WO2023181935A1

WO2023181935A1 - Prediction device, prediction system, and prediction program

Info

Publication number: WO2023181935A1
Application number: PCT/JP2023/008748
Authority: WO
Inventors: 茂小島; 友香子 ▲高▼; 一磨小田; みゆき岡庭
Original assignee: コニカミノルタ株式会社
Priority date: 2022-03-24
Filing date: 2023-03-08
Publication date: 2023-09-28

Abstract

Provided are a prediction device, a prediction system, and a prediction program capable of predicting a residual stress of a fiber composite material.　The prediction device 100 comprises: an acquisition unit 111 that acquires first measurement information obtained by measuring material properties of a fiber composite material with a first measurement device, and second measurement information obtained by measuring the material properties of the fiber composite material with a second measurement device different from the first measurement device; and a prediction unit 113 that predicts a residual stress of the fiber composite material, on the basis of the acquired first measurement information and second measurement information.

Description

Prediction device, prediction system and prediction program

The present invention relates to a prediction device, a prediction system, and a prediction program.

In recent years, lightweight members such as fiber composite materials have been attracting attention in order to popularize electric vehicles and put flying cars into practical use (for example, Patent Document 1). A fiber composite material containing resin and fibers has a three-dimensional structure due to fiber orientation and the like, and its characteristics are easily influenced by fiber orientation, fiber density, and the like.

One of the important properties of fiber composite materials is residual stress. This is because residual stress can affect the occurrence of cracks during long-term use of the fiber composite material.

JP 2021-89195 Publication

It is desired to predict the residual stress of such fiber composite materials.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a prediction device, a prediction system, and a prediction program that can predict the residual stress of a fiber composite material.

The above object of the present invention is achieved by the following means.

(1) First measurement information obtained by measuring the material properties of the fiber composite material using a first measuring device; and second measurement information obtained by measuring the material properties of the fiber composite material by a second measuring device different from the first measuring device. and a prediction unit that predicts residual stress of the fiber composite material based on the acquired first measurement information and second measurement information.

(2) The prediction device according to (1) above, further including a control unit that causes an output unit to output information regarding the predicted residual stress.

(3) The prediction device according to (1) or (2) above, wherein the first measurement information includes information regarding fiber orientation of the fiber composite material.

(4) The prediction device according to (3) above, wherein the first measurement device is an X-ray Talbot-Lau device or an X-ray CT measurement device.

(5) The second measurement information includes information regarding at least one of the resin type, molecular orientation, crystallinity, specific volume, fiber amount, interaction between fibers and resin, and fiber length of the fiber composite material. The prediction device according to any one of (1) to (4) above.

(6) The second measurement device is an infrared spectrometer, an impedance spectrometer, a terahertz wave spectrometer, an ultrasonic measurement device, a Raman spectrometer, an X-ray diffraction device, or a fluorescence microscope (1) above. The prediction device according to any one of (5) to (5).

(7) The acquisition unit further acquires third measurement information measured by a third measurement device different from the first measurement device and the second measurement device, and the prediction unit , the prediction device according to any one of (1) to (6) above, which predicts the residual stress based on the second measurement information and the third measurement information.

(8) The prediction device according to any one of (1) to (7) above, wherein the fiber composite material contains a sensitivity regulator.

(9) The prediction device according to any one of (1) to (8), wherein the prediction unit predicts the residual stress using a learned discriminator.

(10) The prediction unit further includes an extraction unit that extracts feature quantities from each of the acquired first measurement information and second measurement information, and the prediction unit receives the extracted feature quantities as input and predicts the residual stress. The prediction device according to (9) above.

(11) The prediction device according to (10), wherein the discriminator performs machine learning using the feature amount as input data and the residual stress as output data.

(12) A first measuring device that measures the material properties of the fiber composite material; a second measuring device that measures the material properties of the fiber composite material and is different from the first measuring device; 11) A prediction system comprising the prediction device according to any one of item 11).

(13) First measurement information obtained by measuring the material properties of the fiber composite material using a first measuring device; and second measurement information obtained by measuring the material properties of the fiber composite material by a second measuring device different from the first measuring device. and (b) predicting the residual stress of the fiber composite material based on the first measurement information and the second measurement information obtained in step (a). A prediction program that allows a computer to perform processing.

A prediction device, a prediction system, and a prediction program according to the present invention acquire first measurement information and second measurement information that measure material properties of a fiber composite material, and based on the acquired first measurement information and second measurement information, , predicting residual stress in fiber composite materials. This makes it possible to predict the residual stress in the fiber composite material.

FIG. 1 is a diagram showing the overall configuration of a prediction system. FIG. 2 is a block diagram showing a schematic configuration of a prediction device. FIG. 1 is an overall schematic diagram of an X-ray Talbot-Low apparatus. FIG. 2 is a diagram explaining the principle of a Talbot interferometer. FIG. 3 is a schematic plan view of a source grating, a first grating, and a second grating. FIG. 2 is a block diagram showing the functional configuration of a prediction device. It is a figure which shows an example of the display form of the information output by a prediction device. It is a flowchart which shows the procedure of the prediction process performed in a prediction device. 3 is a flowchart showing a machine learning method for a trained model.

Hereinafter, embodiments of the present invention will be described with reference to the attached drawings. In addition, in the description of the drawings, the same elements are given the same reference numerals, and redundant description will be omitted. Furthermore, the dimensional ratios in the drawings are exaggerated for convenience of explanation and may differ from the actual ratios.

<Prediction system configuration>
FIG. 1 is a diagram showing the overall configuration of a prediction system.

As shown in FIG. 1, the prediction system includes, for example, a prediction device 100, a first measurement device 200, and a second measurement device 300. This prediction system predicts residual stress in fiber composite materials. Residual stress is stress that exists within a material.

Fiber composite materials are used as components of a variety of products, including space and aircraft-related products, automobiles, ships, fishing rods, electrical/electronic/home appliance parts, parabolic antennas, bathtubs, flooring materials, roofing materials, etc. It will be done. Examples of fiber composite materials include CFRP (Carbon-Fiber-Reinforced Plastics) using carbon fiber or glass fiber as reinforcing fibers, and CFRTP (Carbon Fiber Reinforced Thermo Plastics). reinforced thermoplastic) and Examples include FRP (Fiber-Reinforced Plastics) represented by GFRP (Glass-Fiber-Reinforced Plastics). In particular, CFRTP is excellent in terms of lightweight and recyclability.

The resin contained in the fiber composite material is, for example, plastic such as general-purpose plastic, engineering plastic, and super engineering plastic, but is not limited to these. For example, plastics include polypropylene, polyamide, ABS, polycarbonate resin, nylon, polyphenylene sulfide (PPS), polyacetal (POM), and the like. The fibers included in the fiber composite material are, for example, carbon fibers, glass fibers, or aramid fibers.

The fiber composite material may contain a sensitivity modifier. A sensitivity adjustment agent refers to a sample that functions like a contrast agent used in X-ray photography and enables measurements of fiber composite materials with higher accuracy and sensitivity. By measuring the fiber composite material containing the sensitivity modifier using the first measuring device 200 and the second measuring device 300, it becomes possible to perform the measurement with higher accuracy. For example, when the second measuring device 300 is a Raman spectrometer, using zirconium tungstate as the sensitivity modifier changes the Raman shift, making it possible to generate information regarding the material properties of the fiber composite material with higher accuracy. It becomes possible. For example, when the second measuring device 300 is a fluorescence microscope, if a fluorescent dye is used as the sensitivity adjusting agent, it becomes possible to generate information regarding the fiber length with higher accuracy.

It is preferable that the sensitivity modifier contained in the fiber composite material has a small effect on the physical properties of the fiber composite material. Thereby, the fiber composite material measured by the first measuring device 200 and the second measuring device 300 can be used for molded products. For measurement with the first measuring device 200 and the second measuring device 300, a test piece of a fiber composite material containing a sensitivity modifier may be created.

(Prediction device 100)
The prediction device 100 is a computer such as a PC, a smartphone, or a tablet terminal, and functions as a prediction device in this embodiment. The prediction device 100 is configured to be connectable to the first measurement device 200 and the second measurement device 300, and transmits and receives various information to and from each device.

FIG. 2 is a block diagram showing a schematic configuration of the information processing device.

As shown in FIG. 2, the prediction device 100 includes a CPU (Central Processing Unit) 110, a ROM (Read Only Memory) 120, a RAM (Random Access Memory) 130, a storage 140, and a communication interface 150. , display section 160, and operation reception 170. Each configuration is communicably connected to each other via a bus.

The CPU 110 controls each of the above components and performs various calculation processes according to programs recorded in the ROM 120 and the storage 140.

The ROM 120 stores various programs and various data.

The RAM 130 temporarily stores programs and data as a work area.

The storage 140 stores various programs including an operating system and various data. For example, the storage 140 is installed with an application for predicting residual stress in a fiber composite material using a learned discriminator. Moreover, the first measurement information and the second measurement information acquired from the first measurement device 200 and the second measurement device 300 may be stored in the storage 140. Furthermore, the storage 140 may store trained models used as classifiers and teacher data used for machine learning.

The communication interface 150 is an interface for communicating with other devices. As the communication interface 150, a wired or wireless communication interface according to various standards is used. The communication interface 150 is used, for example, to receive first measurement information and second measurement information from the first measurement device 200 and the second measurement device 300, and to send prediction results to a server or the like for storage. It will be done.

The display unit 160 includes an LCD (liquid crystal display), an organic EL display, etc., and displays various information. The display unit 160 may be configured by viewer software, a printer, or the like. In this embodiment, the display section 160 functions as an output section.

The operation reception unit 170 includes a touch sensor, a pointing device such as a mouse, a keyboard, etc., and accepts various operations from the user. Note that the display section 160 and the operation reception section 170 may constitute a touch panel by superimposing a touch sensor as the operation reception section 170 on a display surface as the display section 160.

(First measuring device 200)
The first measurement device 200 is a device for measuring material properties of a fiber composite material and acquiring first measurement information. Here, material properties are chemical and physical properties of fiber composite materials, such as acoustic properties, atomic properties, electrical properties, magnetic properties, mechanical properties, optical properties, radiation properties and thermal properties. Characteristics, etc. Preferably, the first measurement information includes information regarding fiber orientation of the fiber composite material. This is because fiber orientation is highly likely to affect residual stress.

The first measurement device 200 is preferably an X-ray Talbot-Low device or an X-ray CT (Computed Tomography) measurement device. This is because such a first measurement device 200 can generate first measurement information including information regarding fiber orientation. In particular, the first measuring device 200 is preferably an X-ray Talbot-Low device. The X-ray Talbot-Low apparatus makes it possible to obtain information on a wide range of fiber composite materials in a short time. The first measuring device 200 may be a transmission type X-ray device or an X-ray diffraction device.

The X-ray Talbot-Lau apparatus uses, for example, a Talbot-Lau interferometer that includes a source grating (the source grating 12 in FIG. 3, which will be described later, is also referred to as a multi-grating, multi-slit, G0 grating, etc.). . The X-ray Talbot-Lau apparatus does not need to have a source grating, for example, a first grating (first grating 14 in FIG. 3 described later, also referred to as G1 grating) and a second grating (described later in FIG. A Talbot interferometer including only the second grating 15 (also referred to as G2 grating) may be used.

FIG. 3 is a diagram showing an example of the configuration of the X-ray Talbot-Low apparatus 1. The X-ray Talbot-Lau apparatus 1 includes an X-ray generator 11, the above-described source grating 12, an object stage 13, the above-described first grating 14, the above-described second grating 15, and an X-ray detector 16. , a support 17 , and a base portion 18 .

According to such an X-ray Talbot-Lau apparatus 1, a moiré image Mo (FIG. 4) of the fiber composite material H located at a predetermined position with respect to the object stage 13 is photographed by a method based on the principle of the fringe scanning method, By analyzing the moire image Mo using a Fourier transform method, at least three types of images (two-dimensional images) can be reconstructed (referred to as reconstructed images). That is, an absorption image (same as a normal X-ray absorption image) that visualizes the average component of the moire fringe in the moire image Mo, a differential phase image that visualizes the phase information of the moire fringe, and a These are three types of small-angle scattering images. Note that it is also possible to generate more types of images by recombining these three types of reconstructed images.

In addition, the fringe scanning method means that one of multiple gratings is 1/M of the slit period of the grating (M is a positive integer, M>2 for absorption images, M>3 for differential phase images and small-angle scattering images). In this method, reconstruction is performed using a moiré image Mo taken M times by moving each moiré image Mo in the slit period direction to obtain a high-definition reconstructed image.

Furthermore, the Fourier transform method is a method in which one moire image Mo is photographed using the X-ray Talbot-Lau apparatus 1 in the presence of the fiber composite material H, and the moire image Mo is subjected to Fourier transform in image processing. This is a method of reconstructing and generating images such as differential phase images.

Here, first, the principle common to Talbot interferometers and Talbot-Lau interferometers will be explained using FIG. 4.

Although FIG. 4 shows the case of a Talbot interferometer, the explanation is basically the same for the case of a Talbot-Lau interferometer. Furthermore, the z direction in FIG. 4 corresponds to the vertical direction in the X-ray Talbot-Low apparatus 1 in FIG. 3, and the x and y directions in FIG. direction).

In addition, as shown in FIG. 5, the first grating 14 and the second grating 15 (also the source grating 12 in the case of Talbot-Lau interferometer) have an A plurality of slits S are arranged and formed at a predetermined period d in the direction. This arrangement of slits S is a one-dimensional lattice, and the arrangement of slits S in the x and y directions is a two-dimensional lattice.

The source grating 12, the first grating 14, and the second grating 15 are, for example, composed of one-dimensional gratings, but may also be composed of two-dimensional gratings.

As shown in FIG. 4, the X-rays irradiated from the X-ray source 11a (in the case of the Talbot-Lau interferometer, the X-rays irradiated from the X-ray source 11a are multiplied by the source grating 12 (not shown in FIG. 4). When the X-rays (which have been converted into a light source) pass through the first grating 14, the transmitted X-rays form images at regular intervals in the z direction. This image is called a self-image (also called a lattice image, etc.), and the phenomenon in which self-images are formed at regular intervals in the z direction is called the Talbot effect.

In other words, the Talbot effect means that when coherent light passes through the first grating 14 in which slits S are provided at a constant period d as shown in FIG. It refers to the phenomenon of forming one's self-image at regular intervals.

Then, as shown in FIG. 4, a second grating 15 provided with slits S like the first grating 14 is placed at a position where the self-image of the first grating 14 focuses. At that time, if the extending direction of the slits S of the second grating 15 (that is, the x-axis direction in FIG. 4) is arranged to be approximately parallel to the extending direction of the slit S of the first grating 14, the second A moire image Mo is obtained on the grid 15.

In addition, in FIG. 4, if the moire image Mo is written on the second lattice 15, the moire fringes and slits S will be mixed and it will be difficult to understand, so the moire image Mo is shown separated from the second lattice 15. There is. However, in reality, the moire image Mo is formed on the second grating 15 and on the downstream side thereof. This moire image Mo is then photographed by an X-ray detector 16 placed directly below the second grating 15.

Furthermore, as shown in FIGS. 3 and 4, when the fiber composite material H exists between the X-ray source 11a and the first grating 14, the phase of the X-rays is shifted by the fiber composite material H, so that the moire image Mo is Moiré fringes are disturbed along the edges of the fiber composite material H. On the other hand, although not shown, if the fiber composite material H does not exist between the X-ray source 11a and the first grating 14, a moire image Mo consisting of only moire fringes will appear. The above is the principle of Talbot interferometer and Talbot-Lau interferometer.

Based on this principle, the X-ray Talbot-Lau apparatus 1 has a second grating 15 in the second cover unit 180 at a position where the self-image of the first grating 14 focuses, as shown in FIG. 3, for example. It is set to be placed. Furthermore, as described above, if the second grating 15 and the X-ray detector 16 are separated, the moire image Mo (see FIG. 4) will become blurred, so the X-ray detector 16 should be placed directly below the second grating 15. is preferred. Alternatively, the second grating 15 may be made of a light emitting material such as a scintillator or amorphous selenium, and the second grating 15 and the X-ray detector 16 may be integrated.

The second cover unit 180 is used to protect the X-ray detector 16, etc. by preventing people or objects from colliding with or touching the first grating 14, the second grating 15, the X-ray detector 16, etc. It is set in.

Although not shown, in the X-ray detector 16, conversion elements that generate electrical signals according to irradiated X-rays are arranged in a two-dimensional form (matrix form), and the electrical signals generated by the conversion elements are converted into images. It is configured to be read as a signal. The X-ray detector 16 is configured to capture, for example, the moire image Mo, which is an X-ray image formed on the second grating 15, as an image signal for each conversion element. The pixel size of the X-ray detector 16 is 10 to 300 (μm), more preferably 50 to 200 (μm).

As the X-ray detector 16, an FPD (Flat Panel Detector) can be used. There are two types of FPDs: an indirect conversion type that converts detected X-rays into electrical signals via a photoelectric conversion element, and a direct conversion type that converts detected X-rays directly into electrical signals. Good too.

In the indirect conversion type, a photoelectric conversion element is arranged two-dimensionally with a TFT (thin film transistor) under a scintillator plate such as CsI or Gd2O2S to form each pixel. When the X-rays incident on the X-ray detector 16 are absorbed by the scintillator plate, the scintillator plate emits light. This emitted light causes charge to be accumulated in each photoelectric conversion element, and the accumulated charge is read out as an image signal.

In the direct conversion type, an amorphous selenium film with a film thickness of 100 to 1000 (μm) is formed on glass by thermal evaporation of amorphous selenium, and the amorphous selenium film and electrodes are placed on a two-dimensional array of TFTs. Deposited. When the amorphous selenium film absorbs X-rays, a voltage is liberated within the material in the form of electron-hole pairs, and the voltage signal between the electrodes is read by the TFT. Note that an imaging means such as a CCD (Charge Coupled Device) or an X-ray camera may be used as the X-ray detector 16.

The X-ray Talbot-Lau apparatus 1 is configured to take a plurality of moiré images Mo using, for example, a so-called fringe scanning method. That is, in this X-ray Talbot-Lau apparatus 1, the relative positions of the first grating 14 and the second grating 15 are set in the x-axis direction (that is, perpendicular to the extending direction (y-axis direction) of the slit S) in FIGS. A plurality of moiré images Mo are photographed while shifting the moiré images Mo in the direction of

For example, the X-ray Talbot-Lau apparatus 1 reconstructs (that is, image reconstruction) an absorption image, a differential phase image, a small-angle scattering image, etc. based on a plurality of captured moiré images Mo. ing. The reconstruction process may be performed by the prediction device 100.

The X-ray Talbot-Lau apparatus 1 is capable of moving the first grating 14 by a predetermined amount in the x-axis direction, for example, in order to capture a plurality of moiré images Mo using the fringe scanning method. Note that it is also possible to move the second grating 15 instead of moving the first grating 14, or to move both of them.

Further, the X-ray Talbot-Lau apparatus 1 captures only one moire image Mo while fixing the relative positions of the first grating 14 and the second grating 15, and other apparatuses (for example, an image processing apparatus or a prediction apparatus) In the image processing in 100), it is also possible to reconstruct an absorption image, a differential phase image, etc. by analyzing this moiré image Mo using a Fourier transform method or the like.

The configuration of other parts of the X-ray Talbot-Lau apparatus 1 will be explained. This X-ray Talbot-Lau apparatus 1 is of a so-called vertical type, and includes an X-ray generator 11, a source grating 12, an object stage 13, a first grating 14, a second grating 15, and an X-ray detector 16 in this order. is placed in the z direction, which is the direction of gravity. That is, here, the z direction is the irradiation direction of the X-rays from the X-ray generator 11.

The X-ray generator 11 includes, as an X-ray source 11a, a Coolidge X-ray source, a rotating anode X-ray source, etc., which are widely used in medical settings, for example. It is also possible to use other X-ray sources. The X-ray generator 11 is configured, for example, to emit X-rays from a focal point in the form of a cone beam. That is, as shown in FIG. 3, X-rays are irradiated with the X-ray irradiation axis Ca coinciding with the z direction as the central axis, and the X-rays are irradiated so as to spread further away from the X-ray generator 11 (that is, the X-ray irradiation range).

In the X-ray Talbot-Lau apparatus 1, for example, a source grating 12 is provided below the X-ray generator 11. In order to prevent vibrations of the X-ray generator 11 caused by rotation of the anode of the X-ray source 11a from being transmitted to the source grating 12, the source grating 12 is not attached to the X-ray generator 11 and is attached to a support. It is preferable to attach it to a fixing member 12a attached to a base part 18 provided at 17.

In the X-ray Talbot-Low apparatus 1, in order to prevent the vibrations of the X-ray generator 11 from propagating to other parts of the X-ray Talbot-Low apparatus 1, such as the column 17 (or to reduce the propagated vibrations), ), for example, a buffer member 17a is provided between the X-ray generator 11 and the support column 17.

In addition to the source grating 12, the fixed member 12a includes, for example, a filtration filter (also referred to as an additional filter) 192 for changing the quality of the X-rays transmitted through the source grating 12, and a filter for changing the quality of the X-rays transmitted through the source grating 12. An irradiation field aperture 193 for narrowing down the irradiation field, and an irradiation field lamp 194 for irradiating the subject with visible light instead of X-rays to perform positioning before irradiating the X-rays are attached.

Note that the source grating 12, filter filter 192, and irradiation field aperture 193 do not necessarily need to be provided in this order. A first cover unit 190 is arranged around the source grating 12 and the like to protect them.

The subject table 13 is a table on which the fiber composite material H is placed, and can function as a rotation stage that rotates the fiber composite material H around the z-axis. When capturing a plurality of moire images Mo using the fringe scanning method described above, it is possible to capture a plurality of moire images Mo while rotating the subject stage 13 at different angles.

In this embodiment, the controller 19 (see FIG. 3) is composed of a computer in which a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), an input/output interface, etc. (not shown) are connected to a bus. has been done. Note that it is also possible to configure the controller 19 as a dedicated control device instead of a general-purpose computer as in this embodiment. Further, although not shown, the controller 19 is provided with appropriate means and devices such as an input means including an operation section, an output means, a storage means, and a communication means. The output means includes a display unit (not shown) that displays information necessary for performing various operations of the X-ray Talbot-Lau apparatus 1 and generated reconstructed images.

The controller 19 is designed to perform general control over the X-ray Talbot-Low apparatus 1. That is, for example, the controller 19 is connected to the X-ray generator 11, and can set the tube voltage, tube current, irradiation time, etc. for the X-ray source 11a. Further, for example, it is also possible to configure the controller 19 to relay transmission and reception of signals and data between the X-ray detector 16 and an external device (for example, the prediction device 100). The prediction device 100 may function as the controller 19.

The X-ray Talbot-Lau apparatus 1 may perform orientation imaging to generate orientation images (amp images, ave images, and pha images, which will be described later). Oriented photography refers to photography in which the relative angle between the grating and the sample (fiber composite material H) is changed by rotating the subject table 13 that functions as a rotation stage. By oriented imaging, the direction in which the signal value is strongest for each pixel can be determined through calculation processing. Orientation photography will be explained below.

First, images are taken while changing the relative angle between the sample and the grid. Prepare at least three types of relative angles (eg, 0°, 60°, 120°). For example, a desired relative angle may be achieved by fixing the device and rotating the sample, or by fixing the sample and rotating the device. In the following, two-dimensional imaging will be explained as an example, but it can also be extended to three-dimensional imaging.

Next, an absorption image, differential phase image, or small-angle scattering image is acquired for each prepared relative angle. Hereinafter, a small-angle scattering image or a small-angle scattering image divided by an absorption image will be used. The small-angle scattering image divided by the absorption image can be said to be an image in which thickness dependence is canceled in the case of a sample with unevenness. For convenience of explanation, both will be collectively referred to as a "small-angle scattering image."

Next, align the prepared small-angle scattering images (three or more) for each relative angle. Here, since the sample is rotated, each image is returned to a predetermined angle.

Finally, fitting is performed using a sine wave for each pixel, and fitting parameters are extracted. A sine wave graph is a graph in which the horizontal axis represents the relative angle between the sample and the grating, and the vertical axis represents the small-angle scattering signal value of a certain pixel. The amplitude, average, and phase of the sine wave are obtained as fitting parameters. An image showing the amplitude value of each pixel will be called an "amp image," an image showing the average value of each pixel will be called an "ave image," and an image showing the phase of each pixel will be called a "pha image." The amp image, ave image, and pha image are collectively referred to as an "orientation image." The fitting method is not limited to sine waves; for example, an ellipse with the angle (phase) of the largest intensity as θ ₀ , the largest intensity as a, and the lowest intensity as b is expressed in polar coordinates as the position r(θ). It is also possible to fit the equation (1). In this case, an image with a value (ab)/2 corresponding to the amplitude of each pixel is an "amp image", and a value corresponding to the average value of each pixel (a+b) corresponds to the name used in sine wave fitting. An image showing /2 may be called an "ave image", and an image showing θ ₀ for each pixel may be called a "pha image". The orientation image may be created by simply assigning the major axis a, the minor axis b, and the phase θ ₀ of the signal intensity to each pixel.

Such an X-ray Talbot-Lau apparatus 1 generates a Talbot image containing information regarding the fiber orientation of the fiber composite material H. That is, the first measurement information generated by the X-ray Talbot-Lau apparatus 1 is information regarding the Talbot image. The Talbot image is an image generated by the Talbot effect described above when the fiber composite material H is photographed by the X-ray Talbot-Lau apparatus 1. Images that have been subjected to image processing, such as the orientation image described above, are also included in the Talbot image. A reconstructed image reconstructed from the moire image Mo is also included in the Talbot image.

(Second measuring device 300)
The second measurement device 300 is a device for measuring the material properties of the fiber composite material and generating second measurement information. This second measuring device 300 is different from the first measuring device 200. The second measurement information preferably includes information regarding at least one of the resin species, molecular orientation, crystallinity, specific volume, fiber amount, interaction between fibers and resin, and fiber length of the fiber composite material. Resin type, molecular orientation, crystallinity, specific volume, fiber amount, interaction between fiber and resin, fiber type, fiber diameter, and fiber length of fiber-reinforced resin may influence residual stress. This is because it has a high level of quality.

The second measurement device 300 is preferably an infrared spectrometer, an impedance spectrometer, a terahertz wave spectrometer, an ultrasonic measurement device, a Raman spectrometer, an X-ray diffraction device, or a fluorescence microscope. Such a second measurement device 300 is capable of obtaining second measurement information including information regarding the resin type, molecular orientation, crystallinity, specific volume, fiber amount, interaction between fibers and resin, or fiber length of the fiber composite material. This is because it can generate .

For example, by measuring the material properties of the fiber composite material using an infrared spectrometer, it is possible to generate second measurement information including information regarding the resin type, the degree of crystallinity by the resin type, the specific volume, etc. By measuring the material properties of the fiber composite material using a terahertz wave spectrometer, it is possible to generate second measurement information including information regarding the interaction between the fibers and the resin, the degree of crystallinity, and the like. By measuring the material properties of the fiber composite material using an ultrasonic measuring device, it is possible to generate second measurement information including information regarding the specific volume, fiber amount, resin type, and the like. By measuring the material properties of the fiber composite material using an impedance spectrometer, it is possible to generate second measurement information including information regarding the amount of fibers, the interaction between the fibers and the resin, and the like. By measuring the material properties of the fiber composite material using an X-ray diffraction device, it is possible to generate second measurement information including information regarding the degree of crystallinity and the like. Furthermore, by using an X-ray diffraction device, it is also possible to analyze the resin species based on the degree of crystallinity.

Such a second measurement device 300 measures each wavelength, including information regarding, for example, the resin type, molecular orientation, crystallinity, specific volume, fiber amount, interaction between fibers and resin, or fiber length of the fiber composite material. generate spectral or fluorescence microscopy images. That is, the second measurement information generated by the second measurement device 300 is information regarding the spectrum of each wavelength or the fluorescence microscopic image.

The prediction system may further include a third measuring device, or a third measuring device and a fourth measuring device. The third measurement device and the fourth measurement device are devices for measuring the material properties of the fiber composite material and generating third measurement information and fourth measurement information. The third measuring device and the fourth measuring device are different from the first measuring device 200 and the second measuring device 300. The third measuring device and the fourth measuring device may be the same measuring device as explained in the second measuring device 300 above.

The third measurement information and the fourth measurement information include information regarding at least one of the resin type, molecular orientation, crystallinity, specific volume, fiber amount, interaction between fibers and resin, and fiber length of the fiber composite material. It is preferable to include. The third measuring device and the fourth measuring device each include information regarding, for example, the resin type, molecular orientation, crystallinity, specific volume, fiber amount, interaction between fibers and resin, or fiber length of the fiber composite material. Generate a spectrum of wavelengths or fluorescence microscopy images. That is, the third measurement information and the fourth measurement information generated by the third measurement device and the fourth measurement device are information regarding the spectrum of each wavelength or the fluorescence microscopic image.

<Function of prediction device 100>
FIG. 6 is a block diagram showing the functional configuration of the prediction device 100.

As shown in FIG. 6, the prediction device 100 functions as an acquisition unit 111, an extraction unit 112, a prediction unit 113, and a control unit 114 when the CPU 110 reads a program stored in the storage 140 and executes the process.

The acquisition unit 111 acquires the first measurement information generated by the first measurement device 200 and the second measurement information generated by the second measurement device 300. The first measurement information and the second measurement information are information obtained by measuring the material properties of the fiber composite material using the first measurement device 200 and the second measurement device 300, respectively. The first measurement information preferably includes information regarding fiber orientation, and is preferably information regarding Talbot images.

The extraction unit 112 extracts feature amounts from each of the first measurement information and the second measurement information acquired by the acquisition unit 111. The feature amount is, for example, a numerical value extracted from the spectrum or image acquired by the acquisition unit 111, and linked to the physical properties of the fiber composite material. When the first measurement information is information related to the Talbot image, the feature amount extracted from the first measurement information may be the Talbot image itself such as an orientation image (amp image, ave image, pha image), or the Talbot image itself. It may be an image signal value obtained from a specific area of .

Further, when the first measurement information is information regarding the Talbot image, the feature amount extracted from the first measurement information may be eccentricity ecc. For example, the eccentricity ecc can be determined by using the signal values amp and ave obtained from the orientation image and calculating σ1=ave+amp (corresponding to the maximum value of the small-angle signal value) and σ2=ave-amp (corresponding to the minimum value of the small-angle signal value). ) is calculated using the following equation (2). The orientation image may include an image (ecc image) showing the degree of eccentricity ecc for each pixel.

When the second measurement information is information regarding the spectrum of each wavelength, the feature amount extracted from the second measurement information is, for example, the peak wavelength. When the second measurement information is an FTIR spectrum, an impedance spectroscopic spectrum, or an X-ray diffraction spectrum, the feature amount extracted from the second measurement information is a peak position, peak intensity, half-value width, or the like. Further, the feature amount extracted from the second measurement information may be a principal component obtained by principal component analysis of the spectrum. When the second measurement information is an impedance spectrum, the feature amount extracted from the second measurement information is a resistance value or capacitance at a specific frequency. When the second measurement information is an X-ray CT image, the feature amount extracted from the second measurement information is the fiber length or fiber orientation obtained by image analysis. When the second measurement information is a fluorescence microscope image, the feature amount extracted from the second measurement information is the fiber length etc. obtained by image analysis. When the second measurement information is ultrasonic measurement, the feature amount extracted from the second measurement information is the speed of sound, reflection on the front surface, reflection on the back surface, or the like. Further, the feature amount extracted from the second measurement information may be a principal component obtained by principal component analysis of the measured waveform. The acquisition unit 111 may acquire information from which feature amounts are extracted. That is, the first measurement information and the second measurement information may have feature amounts extracted from information regarding the fiber composite material measured by the first measurement device 200 and the second measurement device 300.

The prediction unit 113 predicts the residual stress of the fiber composite material based on the first measurement information and the second measurement information acquired by the acquisition unit 111. Specifically, the prediction unit 113 inputs the feature amounts of the first measurement information and the second measurement information extracted by the extraction unit 112 using a learned classifier, and calculates the residual stress of the fiber composite material. Predict.

The control unit 114 causes the display unit 160 to output information regarding the residual stress of the fiber composite material predicted by the prediction unit 113.

FIG. 7 shows an example of information regarding the residual stress of the fiber composite material output to the display section 160. For example, the predicted residual stress value is displayed on the display unit 160 along with information regarding the fiber composite material.

The processing executed in the prediction device 100 will be described in detail below.

<Processing overview>
FIG. 8 is a flowchart showing the procedure of the prediction process executed by the prediction device 100. The processing of the prediction device 100 shown in the flowchart of FIG. 8 is stored as a program in the storage 140 of the prediction device 100, and is executed by the CPU 110 controlling each part.

(Step S101)
The prediction device 100 first acquires first measurement information obtained by measuring the material properties of the fiber composite material by the first measurement device 200 and second measurement information measured by the second measurement device 300. The prediction device 100 obtains first measurement information from the first measurement device 200 and second measurement information from the second measurement device 300, for example. The first measurement device 200 and the second measurement device 300 may store the first measurement information and the second measurement information in another device such as a server, and the prediction device 100 may store the first measurement information and the second measurement information in another device such as a server. Second measurement information may also be acquired.

(Step S102)
The prediction device 100 extracts feature amounts from each of the first measurement information and the second measurement information acquired in the process of step S101.

(Step S103)
The prediction device 100 inputs the feature amounts of each of the first measurement information and the second measurement information extracted in the process of step S102 to a discriminator that has undergone machine learning in advance, and predicts the residual stress of the fiber composite material. For example, the discriminator uses the learning method described below to determine the feature values of each of the first measurement information and second measurement information of a plurality of fiber composite materials prepared in advance, and the residual stress of each of the plurality of fiber composite materials. Machine learning is performed using training data with measured values. Specifically, the discriminator uses the feature quantities of each of the first measurement information and the second measurement information regarding the plurality of fiber composite materials as input data, and uses the measured value of residual stress of each of the plurality of fiber composite materials as output data. be learned. Thereby, the prediction device 100 can predict the residual stress of the fiber composite material by inputting the feature amounts extracted for each of the first measurement information and the second measurement information into the discriminator. Measurements of residual stress in fiber composite materials are obtained using, for example, a perforation method.

The discriminator may perform machine learning using the first measurement information and second measurement information regarding the plurality of fiber composite materials as input data, and using the measured values of residual stress of each of the plurality of fiber composite materials as output data. Further, the information input to the discriminator is not limited to the feature amounts of each of the first measurement information and the second measurement information. For example, in addition to the feature amounts of each of the first measurement information and the second measurement information, information at the time of manufacture may be input to the discriminator and used as information for learning and prediction.

(Step S104)
The prediction device 100 generates a prediction result of the residual stress of the fiber composite material based on the output from the discriminator in the process of step S103.

(Step S105)
The prediction device 100 outputs the prediction result generated in the process of step S104. For example, the prediction device 100 displays the value of the residual stress of the fiber composite material predicted in the process of step S103 on the display unit 160 together with information regarding the fiber composite material (FIG. 7).

<About learning process>
Next, a machine learning method for trained models used in the classifier will be described.

FIG. 9 is a flowchart showing a machine learning method for a trained model.

In the process of FIG. 9, the feature amounts of each of the first measurement information and second measurement information of a plurality of fiber composite materials prepared in advance are input, and the measured value of residual stress of each of the plurality of fiber composite materials is output. Machine learning is performed using a large number (i sets) of data sets as learning sample data. For example, a stand-alone high-performance computer using a CPU and a GPU processor or a cloud computer is used as a learning device (not shown) that functions as a discriminator. In the following, a learning method using a neural network configured by combining perceptrons such as deep learning in a learning device will be described, but the method is not limited to this, and various methods can be applied. For example, random forest, decision tree, support vector machine (SVM), logistic regression, k-nearest neighbor method, topic model, etc. may be applied.

(Step S111)
The learning device reads learning sample data that is teacher data. If it is the first time, the first set of learning sample data is read, and if it is the i-th time, the i-th set of learning sample data is read.

(Step S112)
The learning device inputs input data from the read learning sample data to the neural network.

(Step S113)
The learning device compares the prediction results of the neural network with the correct data.

(Step S114)
The learning device adjusts the parameters based on the comparison results. The learning device adjusts the parameters so that the difference between the comparison results becomes smaller by, for example, executing processing based on back-propagation (error backpropagation method).

(Step S115)
If the learning device completes processing of all data from the 1st to the i-th set (YES), the process proceeds to step S116, and if not (NO), returns the process to step S111 and processes the next learning sample data. is read, and the processing from step S111 onwards is repeated.

(Step S116)
The learning device determines whether or not to continue learning, and when continuing (YES), returns the process to step S111, executes the processes from the 1st group to the i-th group again in steps S111 to S115, and continues. If not (NO), the process advances to step S117.

(Step S117)
The learning device stores the learned model constructed in the previous processing and ends (end). The storage destination includes the internal memory of the prediction device 100. In the process shown in FIG. 8 described above, the residual stress of the fiber composite material is predicted using the learned model generated in this way.

<Effects of prediction device 100 and prediction system>
The prediction device 100 and the prediction system of this embodiment acquire first measurement information and second measurement information that measured the material properties of a fiber composite material, and based on the acquired first measurement information and second measurement information, Predicting residual stresses in composite materials. This makes it possible to predict the residual stress in the fiber composite material. The effects will be explained below.

As mentioned above, residual stress can affect the occurrence of cracks when a molded product made of fiber composite material is used for a long period of time. For this reason, it is desirable to understand the residual stress of the fiber composite material at least before long-term use. However, there are not many methods for measuring residual stress, and the method for measuring residual stress in fiber composite materials is even more limited. The drilling method is mainly used to measure the residual stress in fiber composite materials, but since the drilling method requires precise measurement control, it is possible to efficiently measure the residual stress in a large number of fiber composite materials. It is difficult to do so.

On the other hand, in the prediction system and prediction device 100 of the present embodiment, the residual stress of the fiber composite material is predicted based on the first measurement information and the second measurement information that measured the material properties of the fiber composite material. There is no need to directly measure residual stress using a drilling method or the like.

In addition, since the residual stress of the fiber composite material is predicted based on the measurement information of multiple measurement devices (the first measurement device 200 and the second measurement device 300), the residual stress can be predicted based on the measurement information of a single measurement device. This makes it possible to predict residual stress from multiple angles. Therefore, it becomes possible to predict residual stress with higher accuracy.

Furthermore, by using an X-ray Talbot-Lau apparatus as the first measuring device 200, information regarding the fiber orientation of the fiber composite material can be obtained in a short time and over a wide range. This makes it possible to predict residual stress with higher accuracy.

As explained above, the prediction device 100 and the prediction system of this embodiment make it possible to predict the residual stress of the fiber composite material.

The effects of the present invention will be explained using the following examples. However, the technical scope of the present invention is not limited only to the following examples.

(Creating a trained classifier)
First, we created multiple samples of fiber composite materials to create training data. This sample was produced using the following combinations of four types of resin, three types of fibers, two conditions of fiber concentration (volume ratio), and two conditions of injection pressure. The resin and fibers were mixed in advance at a desired ratio using a Laboplastomill (registered trademark) extruder manufactured by Toyo Seiki Seisakusho Co., Ltd. This produced pellets. Samples of 48 types of fiber composite materials were molded using an injection molding machine SE50D manufactured by Sumitomo Heavy Industries. The sample shape was a dumbbell-shaped test piece type A1 shown in JIS K7139.

Resin: polypropylene (Noblen W101 manufactured by Sumitomo Chemical Co., Ltd.), polyamide 66 (Leona 1300S manufactured by Asahi Kasei Corporation), ABS (Toyolac 700 314 manufactured by Toray Industries, Inc.), polycarbonate (Iupilon H-3000R manufactured by Mitsubishi Engineering Plastics Co., Ltd.);
Fiber: PAN (polyacrylonitrile) carbon fiber (CF-N manufactured by Nippon Polymer Sangyo Co., Ltd.), PAN carbon fiber (TC-33 manufactured by Taiwan Plastics Co., Ltd.), glass fiber (CS3J-960 manufactured by Nitto Boseki Co., Ltd.);
Fiber concentration: 5%, 20%;
Injection pressure: 50MPa, 100MPa.

Next, each of these 48 types of fiber composite material samples was measured using the following measuring device, and the discriminator was made to learn the feature amounts. The measurement was performed near the center of the dumbbell-shaped test piece.

FTIR (Fourier Transform Infrared Spectroscopy) device (AVATAR370 manufactured by Thermo Fisher Scientific);
Terahertz wave spectrometer (C12068-01 manufactured by Hamamatsu Photonics Co., Ltd.);
Ultrasonic measuring device (UVM-2 manufactured by Ultrasonic Industry Co., Ltd. Measurement was performed in reflection mode);
Impedance spectrometer (Model 126096 manufactured by Solartron. Conductive tapes with a diameter of 5 mm were pasted on the sample at two locations 50 mm apart, and measurements were performed using these as electrodes.);
X-ray diffraction device (Smart Lab manufactured by Rigaku Co., Ltd.);
X-ray Talbot-Low device (device described in JP 2019-184450);
X-ray CT device (SKYSCAN1272 manufactured by Bruker Japan);
Fluorescence microscope (Used with mirror unit U-MWU2 attached to BX51 manufactured by Olympus Corporation. As a sensitivity adjuster, fluorescent dye 2,5-thiophenediylbis (5-tert-butyl-1,3-benzoxazole) was used at 0.1 weight % and prepared the sample).

Additionally, the residual stress of each of the 48 types of fiber composite material samples was measured using the perforation method, and the discriminator was trained. The drilling method was performed as follows. First, a bending test was performed on the test piece to determine the Young's modulus. Next, a commercially available rosette strain gauge is attached to the measurement location, a hole with a diameter of 2 mm and a depth of 1 mm is made using a commercially available drill, and the residual stress is calculated using the amount of displacement and the Young's modulus determined in advance. I asked for it.

(Examples 1 to 11, Comparative Examples 1 to 7)
First, samples of four types of fiber composite materials were produced. This sample was produced using the following combinations of two types of resin, two types of fibers, one condition of fiber concentration (volume ratio), and one condition of injection pressure. The samples were prepared in the same manner as for the training data described above.

Resin: polypropylene (Noblen W101 manufactured by Sumitomo Chemical Co., Ltd.), polyamide 66 (Leona 1300S manufactured by Asahi Kasei Corporation);
Fiber: PAN-based carbon fiber (CF-N manufactured by Nippon Polymer Sangyo Co., Ltd.), PAN-based carbon fiber (TC-33 manufactured by Taiwan Plastics Co., Ltd.);
Fiber concentration: 10%;
Injection pressure: 80MPa.

In Examples 1 to 11, samples of each of these four types of fiber composite materials were measured using a plurality of measuring devices shown in Table 1 below. In Comparative Examples 1 to 7, samples of each of the four types of fiber composite materials were measured using a single measuring device shown in Table 1 below. Thereafter, the spectra or image features measured by the measuring device were input into a trained discriminator to obtain predicted values of residual stress.

Additionally, the residual stress of each of the above four types of fiber composite materials was measured using a perforation method, and the measured values were determined. Next, after calculating the error between the predicted value and the measured value using the following formula (3), the average of the errors for the four types of fiber composite materials was determined. In Table 1 below, the average error of each of the four types of fiber composite materials of Comparative Example 1 is set as 1, and the average error of Examples 1 to 11 and Comparative Examples 1 to 7 is listed as a relative value. That is, the smaller the value in the "error" column in Table 1, the higher the accuracy of the residual stress predicted using the learned discriminator.

Examples 1 to 11, in which spectra or image feature quantities measured by multiple measurement devices were input to the discriminator, had smaller errors than Comparative Examples 1 to 7. Further, among Examples 1 to 11, in Examples 1 and 4 to 11 using the X-ray Talbot-Rho measuring device, the error was able to be made smaller than in Examples 2 and 3. Furthermore, by effectively combining various types of measuring devices, it was possible to predict residual stress with higher accuracy (Examples 8 to 11).

The configurations of the prediction device 100 and the prediction system described above are the main configurations explained in explaining the features of the above-mentioned embodiments and examples, and are not limited to the above-mentioned configurations, but within the scope of the claims. Various modifications can be made. Moreover, the configuration provided in a general prediction system is not excluded.

For example, the prediction device 100 may include components other than the above components, or may not include some of the above components.

Furthermore, the prediction device 100, the first measuring device 200, and the second measuring device 300 may each be configured by a plurality of devices, or may be configured by a single device.

Furthermore, the functions of each configuration may be realized by other configurations. For example, the first measurement device 200 or the second measurement device 300 is integrated into the prediction device 100, and some or all of the functions of the first measurement device 200 and the second measurement device 300 are realized by the prediction device 100. Good too.

Furthermore, the processing units in the flowchart in the above embodiment are divided according to the main processing contents in order to facilitate understanding of each process. The present invention is not limited by how the processing steps are classified. Each process can also be divided into more process steps. Also, one processing step may perform more processing.

The means and methods for performing various processes in the system according to the embodiments described above can be realized by either a dedicated hardware circuit or a programmed computer. The program may be provided on a computer-readable recording medium such as a flexible disk or CD-ROM, or may be provided online via a network such as the Internet. In this case, the program recorded on the computer-readable recording medium is usually transferred and stored in a storage unit such as a hard disk. Further, the above program may be provided as a standalone application software, or may be incorporated into the software of the device as a function of the system.

This application is based on a Japanese patent application (Japanese Patent Application No. 2022-47924) filed on March 24, 2022, the disclosure content of which is incorporated by reference in its entirety.

100 prediction device,
110 CPU,
111 Acquisition Department;
112 Extraction part,
113 Prediction Department,
114 control unit,
120 ROM,
130 RAM,
140 storage,
150 communication interface,
160 display section,
170 Operation reception department,
200 first measuring device,
300 Second measuring device.

Claims

Obtain first measurement information obtained by measuring the material properties of the fiber composite material with a first measurement device, and second measurement information obtained by measuring the material properties of the fiber composite material with a second measurement device different from the first measurement device. an acquisition unit to
A prediction device comprising: a prediction unit that predicts residual stress of the fiber composite material based on the acquired first measurement information and second measurement information.
The prediction device according to claim 1, further comprising a control unit that causes an output unit to output information regarding the predicted residual stress.
The prediction device according to claim 1 or 2, wherein the first measurement information includes information regarding fiber orientation of the fiber composite material.
The prediction device according to claim 3, wherein the first measurement device is an X-ray Talbot-Lau device or an X-ray CT measurement device.
The second measurement information includes at least one of the resin type, molecular orientation, crystallinity, specific volume, fiber amount, interaction between fiber and resin, fiber type, fiber diameter, and fiber length of the fiber composite material. The prediction device according to any one of claims 1 to 4, comprising information regarding.
Any one of claims 1 to 5, wherein the second measurement device is an infrared spectrometer, an impedance spectrometer, a terahertz wave spectrometer, an ultrasonic measurement device, a Raman spectrometer, an X-ray diffraction device, or a fluorescence microscope. A prediction device described in .
The acquisition unit further acquires third measurement information measured by a third measurement device different from the first measurement device and the second measurement device,
7. The prediction device according to claim 1, wherein the prediction unit predicts the residual stress based on the acquired first measurement information, second measurement information, and third measurement information.
The prediction device according to any one of claims 1 to 7, wherein the fiber composite material contains a sensitivity adjuster.
The prediction device according to any one of claims 1 to 8, wherein the prediction unit predicts the residual stress using a learned discriminator.
further comprising an extraction unit that extracts a feature amount from each of the acquired first measurement information and second measurement information,
The prediction device according to claim 9, wherein the prediction unit receives the extracted feature amount as input and predicts the residual stress.
The prediction device according to claim 10, wherein the discriminator performs machine learning using the feature amount as input data and the residual stress as output data.
a first measuring device that measures material properties of the fiber composite material;
a second measuring device that measures material properties of the fiber composite material and is different from the first measuring device;
A prediction system comprising: the prediction device according to any one of claims 1 to 11.
Obtain first measurement information obtained by measuring the material properties of the fiber composite material with a first measurement device, and second measurement information obtained by measuring the material properties of the fiber composite material with a second measurement device different from the first measurement device. step (a) of
A prediction program for causing a computer to execute a process comprising a step (b) of predicting the residual stress of the fiber composite material based on the first measurement information and the second measurement information acquired in step (a). .