforcing fiber bundle, for example, as a pretreatment, a sample can be prepared by the following procedure. 1) Cut the reinforcing fiber bundle to a length of approximately 2 cm. 2) The obtained sample is fixed to the splint with double-sided tape. 3) Place the above item 2) on the fixed jig (notch jig).

[0026] Next, the obtained sample is subjected to X-ray CT observation at the highest resolution (270 to 325 nm / pixel) using, for example, a 3D micro X-ray CT imaging device. The observation conditions may be as follows. <Observation conditions> Equipment used: Rigaku Corporation, nano3DX Measurement conditions: Target Cu, Lens 0270, Binning 1 Scanning angle: 1000 points Accumulation time at each angle: 45 seconds FOV: 0.9mmφ×0.6mm Pixel size: 0.26μm / voxel In addition, since the cut end of the sample is damaged, it is removed from the field of view (FOV) during CT observation, and the part approximately 1 cm from the end is included in the field of view (FOV) during CT observation.

[0027] The tomographic image may be an original image obtained by X-ray CT observation or the like, or may be an image obtained by constructing a three-dimensional image based on the obtained original image and then reconstructing the cross section of the reinforcing fiber in the longitudinal direction.

[0028] When reconstructing, the sliding (division) of the 3D image can be performed using the integrated mesh function of the observation device used to obtain the 3D image. This allows multiple tomographic images to be obtained. The number of divisions is not particularly limited, but in CT measurements at high resolution where the convergence agent can be observed, a smaller number of divisions is preferable from the viewpoint of improving the accuracy of the analysis.

[0029] Furthermore, as a cross-sectional image obtained by a method other than the above-mentioned X-ray CT observation, an observation image obtained by solidifying a sample of a reinforcing fiber bundle, creating a cross-section of the sample (along the length direction of the reinforcing fibers) by a method such as mechanical polishing, ion milling (CP), or microtome, and observing the cross-section by FIB-SEM observation or the like may be used.

[0030] Furthermore, in order to easily obtain a highly accurate differential image, the tomographic image may be subjected to noise removal processing (step 6) described below) or normalization as necessary.

[0031] 2) Process for obtaining differential images The obtained tomographic image is subjected to a differential process to obtain a differential image (see FIG. 1B).

[0032] The differentiation process is an operation for obtaining the gradient (slope) of the gray value (pixel value) in the vicinity of the pixel of interest. In the present invention, the differential value of the gray value with respect to the position in the longitudinal direction of the reinforcing fiber is calculated, and a differential image is obtained from the calculated differential value and the corresponding position.

[0033] Fig. 4A is a graph showing the gray value of each pixel on the line BB in the longitudinal direction of the reinforcing fiber in the tomographic image shown on the left. In Fig. 4A, the vertical and horizontal axes on the left both indicate the position on the line BB, and the vertical axis on the right indicates the gray value. Fig. 4B is a differential image based on the differential value data of the tomographic image of Fig. 4A (differential image before taking the absolute value), and Fig. 4C is a diagram showing a differential image based on the absolute value of the differential value data of the tomographic image (differential image after taking the absolute value).

[0034] That is, as shown in FIG. 4A, there is a large gray value difference (a position where the change in gray value is large relative to the position in the longitudinal direction of the reinforcing fiber) between the convergence agent and the voids (the boundary portion between the convergence agent and the voids) in the tomographic image. That is, by identifying the region where the change in gray value is large, it is possible to identify the boundary portion between the convergence agent (to be analyzed) and the voids or the reinforcing fiber, and further, the starting point of the convergence agent region (a position corresponding to a differential value equal to or greater than a certain threshold value) in step 3) described below. Therefore, in this step, the change in gray value relative to the position in the longitudinal direction of the reinforcing fiber, i.e., the differential value, is calculated.

[0035] The differentiation process can be performed using a differential filter in the longitudinal direction of the reinforcing fibers. The type of such a differential filter is not particularly limited, but for example, a Sobel filter can be used.

[0036] The differential image may be a differential image based on the calculated differential value itself (see FIG. 4B and the above-mentioned FIG. 1B), or may be a differential image based on the absolute value of the calculated differential value (see FIG. 4C). In a differential image based on the calculated differential value itself, even if the absolute value of the differential value is the same, positive and negative differential values ​​are displayed with different contrasts (for example, in FIG. 4B, only areas corresponding to negative differential values ​​equal to or greater than a certain threshold are displayed in white), and therefore, the area corresponding to the boundary portion between the sizing agent and the voids is extracted in small amounts.

[0037] In contrast, in an image based on the absolute value of the calculated differential value, if the absolute value is the same, it is displayed with the same contrast (for example, in FIG. 4C, both the area corresponding to a negative differential value equal to or greater than a certain threshold and the area corresponding to a positive differential value equal to or greater than a certain threshold are displayed in white), so it is possible to extract many areas corresponding to the boundary between the convergence agent and the voids. Therefore, it is easy to obtain a highly accurate mask image.

[0038] 3) About the process of obtaining the mask image Based on the obtained differential image, a mask image is obtained (see FIG. 2A).

[0039] The mask image is a filter for extracting an analysis region. In this embodiment, since one of the analysis objects is the convergence agent, an analysis region including the convergence agent is extracted. Specifically, in the differential image obtained in the above step 2), a position where the differential value is equal to or greater than a certain threshold (the region displayed in white in Figs. 4B and C) is used as a starting point, and the region is expanded in the length direction of the reinforcing fiber to define a non-masked region (analysis region). It is preferable that the expansion direction is in both directions along the length of the reinforcing fiber. In Fig. 2A, the black filled area is the masked area (non-analysis region), and the unfilled area is the non-masked area (analysis region).

[0040] Specifically, the differential value being equal to or greater than a certain threshold value means that, when the absolute value of the differential value is converted to a value from 0 to 255, the absolute value of the differential value is preferably equal to or greater than 16, more preferably equal to or greater than 30, and even more preferably equal to or greater than 32. The differential value being equal to or greater than a certain threshold value may be a maximum or minimum value of the differential value (i.e., a maximum value of the absolute value of the differential value). The position corresponding to the differential value being equal to or greater than a certain threshold value may correspond to the boundary between the sizing agent and voids, or the boundary between the sizing agent and reinforcing fibers.

[0041] The mask image can be obtained by any method. For example, the mask image can be obtained by blurring the differential value data of the differential image (data of the differential value) in the length direction of the reinforcing fiber, starting from a point where the absolute value of the differential value is equal to or greater than a certain threshold value.

[0042] 4) Clustering process Then, the analysis region extracted from the tomographic image by the mask image is subjected to clustering (see FIGS. 2B and 3A to 3C).

[0043] The analysis region can be extracted as the region not covered by the mask image (non-masked region) when the mask image obtained in 3) above is superimposed on the tomographic image obtained in 1) above (see Figure 2B).

[0044] Then, the tomographic images are clustered in the extracted analysis region. Specifically, the tomographic images in the extracted analysis region are classified into two or more regions, preferably three or more regions, based on the gray value. The two or more regions include the region of the sizing agent (to be analyzed) (see FIG. 3A). That is, in this embodiment, the region can be classified into a background region (see FIG. 3C), a sizing agent region (see FIG. 3A), and other regions (see FIG. 3B). Alternatively, the region may be classified into three or more regions including the sizing agent region, the void region, and the reinforcing fiber region.

[0045] Clustering can be performed by any method, for example, by classifying each region based on the gray value (of the tomographic image) in the analysis region, the distance between observation points (pairwise distance), etc. Examples of such clustering methods include Gaussian mixture models, hierarchical clustering, k-means clustering, k-medoids clustering, DBSCAN (Density-Based Spatial Clustering of Applications with Noise), nearest neighbor, and spectral clustering (data division using spectral clustering).

[0046] Among them, a method of classifying each region based on the gray value (of a tomographic image) in the analysis region is preferable, and a Gaussian mixture model, k-means, and spectral clustering are more preferable, and a Gaussian mixture model is even more preferable. The Gaussian mixture model is an analysis method that uses an algorithm to separate the distribution of the original set by regarding it as a "collection of Gaussian distributions," and is particularly suitable for clustering values ​​whose values ​​themselves can be regarded as a Gaussian distribution (or a distribution close to it), such as the distribution of gray values ​​in a tomographic image (such as a SEM image).

[0047] FIG. 5A is a diagram showing the region of the converging agent obtained by clustering, and FIG. 5B is a diagram showing the region of the converging agent of FIG. 5A superimposed on a tomographic image.

[0048] The convergence agent regions obtained by clustering can be displayed alone (see FIG. 5A) or overlaid on the original tomographic image (see FIG. 5B). By overlaying the regions on the original tomographic image, the distribution of the convergence agent can be grasped in more detail.

[0049] The method for analyzing a reinforcing fiber bundle according to the present embodiment may further include other steps as necessary. For example, in the above step 1), a plurality of cross-sectional images (along the length direction of the reinforcing fibers) may be obtained, and after the above step 4), a step 5) of obtaining a three-dimensional image of the sizing agent may be further performed.

[0050] 5) The process of obtaining 3D images The identified areas of the converging agent for each of the multiple tomographic images may be combined to obtain a three-dimensional image of the converging agent.

[0051] Specifically, multiple tomographic images can be synthesized by carrying out steps 2) to 4) above for each of the multiple tomographic images, identifying the areas of reinforcing fibers, bundling agents, and voids, and adding them together (i.e., stacking them in the slice direction).

[0052] As another step, between steps 1) and 2), a step 6) of removing noise from the tomographic image may be further carried out, and a step 8) of obtaining a distribution of particle volume of the converging agent based on data such as the three-dimensional image obtained in step 5) may be further carried out.

[0053] Regarding step 6 (noise removal process) Noise removal (default setting) specifically refers to a process that amplifies the gray value gap at the boundary between the reinforcing fiber and the converging agent in the tomographic image. Noise removal can be performed, for example, by a Non-Local Means filter.

[0054] 2. Reinforcement fiber bundle analysis device FIG. 6 is a block diagram of a control system of the reinforcing fiber bundle analysis device 100 according to this embodiment.

[0055] As shown in FIG. 6, analysis device 100 has an imaging section 110, a processing section 120, and a display section .

[0056] The imaging unit 110 is not particularly limited, and can be an imaging unit such as an X-ray CT observation device.

[0057] The processing unit 120 acquires, processes, and analyzes image information (tomographic images) captured by the imaging unit 110. Specifically, the processing unit 120 has an image acquisition unit 121, a differential processing unit 122, a mask image generation unit 123, and a clustering unit .

[0058] The image acquiring unit 121 acquires image information (multiple tomographic images constituting a three-dimensional image) captured by the imaging unit 110, specifically, tomographic images parallel to the length direction of the reinforcing fiber. If necessary, the image acquiring unit 121 may construct a three-dimensional image from the acquired image information, and then slide (divide) the reinforcing fiber in the length direction to reconstruct a cross section in any direction to obtain a tomographic image.

[0059] The differential processing unit 122 obtains a differential image by performing differential processing on at least some (or each) of the multiple tomographic images obtained by the image acquisition unit 121 in the length direction of the reinforcing fiber. Specifically, for the tomographic image, a differential value of the gray value with respect to the position in the length direction of the reinforcing fiber is calculated, and a differential image is obtained from the calculated differential value and the position corresponding to it.

[0060] The mask image generating unit 123 generates a mask image based on the differential image obtained by the differential processing unit 122. Specifically, a region expanded in the length direction of the reinforcing fiber is set starting from a position corresponding to a differential value equal to or greater than a certain threshold in the differential image, and a mask image for extracting an analysis region including the treatment agent is obtained.

[0061] The clustering unit 124 performs clustering on the tomographic image for the analysis region extracted by the mask image. Specifically, the clustering unit 124 superimposes the tomographic image and the mask image, and performs clustering on the analysis region extracted by the mask image. The clustering is preferably performed by a Gaussian mixture model.

[0062] The processing unit 120 may further include, as necessary, a three-dimensional image creation unit (not shown) and a histogram creation unit (not shown) that analyzes the volume of each particle of the sizing agent based on the three-dimensional image data obtained by the three-dimensional image creation unit and creates a histogram of the particle volume of the sizing agent contained in the reinforcing fiber bundle.

[0063] The three-dimensional image creation unit synthesizes (stacks in the slice direction) the information obtained by the clustering unit 124 for a plurality of tomographic images, for example, to three-dimensionalize at least one of the reinforcing fiber region, the bundling agent region, and the void region. This makes it possible to obtain a three-dimensional image of at least one of the reinforcing fiber region, the bundling agent region, and the void region.

[0064] The processing unit 120 may be realized by software using a processing device such as a CPU (Central Processing Unit), or may be realized by hardware such as an IC (Integrated Circuit), or may be realized by a combination of software and hardware.

[0065] Display unit 130 is, for example, a display screen of an image display device such as a personal computer, and displays the two-dimensional image (data) obtained by clustering unit 124 and the three-dimensional image (data) obtained by a three-dimensional image creation unit (not shown).

[0066] (Modification) In the above embodiment, in step 3) (step of obtaining a mask image), an example is shown in which the mask image is generated based on a differential image based on the absolute value of the differential value. However, this is not limited to this, and the mask image may be generated based on a differential image based on a differential value.

[0067] In the above embodiment, the reinforcing fiber bundle contains a sizing agent as a treatment agent, but the present invention is not limited to this. For example, the present invention can be applied to cases where the treatment agent contains a texture improving agent such as a silicone-based agent, various water repellents, textile oils, water-absorbing agents, antibacterial agents, etc.

[0068] In addition, in the above embodiment, an example was shown in which a tomographic image was used as the original image, but the present invention is not limited to this, and a surface image (an observation image such as an SEM image or a two-dimensional image along the length direction of the reinforcing fiber) may also be used. In other words, the analysis method of the present invention can be applied not only to the internal observation of a reinforcing fiber bundle, but also to the surface observation.

[0069] 3. Reinforced fiber bundles The method for analyzing a reinforcing fiber bundle according to the present embodiment can be applied to the analysis of an object (e.g., a reinforcing fiber bundle or a prepreg) that contains a fiber component and a resin component and has a low content of the resin component relative to the fiber component, preferably an arbitrary reinforcing fiber bundle. A reinforcing fiber bundle usually contains a plurality of reinforcing fibers and a sizing agent as a treatment agent.

[0070] (Reinforced Fiber) Examples of reinforcing fibers include glass fibers, carbon fibers, graphite fibers, aramid fibers, boron fibers, alumina fibers, and silicon carbide fibers. Among these, carbon fibers and graphite fibers are preferred because they are easy to use to obtain a lightweight, highly durable molded product.

[0071] From the viewpoint of impact resistance, the carbon fiber preferably has a tensile modulus of 400 GPa or more, and from the viewpoint of rigidity and mechanical strength, the tensile strength is preferably 4.4 to 6.5 GPa. The tensile elongation of the carbon fiber is preferably 1.7 to 2.3%. Therefore, the carbon fiber having a tensile modulus of at least 230 GPa, a tensile strength of at least 4.4 GPa, and a tensile elongation of at least 1.7% is most preferable.

[0072] Examples of commercially available carbon fibers include Torayca (registered trademark) T800G-24K, Torayca (registered trademark) T800S-24K, Torayca (registered trademark) T700G-24K, Torayca (registered trademark) T300-3K, and Torayca (registered trademark) T700S-12K (all manufactured by Toray Industries, Inc.).

[0073] The average fiber diameter (diameter of the fiber cross section) of the reinforcing fibers can be, for example, 1 to 20 μm, and preferably 4 to 15 μm, from the viewpoints of moldability and mechanical strength.

[0074] The average fiber diameter and the average fiber length of the reinforcing fibers can be calculated by measuring the fiber diameter and fiber length of 10 arbitrarily selected reinforcing fibers with an electron microscope and taking the average value. The fiber diameter of the reinforcing fibers is the maximum Feret diameter of the cross section of a single fiber.

[0075] (Convergence agent) The type of sizing agent is not particularly limited, and for example, when the reinforcing fiber is carbon fiber, from the viewpoints of good affinity therewith and ease of increasing the mechanical strength of the molded body, water-dispersible resins such as urethane-based resins, epoxy-based resins, polyamide-based resins, olefin-based resins, acrylic-based resins, butadiene-based resins, etc. are preferred. Among these, urethane-based resins and epoxy-based resins are preferred, and epoxy-based resins are more preferred from the viewpoint of further improving the mechanical strength.

[0076] The content of these resins may be 1 to 20% by mass based on the total amount of the sizing agent.

[0077] The sizing agent may further contain additives other than those described above, if necessary. Examples of the additives include a surfactant, a leveling agent, an emulsifier, etc.

[0078] The amount of the sizing agent attached is, for example, 0.1 to 10 parts by mass, preferably 0.2 to 3 parts by mass, relative to 100 parts by mass of the reinforcing fibers. Thus, the amount of the sizing agent attached is very small relative to the total amount of the reinforcing fibers constituting the reinforcing fiber bundle, and voids are likely to be present between the reinforcing fibers. The present invention is particularly effective for such an analysis target.

[0079] The reinforcing fiber bundle may be obtained by impregnating a plurality of reinforcing fibers with a sizing agent liquid. The sizing agent liquid may be an emulsion containing the water-dispersible resin. The sizing agent liquid may further contain the additives described above in addition to the water-dispersible resin as the main component. [Industrial Applicability]

[0080] According to the present invention, it is possible to provide a method and apparatus for analyzing reinforcing fiber bundles that can accurately separate reinforcing fiber areas from bundling agent areas in a tomographic image and can analyze the distribution state of a treatment agent inside a reinforcing fiber bundle in a short period of time with good reproducibility. [Explanation of symbols]

[0081] 100 Analyzer 110 Imaging unit 120 Processing section 121 Image acquisition unit 122 Differential Processing Section 123 Mask Image Generation Unit 124 Clustering Department 130 Display section

Claims

1. A method for analyzing a reinforcing fiber bundle including a plurality of reinforcing fibers and a treatment agent, comprising: Obtaining a tomographic image parallel to the length direction of the reinforcing fibers of the reinforcing fiber bundle; A step of calculating a differential value of the gray value with respect to a position in the longitudinal direction of the reinforcing fiber for the tomographic image, and obtaining a differential image from the calculated differential value and the corresponding position; a step of setting an area expanded in a longitudinal direction of the reinforcing fiber from a position corresponding to a differential value equal to or greater than a certain threshold in the differential image as a starting point, and obtaining a mask image for extracting an analysis area including the treatment agent; clustering an analysis region extracted by the mask image from the tomographic image; Including, Analysis method for reinforcing fiber bundles.

2. The mask image is obtained by blurring the differential value data in the length direction of the reinforcing fiber. The method for analyzing a reinforcing fiber bundle according to claim 1.

3. The position corresponding to the differential value equal to or greater than the certain threshold corresponds to the boundary between the treatment agent and the void or the reinforcing fiber. The method for analyzing a reinforcing fiber bundle according to claim 1 or 2.

4. The clustering is performed by classifying the tomographic images in the analysis region into three or more regions. The method for analyzing a reinforcing fiber bundle according to any one of claims 1 to 3.

5. The clustering classifies each area based on the gray value in the analysis area. The method for analyzing a reinforcing fiber bundle according to any one of claims 1 to 4.

6. The clustering is performed by analysis using a Gaussian mixture model. The method for analyzing a reinforcing fiber bundle according to any one of claims 1 to 5.

7. Identifying a region of the treatment agent by the clustering. The method for analyzing a reinforcing fiber bundle according to any one of claims 1 to 6.

8. The clustering further identifies the reinforcing fiber regions and void regions. The method for analyzing a reinforcing fiber bundle according to claim 7.

9. In the step of obtaining a tomographic image, a plurality of tomographic images are obtained, and further comprising a step of synthesizing the areas of the treatment agent identified for each of the plurality of tomographic images to obtain a three-dimensional image of the treatment agent. The method for analyzing a reinforcing fiber bundle according to claim 7.

10. The treatment agent is a sizing agent. The method for analyzing a reinforcing fiber bundle according to any one of claims 1 to 9.

11. The reinforcing fiber is a carbon fiber. The method for analyzing a reinforcing fiber bundle according to any one of claims 1 to 10.

12. An analysis device for a reinforcing fiber bundle including a plurality of reinforcing fibers and a treatment agent, An image acquisition unit for acquiring a tomographic image parallel to the longitudinal direction of the reinforcing fibers of the reinforcing fiber bundle; a differential processing unit that calculates a differential value of a gray value with respect to a position in a length direction of the reinforcing fiber for the tomographic image and obtains a differential image from the calculated differential value and the corresponding position; a mask image generating unit that sets an area expanded in a longitudinal direction of the reinforcing fiber from a position corresponding to a differential value equal to or greater than a certain threshold value in the differential image as a starting point, and obtains a mask image for extracting an analysis area including the treatment agent; a clustering unit that performs clustering on an analysis region extracted by the mask image from the tomographic image; Including, Analysis device for reinforcing fiber bundles.

13. The image acquisition unit acquires a plurality of the tomographic images, the analysis device further includes a three-dimensional image creation unit that synthesizes the areas of the treatment agent identified by the clustering for each of the plurality of tomographic images to obtain a three-dimensional image of the treatment agent. The reinforcing fiber bundle analysis device according to claim 12.

Images