Strain distribution testing device

The strain distribution testing apparatus uses X-ray imaging and digital image correlation to measure three-dimensional strain in opaque rubber, addressing the limitations of previous methods by enabling accurate strain analysis in a single test piece.

WO2026240119A1PCT designated stage Publication Date: 2026-11-19TOHOKU UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
PCT/JP2026/018025
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-05-12
Filing Date
2026-05-12
Publication Date
2026-11-19

AI Technical Summary

Technical Problem

Existing methods for measuring strain distribution in soft materials like rubber are limited to transparent materials and require multiple test pieces, lacking versatility and complicating the measurement process.

Method used

A strain distribution testing apparatus using X-ray imaging and digital image correlation to measure three-dimensional strain distribution within a single rubber test piece, employing observation particles that do not transmit X-rays, a pressing unit, and an imaging unit to capture X-ray images from intersecting directions, allowing for quantitative measurement.

Benefits of technology

Enables the quantitative measurement of three-dimensional strain distribution in opaque rubber using a single test piece, overcoming limitations of previous methods by providing accurate and efficient strain analysis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure JP2026018025_19112026_PF_FP_ABST
    Figure JP2026018025_19112026_PF_FP_ABST
Patent Text Reader

Abstract

The present invention provides a strain distribution testing device that makes it possible to quantitatively measure three-dimensional strain distribution inside rubber using only a single rubber test piece, not limited to transparent rubber. This control unit presses a measurement surface so that pressure applied by a pressing unit is orthogonal to a moving surface, moves the pressing unit, and then irradiates a test piece with X-rays from two directions, namely the movement direction of the pressing unit and the direction intersecting the movement direction, deforming the test piece with the pressing force of the pressing unit, and measures the positional change in observed particles inside the test piece caused by deformation of the test piece.
Need to check novelty before this filing date? Find Prior Art

Description

Strain distribution testing apparatus

[0001] The present invention relates to a strain distribution testing apparatus. This application claims priority based on Japanese Patent Application No. 2025-079608 filed in Japan on May 12, 2025, the content of which is incorporated herein by reference.

[0002] For example, materials that include polymers such as macromolecules, colloids, and liquid crystals and exhibit elasticity, such as rubber, resins, and biological substances, are generally collectively referred to as soft matter, and the aforementioned soft matter may come into contact and generate friction in industrial products such as shoe soles and tires, artificial joints, and living human skin.

[0003] Compared with hard materials such as metals and ceramics, soft matter has low elasticity, so it is accompanied by large strain and deformation during contact and friction, and a larger real contact portion is easily formed. In addition, since it is a material composed of polymers, it has viscoelasticity, and has the characteristic that a hysteresis term arises from changes in the strain distribution accompanying friction. That is, consideration of strain distribution is indispensable for understanding and controlling the contact and friction of soft matter.

[0004] Conventionally, as methods for determining such strain distributions, many predictions of strain distribution based on computational simulations have been reported. However, on the other hand, reported cases of strain distributions based on experiments are limited.

[0005] For example, Non-Patent Document 1 describes the behavior of the strain distribution within the cross-section of a rubber test piece accompanying friction by optically tracking markers arranged on the horizontal cross-section of a transparent rubber test piece. It also describes that quantification of the three-dimensional strain distribution in the horizontal cross-section can be achieved by imaging the markers on the horizontal cross-section of a test piece made of transparent rubber with two cameras.

[0006] Ueno, Nomoto, Nishi, Sugisawa, Hanzawa, Hata, Yamaguchi, Kuwayama, Yamaguchi: Relationship between strain distribution and vertical load in sliding friction of rubber and resin spheres, Tribology Conference 2024 Autumn Nago, Nago, B27.

[0007] However, the strain distribution measurement method described in Non-Patent Document 1 is limited to transparent rubber as the measurement target, and when measuring at multiple horizontal cross-sectional heights that are different from each other, it is necessary to prepare multiple different rubber test pieces, which has the drawbacks of low versatility and complicated measurement.

[0008] This invention was proposed in view of the above-mentioned problems, and aims to provide a strain distribution testing device that can quantitatively measure the three-dimensional strain distribution inside rubber using only one rubber test piece, not limited to transparent rubber.

[0009] The inventors measured the displacement inside the rubber test piece due to sliding by tracking the marker positions using X-ray imaging and image correlation in a friction test of a rubber test piece with markers placed on its vertical cross-section. They newly discovered that this measurement method is not limited to transparent rubber as the target material, and that by extracting multiple minute regions and calculating their displacements, it is possible to quantify the three-dimensional strain inside the rubber test piece.

[0010] To solve the above-mentioned problems, the strain distribution testing apparatus of one embodiment of the present invention proposes the following means: (1) The strain distribution testing apparatus of embodiment 1 of the present invention is a strain distribution testing apparatus capable of observing deformation behavior due to strain inside a test piece, wherein a plurality of observation particles that do not transmit X-rays are arranged inside the test piece so as to fit within a measurement surface which is an arbitrary plane, and is movable along a movable surface, and comprises a pressing unit that presses the measurement surface, an imaging unit capable of capturing a transmitted image of the observation particles arranged in the test piece, and a control unit that controls the pressing unit and the imaging unit, wherein the measurement surface is arranged perpendicular to the movable surface, and the control unit presses the measurement surface so that the pressure applied by the pressing unit is perpendicular to the movable surface, moves the pressing unit, and then irradiates the test piece with X-rays from two directions that intersect each other, deforms the test piece with the pressing force of the pressing unit, and controls the apparatus to measure the change in position of the observation particles inside the test piece caused by the deformation of the test piece.

[0011] (2) Aspect 2 of the present invention is the strain distribution testing apparatus of Aspect 1, wherein the test specimen includes a material that transmits X-rays.

[0012] (3) Embodiment 3 of the present invention is the strain distribution testing apparatus of Embodiment 1 or 2, wherein the observed particles include a metal with an atomic number of 26 or higher.

[0013] (4) Embodiment 4 of the present invention is a strain distribution testing apparatus according to any one of embodiments 1 to 3, wherein the material constituting the test piece includes an element with an atomic number of 14 or less.

[0014] (5) Embodiment 5 of the present invention is a strain distribution testing apparatus according to any one of embodiments 1 to 4, wherein the X-rays are white X-rays.

[0015] According to the present invention, it is possible to provide a strain distribution testing device that can quantitatively measure the three-dimensional strain distribution inside rubber using only one rubber test piece, which is not limited to transparent rubber.

[0016] This is a schematic diagram showing a strain distribution testing apparatus according to one embodiment of the present invention. This is a photograph showing an example of a test piece S. This is a schematic diagram of an embodiment of the present invention. This is a photograph showing the actual experimental environment of an embodiment of the present invention. This is a schematic diagram of an experiment conducted using the measurement method of the embodiment. This is a schematic diagram showing the procedure of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the embodiment. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example. This is data showing the results of the example.

[0017] The strain distribution testing apparatus according to one embodiment of the present invention will be described below with reference to the drawings. The embodiments and examples shown below are provided specifically to help the spirit of the invention be better understood, and do not limit the present invention unless otherwise specified. In addition, the drawings used in the following description may be enlarged for convenience to make the features of the present invention easier to understand, and the dimensional ratios of each component may not be the same as in reality.

[0018] Furthermore, in the following embodiments and examples, when we say "A and B are perpendicular" or "A is orthogonal to B", in addition to the angle of B relative to A being 90°, an angle range of approximately 85° to 95° is also permitted.

[0019] Furthermore, in the following embodiments and examples, the term "transparent" refers to a state in which electromagnetic waves in the visible light wavelength range (generally 380 nm to 780 nm) can be transmitted.

[0020] (Strain Distribution Testing Apparatus) Figure 1 is a schematic diagram showing a strain distribution testing apparatus according to one embodiment of the present invention. The strain distribution testing apparatus 10 of this embodiment includes a stage 11 on which a test piece S, which is the object to be measured, is placed, a pressing unit 12 for pressing the test piece S, an imaging unit 13 for imaging the test piece S, and a control unit 14 for controlling each of these components.

[0021] In this embodiment, the test piece S used to measure strain in the strain distribution testing apparatus 10 is, for example, shaped into a rectangle, and a plurality of observation particles D that are not transmitted by X-rays are arranged inside so as to fit within a measurement surface F, which is any plane within the test piece S.

[0022] Stage 11 only needs to be movable in the X and Y directions in Figure 1, and also rotatable around the Z direction.

[0023] In this embodiment, the measuring surface F of the test piece S is a surface that intersects the X and Y directions in Figure 1, respectively, and extends parallel to the Z direction. For example, the measuring surface F can be a surface that forms a 45° angle with the X and Y directions in Figure 1, respectively, and extends parallel to the Z direction.

[0024] The constituent material of such test specimen S may be, for example, transparent rubber, resin, etc. Preferably, the constituent material of test specimen S is composed of a material containing an element with an atomic number of 14 or less (Si or less).

[0025] Furthermore, the observed particle D is preferably composed of, for example, an X-ray shielding metal, such as a material containing a metallic element with an atomic number of 26 or higher (Fe or higher). A photographic example of such a test specimen S is shown in Figure 2.

[0026] The pressing section 12 of the strain distribution testing apparatus 10 is equipped with an indenter 12a that presses the measurement surface F in the Z direction with an arbitrary pressing force. This indenter 12a is formed to be movable along a movable surface M which is one surface (top surface) of the test piece S.

[0027] The imaging unit 13 includes two X-ray sources 13a and 13b capable of irradiating X-rays in directions that intersect each other, and an X-ray imaging device 13c that images a region containing observation particles D placed on the measurement surface F of the test piece S. In this embodiment, the two X-ray sources 13a and 13b are arranged to irradiate X-rays in directions that are orthogonal to each other.

[0028] The X-rays emitted from these X-ray sources 13a and 13b can be, for example, white X-rays in any X-ray wavelength range, which are a mixture of multiple consecutive wavelengths.

[0029] The control unit 14 may consist of a personal computer or interface that controls the operation of the pressing unit 12 and the imaging unit 13.

[0030] In the strain distribution testing apparatus 10 of this embodiment, the control unit 14 first controls the pressing unit 12 so that the pressure applied by the indenter 12a presses the measuring surface F in a direction perpendicular to the moving surface M (the Z direction in Figure 1). Then, while pressing the measuring surface F, the indenter 12a is moved in any direction along the moving surface M. As a result, the measuring surface F of the test piece S deforms.

[0031] Next, the control unit 14 controls the imaging unit 13 to irradiate the test piece S with X-rays from the X-ray sources 13a and 13b in two directions that intersect each other, relative to the pressing position by the indenter 12a of the pressing unit 12. Then, the X-ray imaging device 13c captures an X-ray image of the measurement surface F and measures the change in position of the observation particles D placed on the measurement surface F, which is caused by the deformation of the test piece S.

[0032] The control unit 14 then calculates the strain distribution generated on the measurement surface F by the pressing force applied by the indenter 12a based on the positional change of the observed particle D, and outputs the result as a distribution map or the like.

[0033] In the following, more detailed examples of the present invention will be described using embodiments of the present invention. In this embodiment, the focus is on the friction between the rubber block and the resin sphere, and the objective is to develop a technique for quantifying the three-dimensional strain distribution. Specifically, the objective is to quantify the three-dimensional strain distribution within the rubber block by incident X-rays from two directions and capturing an X-ray transmission image.

[0034] <Development and accuracy verification of a strain measurement method using synchrotron X-rays> Quantifying the strain distribution inside a material is a very effective means of understanding the friction phenomena of soft materials such as rubber, but its achievement has been limited until now. In order to quantify the strain distribution in three dimensions, it was assumed that the strain distribution inside the material could be quantified by calculating the three-dimensional coordinates of singularities such as metal markers inside the material and analyzing the change in their position.

[0035] Generally, measuring the location of singularities inside an object can be achieved by continuously acquiring X-ray projection images while rotating the object and constructing computed tomography (CT) images. To acquire CT images inside materials with friction, it is necessary to acquire X-ray projection images at high speed (100 to 10,000 fps) while rotating the X-ray irradiation direction or friction mechanism (6,000 to 600,000 rpm) at a speed that sufficiently ignores the sliding velocity (0.1 to 10 mm / s). However, in reality, experiments involving high-speed rotation of the X-ray irradiation direction or friction mechanism are technically impossible.

[0036] Therefore, in the following embodiment, a rubber test piece with metal markers scattered on one cross-section in the vertical direction is used, and the friction test is captured by X-ray imaging. At this time, instead of rotating the friction tester, X-ray imaging is performed twice from two directions that are 90° apart from each other, and the time-series change of the metal markers is obtained by applying the digital image correlation method (DICM) to the acquired imaging data, and an attempt is made to quantify the strain distribution of the rubber material during friction in three-dimensional space. A schematic diagram of this embodiment is shown in Figure 3. Note that the measurement method using the strain distribution tester assumes that the strain distribution is the same when measured from two intersecting directions.

[0037] The measurement accuracy of the measurement method described above depends on the accuracy of identifying the internal marker position of the material based on DICM. Therefore, in this embodiment, we verify the accuracy of identifying the internal marker position of the material. Specifically, a resin block containing a stainless steel sphere was placed on a rotating stage, and X-ray transmission images were acquired as it rotated 360 degrees. The two-dimensional position of the stainless steel sphere was calculated based on DICM from each of the X-ray transmission images with irradiation angles differing by 90°, and the three-dimensional position of the stainless steel sphere was calculated by solving the simultaneous equations of the results from the two directions.

[0038] In addition, we constructed CT images using an existing method with the highest positional resolution (0.01–10 μm) and calculated the three-dimensional position of the stainless steel sphere. By comparing the three-dimensional positions of the stainless steel sphere calculated from these two measurement methods, we verified the accuracy of this measurement method.

[0039] <Verification Example 1> [Measurement Overview] The accuracy verification of this measurement method was conducted at BL28B2 of the large synchrotron radiation facility SPring-8 located in Sayo-cho, Hyogo Prefecture. The reason for using this facility is its outstanding high brightness characteristics that enable X-ray imaging at high speeds of 1000 fps or more in order to observe the strain due to friction in rubber in real time, and its high straightness that enables X-ray imaging without distortion due to position within the material.

[0040] When performing observations, imaging at a high frame rate is required to elucidate the internal behavior of materials during friction tests. However, with laboratory X-ray sources, the X-ray intensity is insufficient, leading to a significant influence of noise contained in captured images, and X-ray images with a sufficient signal-to-noise ratio cannot be obtained. The synchrotron radiation of SPring-8 has higher luminance compared to laboratory X-ray sources, thus enabling sufficient transmitted X-ray intensity to reach the detector even with short exposures on the microsecond order. In addition, to measure minute displacements inside materials with high accuracy, it is necessary to minimize the difference in appearance between the front and back of the sample caused by X-ray divergence, and the high straight-line propagation property of synchrotron radiation is indispensable.

[0041] Figure 4 shows a photograph of the actual experimental environment. X-rays entered through a beryllium window located on the front left side of the photograph in Figure 4, passed over the upper part of the rotation stage installed in the testing machine, and entered the scintillator placed deep inside the opening on the right side of the photograph. Figure 5 shows a schematic diagram of an experiment conducted using the measurement method of the present example.

[0042] Three stainless steel balls made of SUS304 with a diameter of 0.5 mm were bonded to one side surface of an acrylic block with an adhesive, and this acrylic block was mounted on a rotation stage. Then, while rotating at a rotation speed of 20 rpm, X-ray imaging was performed for 3.0 seconds at an imaging rate of 1250 fps using one high-speed camera (TMX6410, manufactured by Vision Research, Inc.).

[0043] At this time, the magnification of the lens (AF-S, Micro NIKKOR, manufactured by Nikon Corporation) was set to 1.0x, the focal length was 105 mm, and a GAGG scintillator with a thickness of 10 μm was used as the X-ray-to-visible light conversion element. In addition, the X-rays usable at BL28B2 are white X-rays, and a 16.0 mm aluminum plate was installed on the front surface of the resin block to absorb synchrotron radiation in the high-energy region. The ambient temperature in the beamline was 24.8°C, and the relative humidity was 35%. A list of these experimental conditions is shown in Table 1.

[0044]

[0045] [Displacement Calculation of Stainless Steel Balls Using DICM] Since the X-ray transmittance of stainless steel balls is sufficiently lower than that of the resin block (acrylic resin), the stainless steel balls are observed as low-brightness regions in X-ray transmission images. In this example, the temporal change of the three-dimensional position of each of the three stainless steel balls was quantified based on the following two procedures.

[0046] (Procedure 1) As shown in FIG. 6, X-ray transmission images from two directions differing by 90° are extracted, and the positions of the stainless steel balls in each X-ray transmission image are tracked by digital image correlation method (DICM), to calculate the positions of the stainless steel balls in the xz plane and the yz plane, respectively. (Procedure 2) The three-dimensional positions of the stainless steel balls were derived by combining the results quantified in Procedure 1 described above.

[0047] Note that in the DICM of Procedure 1, a rectangular region expanded outward by 7 pixels each in the up, down, left, and right directions from the contour of the transmission image of the stainless steel ball was defined as a subset, and the displacement amount was calculated by tracking the subset position in each frame.

[0048] FIG. 7 shows the captured image of the stainless steel ball obtained from the X-ray transmission image and the calculated subset. The κ×κ pixel subset centered at the position (xi, yi) shown in (a) of FIG. 7 is displaced to the position (xi+1, yi+1) as shown in (b) of FIG. 7 along with the rotation of the stage. In DICM, a region having a high correlation with the luminance value distribution of the subset before deformation is searched from the captured image after deformation, to calculate the subset position. The correlation of the luminance distribution was calculated by the residual minimum method shown in the following formula (1).

[0049] Here, C(X+u,Y+v) in formula (1) is the difference between the luminance values before and after deformation, and Iu(x+i,y+j) and Id(X+u,Y+v) represent the luminance values before movement and after movement, respectively. In addition, X and Y are the center positions of the subset, and u and v represent the displacement amounts in the x-axis direction and the y-axis direction, respectively. The values of u and v that minimize the value of C(X+u,Y+v) were calculated, and the displacement direction and displacement amount were derived.

[0050] [Calculation of Displacement of Stainless Steel Spheres Using CT] To verify the validity of the displacement quantification method using DICM, CT images were reconstructed from the X-ray transmission images of the stainless steel spheres obtained by the displacement calculation of stainless steel spheres using DICM described above, and the three-dimensional positions of each of the three stainless steel spheres were calculated.

[0051] Figures 8 and 9 show the detection results of the center coordinates of three stainless steel spheres and the output results of the three-dimensional positions of the three stainless steel spheres in the CT images, respectively. Here, the three stainless steel spheres bonded to the resin block are defined as sphere 1, sphere 2, and sphere 3. For each frame of the transmitted image data of the captured stainless steel spheres, the center coordinates of a subset of spheres 1 to 3 were detected, and the time-series change of the center coordinates of each sphere was determined.

[0052] [Statistical Analysis] To evaluate the estimation accuracy of the method for quantifying the three-dimensional position of stainless steel spheres using DICM, the Pearson product-moment correlation coefficient (r) in each frame shown in equation (2) and the root mean square error (RMSE) shown in equation (3) were calculated for all three stainless steel spheres.

[0053]

[0054] However, ui is the center coordinate of the stainless steel sphere calculated by DICM, ▲u ̄▼ is the average value of the center coordinates of the stainless steel sphere calculated by DICM, ui' is the center coordinate of the stainless steel sphere calculated by CT, ▲ui' ̄▼ is the average value of the center coordinates of the stainless steel sphere calculated by CT, and n is the total number of data points.

[0055] Furthermore, the estimation accuracy of the method for quantifying the change in position of a stainless steel sphere in three-dimensional space using DICM was visualized using a Bland-Altman plot. In the Bland-Altman plot, the horizontal axis represents the average of the measurements from both methods, and the vertical axis represents the difference between the measurements from both methods.

[0056] The solid line and the two dashed lines on the vertical axis of the graph are defined as the mean difference between the methods and the 90% agreement limit, which indicates the acceptable margin of error relative to the mean, respectively. If the difference between the two measurement methods is within the acceptable margin of error, the two measurement methods can be considered to be consistent.

[0057] Furthermore, Bland-Altman plots can also be used to check for systematic errors such as fixed errors and proportional errors. Fixed errors occur when there is a constant offset between the measured and estimated values, and can be identified by a large deviation from zero in the mean of the difference across the entire plot. On the other hand, proportional errors occur when the error changes depending on the magnitude of the measured value, and should be suspected when the difference tends to change with a constant slope relative to the horizontal axis (the mean of the measured and estimated values).

[0058] [Time-series changes in the displacement of stainless steel spheres using DICM and CT] Figures 10 and 11 show the time-series changes in the center coordinates of spheres 1 to 3, calculated by applying DICM to transmission image data acquired by X-ray imaging, and the time-series changes in the center coordinates of spheres 1 to 3, obtained from the reconstruction of CT images, respectively. The vertical axis of the graph represents the position in each axial direction, and the horizontal axis represents the frame number, with the center coordinates of sphere 1 in the first frame being used as the origin.

[0059] As shown in Figure 10, the time-series changes in the center coordinates of each sphere, calculated using DICM, were observed. For sphere 1, the displacement in the x-axis direction initially increased and then decreased, while the displacement in the y-axis direction initially decreased and then increased. For sphere 2, the displacement in the x-axis direction initially decreased and then increased, while the displacement in the y-axis direction showed an increasing trend. Furthermore, for sphere 3, the displacement in the x-axis direction initially increased and then decreased, while the displacement in the y-axis direction showed a decreasing trend. The displacement in the z-axis direction remained almost constant for all spheres.

[0060] Regarding the displacement behavior of each sphere, it can be seen that the displacement behavior and its rate of change differ for each sphere depending on the distance from the rotation center of the rotating stage on which the acrylic block is installed to each sphere, as well as the difference in the position at the start of imaging.

[0061] In particular, sphere 1 is closest to the rotation center of the rotating stage and has a small rotation radius, so its rate of change is thought to be smaller compared to spheres 2 and 3. While these trends are observed when looking at the entire time series, when focusing on the changes in each frame, minute vibrations were observed that are thought to be due to the accuracy of sphere center position detection by DICM. The validity of the accuracy of sphere center position detection using DICM will be discussed in detail later.

[0062] Next, regarding the time-series changes in the center coordinates of each sphere calculated using CT, as shown in Figure 11, the displacement trends of spheres 1 to 3 were generally similar to those obtained using DICM. However, even in these results, minute vibrations were observed when focusing on the changes from frame to frame.

[0063] This is thought to be due to a decrease in the accuracy of sphere recognition caused by the blurring of the sphere's outline due to noise associated with X-ray imaging, as well as the influence of position determination errors due to the reconstructed CT image being represented in voxel steps. Taking these points into consideration, in the next section, we compared the two methods using the results obtained from CT and DICM, and verified the validity of the displacement quantification method using DICM.

[0064] [Validation of the proposed method by measuring the accuracy of the measurement results] Figure 12 shows the correspondence between the time-series changes in the central coordinates of spheres 1 to 3, calculated by applying DICM to transmission image data acquired by X-ray imaging, and the time-series changes in the central coordinates of spheres 1 to 3 obtained from CT reconstruction results. The vertical axis of the graph represents the time-series displacement of the central coordinates of the spheres calculated by DICM, and the horizontal axis represents the time-series displacement of the central coordinates of the spheres calculated by CT.

[0065] As shown in Figure 12, in the x-axis and y-axis directions, a difference was observed in the amount of change in the relative position of the sphere calculated by the two methods, corresponding to the magnitude of the rotation radius of the rotating stage. On the other hand, in the z-axis direction, a tendency was observed for most of the plots to be distributed near the origin of the graph.

[0066] Tables 2 and 3 show the correlation coefficient r and mean squared error (RMSE) obtained based on the correspondence between the results calculated by DICM and CT.

[0067]

[0068] The center coordinates of the stainless steel spheres calculated by DICM and CT showed a very high correlation coefficient of approximately 0.99 (p < 0.05) in the x and y directions, indicating a strong positive correlation. On the other hand, the correlation coefficient in the z direction remained at 0.07 to 0.49 (p < 0.05), and no clear correlation was confirmed.

[0069] The reason for the low correlation coefficient in the z-axis direction is thought to be the influence of the rotation direction and spatial resolution of the rotating stage during X-ray imaging. In this verification example, the motion associated with the rotation of the rotating stage mainly occurred in the horizontal direction, and the change in the height direction was extremely small. As a result, the displacement in the z-axis direction was sufficiently small compared to other directions, and it is thought that the influence of measurement errors became relatively apparent in both the DICM and CT methods, leading to a decrease in the correlation between the two methods.

[0070] On the other hand, the magnitude of the RMSE values ​​in each direction of the stainless steel sphere was very small, ranging from 2.18 to 7.82 μm, as shown in Table 3, and was below the size of one pixel in this study, which is 8.81 μm.

[0071] These results indicate that the displacement calculated by DICM was quantified with an accuracy exceeding the pixel resolution. Furthermore, the RMSE in the z-axis direction showed a similar value to that in the x-axis and y-axis directions, and no tendency for the error to increase in any particular direction was observed.

[0072] Based on these results, it is highly likely that the decrease in the correlation coefficient in the z-axis direction is not due to an increase in RMSE, but rather to the extremely small amount of displacement during X-ray imaging.

[0073] Based on the above results, the method for calculating the displacement of the strain distribution using the strain distribution testing apparatus of the present invention shows a good correspondence with the calculation results obtained by CT, and its validity has been confirmed.

[0074] [Consistency Evaluation of the Proposed Method Using Bland-Altman Analysis] Next, in order to quantitatively evaluate the consistency between DICM and CT as described above, Bland-Altman analysis was performed to examine the consistency. Figure 13 shows the Bland-Altman plot created based on a comparison of the time-series changes in the central coordinates of spheres 1 to 3, calculated by applying DICM to transmission image data acquired by X-ray imaging, and the time-series changes in the central coordinates of spheres 1 to 3 obtained from the CT reconstruction results. Tables 4 to 6 show the upper and lower limits of the 95% consistency limit in the x, y, and z axes for each stainless steel sphere, as well as the average difference between methods representing the fixed error, respectively.

[0075]

[0076] For sphere 1, the fixed errors in the x-axis and y-axis directions were 0.7 μm and -0.03 μm, respectively, which are close to zero, indicating good agreement. However, in the Bland-Altman plot in the y-axis direction, some sets of plots were observed that exceeded the upper and lower limits of the 95% agreement limit, respectively. In addition, a large positive bias of 7.51 μm was observed in the z-axis direction, and the Bland-Altman plot showed a tendency for the difference between methods to increase with increasing measured values ​​in the z-axis direction, suggesting the existence of proportional error.

[0077] For sphere 2, the fixed errors in each direction were distributed within the range of -1.39 to 1.62 μm, and since they were close to zero, the agreement between the two methods was generally good. However, similar to sphere 1, in the Bland-Altman plot in the y-axis direction, some sets of plots were observed that exceeded the upper and lower limits of the 95% agreement limit, respectively. In the plot in the z-axis direction, a tendency was observed for the difference between the methods to increase with increasing measured values, suggesting the existence of proportional error.

[0078] For sphere 3, the fixed errors in the x-axis and y-axis directions were 1.34 μm and -1.81 μm, respectively, which are close to zero, indicating good agreement between the two methods. In particular, for the Bland-Altman plot in the y-axis direction, the bias in the distribution of the mean of the difference observed in spheres 1 and 2 was not observed, and the majority of the plots were distributed within the 95% agreement limit, suggesting good correlation between DICM and CT. However, for the z-axis direction, a large bias in the positive direction was observed at 5.12 μm, and the presence of proportional error was also observed, similar to the other spheres.

[0079] Unlike imaging under visible light, transmission image data obtained by X-ray imaging integrates the absorption coefficient in the thickness direction of the object. Therefore, because structures in front of and behind the object overlap, the contrast of the brightness distribution pattern in the area to be tracked decreases, and the phenomenon of blurred edges is unavoidable. Since the search algorithm in DICM is affected by the brightness gradient, this suggests that the stability of subpixel estimation is likely to decrease significantly in low-contrast environments, which are common in transmission image data.

[0080] Furthermore, in this embodiment, a 10 μm thick GAGG scintillator was used to detect the displacement of fine patterns, and a high-resolution imaging system with a pixel resolution of 8.81 μm was constructed. Therefore, while light diffusion was suppressed and high spatial resolution was achieved, it is presumed that the decrease in X-ray stopping power caused signal disturbance due to noise, resulting in systematic errors in each direction.

[0081] Generally, the accuracy of 3D CT using X-ray imaging is high. According to a known international comparative study, a deviation evaluation comparing measurement results from 15 research institutions and companies showed that the average absolute value of the deviation in measuring the diameter of a ruby ​​sphere was 18.1 μm. Furthermore, Broden et al. constructed a 3D CT of a plastic model simulating a human pelvis with tantalum beads attached, subjecting it to translational and rotational motion, and reported an accuracy of 70–320 μm for translation and 0.21–0.82° for rotation.

[0082] Therefore, the fixed and proportional errors that occurred at the very small spatial resolution in this embodiment can be said to be good results for measuring the amount of change under X-ray imaging. In addition, in the friction test of the rubber material targeted in this embodiment, it is expected that large deformation will occur near the contact surface.

[0083] In such large deformation behaviors, the ratio of measurement error to total displacement becomes extremely small. Therefore, it is predicted that the impact of the error observed in this embodiment on the final analysis results will be reduced to a level that is practically negligible. Thus, this embodiment suggests that even in environments with physical image quality constraints such as noise and contrast, measurement with sufficient accuracy is possible for detecting the displacement of fine patterns.

[0084] [Summary of Verification Example 1] (1) The center coordinates of the stainless steel sphere calculated by DICM showed a strong positive correlation (r = 0.99) with the center coordinates calculated by CT in the x-axis and y-axis directions, and the RMSE was 2.18 to 4.02 μm. On the other hand, although the correlation coefficient was low at 0.07 to 0.49 in the z-axis direction, the RMSE was 2.97 to 7.82 μm, which was below the pixel resolution (8.81 μm). From the above, it is thought that the decrease in the correlation coefficient in the z-axis direction was not due to an increase in error, but rather because the amount of displacement in the z-axis direction due to rotation was extremely small, and the effect of measurement error became relatively apparent.

[0085] (2) The Bland-Altman plot created using the center coordinates of the stainless steel sphere calculated by DICM and the center coordinates of the stainless steel sphere calculated by CT showed that the fixed error in each direction was 1.62 to 7.51 μm, suggesting that high-precision measurement is possible.

[0086] <Verification Example 2> In Verification Example 2, we verified a method for quantifying the three-dimensional strain distribution inside a rubber material, which is an opaque material, during friction by using the displacement measurement method based on DICM described in Verification Example 1.

[0087] Generally, during the friction process, at the contact interface between rubber and the mating material, compressive deformation occurs at the leading end in the sliding direction, while recovery and separation occur at the trailing end. It is expected that the asymmetry of the stress response during this compression and recovery process contributes to the generation of a hysteresis term when friction occurs.

[0088] However, the quantitative contribution of the hysteresis term to frictional force has not yet been clarified. The main reason for this is that a method for measuring the three-dimensional strain distribution inside the material during friction has not yet been established.

[0089] [Test Method] In Verification Example 2, a friction test is performed using a rubber test piece in which metal markers are scattered on a vertical cross-section under X-ray imaging, as shown in Figure 2. In this verification example, the three-dimensional strain distribution of the rubber material is quantified by applying digital image correlation (DICM) to each of the X-ray transmission images acquired from two directions that are 90° apart from each other.

[0090] The effectiveness and measurement accuracy of this verification example are as verified in Verification Example 1. In Verification Example 2, multiple analysis regions (subsets) are set for numerous metal markers scattered on a cross-section of a rubber test piece, and the time-series change of displacement at any position within the marker layer is calculated.

[0091] Next, the three-dimensional displacement and strain distribution within the marker layer is calculated by solving simultaneous equations of the displacements obtained from the X-ray transmission images from the two directions. Since it is technically difficult to perform X-ray imaging from two directions simultaneously, in this verification example, X-ray imaging is performed in one direction at a time, and it is assumed that the strain distribution in each friction test is the same.

[0092] The details of the opaque rubber test specimens used in this verification example, as shown in Figure 2, are described below. A rectangular prism-shaped rubber test specimen was used, consisting of three parts: two triangular prism-shaped specimens with an ochre color composed solely of styrene butadiene rubber (SBR) as the base material, and one 0.12 mm thick black rubber sheet with 13.3 volume% of tungsten powder (WL, manufactured by Nippon Shinkinzoku Co., Ltd.) with a particle size of 10.0 to 40.0 μm added. The tungsten powder layer was placed on the vertical cross section along the diagonal of the rubber test specimen. Table 7 shows the dimensions and physical properties of the rubber test specimens.

[0093]

[0094] The specific procedure for preparing the rubber test specimen is as follows: (1) Approximately 7.5 g of SBR mixed with tungsten powder is rolled into a thin film using a roller, then sandwiched between two mirror-finished stainless steel plates and subjected to 200 kg / cm². 2 (1) The rubber sheet was formed by pressing it and letting it stand at 160°C for 5 minutes. (2) The rubber sheet obtained in (1) above was cut into a rectangle and placed along the hypotenuse of an isosceles triangle-shaped mold (62 mm x 5 mm). Then, 11 to 12 g of SBR without tungsten, which is the base material, was filled into the void in the mold and heated at 160°C for 5 minutes to produce a triangular prism test piece with a tungsten layer on the side. (3) The triangular prism test piece obtained in (2) above was placed in a mold with a cavity shape of a rectangular parallelepiped (62 mm x 62 mm x 5 mm), and 11 to 12 g of unvulcanized SBR without tungsten was filled so as to be in contact with the side of the tungsten layer and heated at 160°C for 5 minutes to obtain a rectangular prism-shaped rubber test piece with a tungsten layer on one cross section in the vertical direction.

[0095] This verification example was conducted at BL28B2 of the large synchrotron radiation facility SPring-8, similar to Verification Example 1. Figure 14 shows a schematic of the friction test of a resin ball (16.0 mm in diameter, precision grade, manufactured by Tenji Steel Ball Manufacturing Co., Ltd.) made of rubber and nylon 66.

[0096] A rubber test piece was fixed on the stage, and a resin ball was brought into contact with it under a vertical load of 4.90 N. A preliminary test was then conducted with 50 reciprocating movements at a sliding distance of 40 mm to stabilize the contact. Following this, the main test was performed using a linear sliding test under conditions of a sliding speed of 10.0 mm / s and a sliding distance of 40.0 mm.

[0097] In this verification example, the time when the sliding distance of the resin sphere reached 20 mm was defined as t = 0.0 s. After the start of sliding, X-ray transmission images were continuously captured up to -2.000 ≤ t < 0.750 s, which is considered to be the time when the resin sphere passed over the black layer and the rubber test piece entered the recovery process.

[0098] Furthermore, images were acquired in two separate sessions from two directions using a high-speed camera (TMX 6410, Vision Research, Inc.). The high-speed camera had a resolution of 1280 x 800 pixels, a pixel size of 8.810 μm x 8.810 μm, a field of view of 11.28 x 7.05 mm, a sampling rate of 1000 Hz, and a bit depth of 12 bits.

[0099] Similar to Verification Example 1, the magnification of the lens (AF-S Micro NIKKOR, manufactured by Nikon Corporation) was set to 1.0x, and a 10 μm thick GAGG scintillator was used as the X-ray-to-visible light conversion element. Also similar to Verification Example 1, a 16 mm aluminum plate was installed in the beryllium window to absorb high-energy synchrotron radiation.

[0100] Table 8 shows a list of the experimental conditions. The ambient temperature inside the line during the friction test was 24.8°C and the relative humidity was 35%.

[0101]

[0102] Furthermore, the frictional force F generated between the resin ball and the rubber test piece during linear sliding was measured using a force gauge (TL201Tt, manufactured by Trinity Lab) at a sampling rate of 1000 Hz, and the friction coefficient μ was calculated using the vertical load W and the frictional force from the following equation (4): μ = F / W ... (4)

[0103] [Derivation of Strain Distribution Based on DICM] Based on DICM, the in-plane strain distribution of the metal powder layer was derived from the random pattern of the metal powder layer of a rubber test specimen that deforms due to sliding. In DICM, the captured data was divided into a specified area at arbitrary intervals (number of pixels), and the displacement of each subset was calculated by setting subsets.

[0104] Furthermore, the strain distribution within the region was derived from the relative changes in the spacing between each subset. Figure 15 shows images obtained from friction tests of resin spheres and rubber test pieces, as well as examples of defined subsets. The 1.762 mm × 5.286 mm rectangular region enclosed by the frame shown in Figure 15 was used as the analysis region, with a subset size of 36 × 36 pixels (0.317 mm × 0.317 mm), a subset spacing of 13 pixels (0.115 mm), and strain calculated for each of the 752 subsets within the analysis region. The analysis conditions are summarized in Table 9.

[0105]

[0106] As the stage rotates, the subset of κ×κ pixels centered at position (xi, zi) in Figure 16(a) is displaced to position (xi+1, zi+1) as shown in Figure 16(b). In this verification example, the method proposed in Verification Example 1 was applied to the entire analysis domain to determine the displacement of all subsets within the domain.

[0107] Specifically, the region with the maximum correlation was searched for in each frame, using the luminance distribution in the subset before deformation as a reference. In doing so, the time-series positional change of each subset was calculated using the residual minimization method shown in equation (1) above. When observing within the XZ plane, the strains εx(x,z) and εz(x,z) in the x and z axes at coordinate (x,z), and the shear strain γ(x,z) were calculated using the following equations (5), (6), and (7).

[0108]

[0109] These processes were performed on all subsets within the measurement area, and the displacement and strain distributions were calculated. Similarly, when observing in the yz plane, εy(y,z), εz(y,z), and shear strain γ(y,z) were calculated.

[0110] In this verification example, the three-dimensional displacement and three-dimensional strain distribution inside the material when friction occurs in three-dimensional space were calculated by applying DICM to X-ray transmission images from two mutually orthogonal directions using the procedure described above, and then solving the two-dimensional displacements of each subset in the xz and yz planes simultaneously.

[0111] [Evaluation of data reproducibility in friction tests using a single specimen] Figure 17 shows the change in the coefficient of friction μ with respect to the sliding distance d obtained by the friction test. The vertical and horizontal axes of the graph represent μ and d, respectively, and the solid and dotted lines in Figure 17 represent the results obtained when X-ray transmission images were obtained from the x-axis direction (sliding direction) and the y-axis direction (direction perpendicular to the sliding direction), respectively.

[0112] In all friction tests, the average values ​​of the coefficient of dynamic friction for the sliding distance d = 20.0 to 30.0 mm over the metal marker layer were approximately the same, at 1.287 and 1.291, respectively. Regarding the time-series changes of each plot, μ increased in the range of d = 0.0 to 2.0 mm, and then remained roughly constant in the range of d = 2.0 to 12.0 mm. Subsequently, it showed an increasing trend again in the range of d = 12.0 to 18.0 mm, decreased gradually in the range of d = 18.0 to 26.0 mm, and then remained roughly constant again.

[0113] These results indicate that the reproducibility of the friction behavior was sufficiently high regardless of the X-ray imaging direction. Furthermore, the increasing and decreasing trends in the range d = 12.0 to 26.0 mm generally coincide with the slip distance before and after passing over the metal marker layer. This is thought to be due to the difference in physical properties between the portion composed solely of SBR and the portion with the metal marker.

[0114] As shown in Table 7 above, the elastic modulus of the metal marker layer was 6.17 MPa (calculated from the stress-strain curve measured based on JIS K6251:2017) for a strain of 0-5%, while the elastic modulus of the base material was 3.82 MPa, indicating a difference in Young's modulus between the two.

[0115] In other words, it is presumed that the metal marker layer had higher rigidity compared to the rubber alone, causing a momentary increase in frictional force when passing directly over the metal marker layer compared to other regions. In addition, the observation that the coefficient of friction gradually asymptotically approaches the value before entering the layer after approximately d = 20.0 mm, where the metal marker layer is thought to have been passed, is an important result that clearly supports this behavior.

[0116] [Relationship between distance from the contact point and sliding direction on displacement within the specimen during sliding friction between a rubber specimen and a resin ball] Here, we focus on the three-dimensional displacement at characteristic points near the contact area (a total of four subsets: the contact point and the surface layer and directly below its edges) during the friction test between the resin ball and the rubber specimen shown in Figure 18.

[0117] Subsets A and B are defined as the vicinity of the analysis region edge and the contact point on the surface of the specimen, respectively, while Subsets C and D are defined as the vicinity of the analysis region edge and the contact point directly beneath the specimen.

[0118] Figures 19 and 20 show the time-series changes in the positions of each subset within the xy plane. The vertical and horizontal axes of these graphs represent the xy, xz, and yz planes on the surface and directly beneath the specimen, respectively. With friction, the time at which the sliding distance reaches 20 mm is set to t = 0.000 s, and the plots of each subset in Figures 19 and 20 represent the trajectories at -2.000 ≤ t ≤ 0.750 s.

[0119] Furthermore, the coordinate system used in this analysis was defined as the contact point (x, y) = (0, 0) between the resin sphere and the rubber test piece at t = 0.000 s, and for the vertical direction, the upper surface of the test piece before the start of the sliding test was defined as z = 0.

[0120] In Figures 19 and 20, Subset A and Subset C underwent significant displacement in the positive x-axis direction as the resin sphere approached the metal marker layer. Subsequently, they followed roughly similar paths, displacing to approach their initial positions. On the other hand, Subset B and Subset D, during the same approach process, underwent displacement in both the positive x-axis and negative y-axis directions, and then exhibited behavior of approaching their initial positions while tracing a counterclockwise arc.

[0121] The differences in displacement behavior during friction are thought to be due to the distance from the contact point between the resin sphere and the rubber test piece. In Subset A, located near the contact point, the influence of adhesion between the resin sphere and the rubber test piece at the contact interface was dominant, and it is thought that the region around the contact point followed the movement of the resin sphere, resulting in a large displacement only in the x-axis direction.

[0122] In contrast, Subset B, although located near the surface of the specimen, was situated approximately 2.6 mm away from the contact point in the negative y-axis direction. Therefore, as the resin sphere approached, it was displaced outside the trajectory of the contact point. This difference in deformation modes resulted in a difference in the trajectories of the subsets during the compression and recovery processes. It is thought that this asymmetry in the trajectories led to energy dissipation, resulting in the emergence of the hysteresis term.

[0123] These differences in displacement showed a similar trend for Subset C and Subset D, which are located inside the rubber specimen. However, it is thought that the influence of adhesion, as mentioned above, decreased significantly as the distance from the contact interface in the depth direction increased.

[0124] In addition, the stress generated by the penetration of the resin sphere is thought to have been dispersed and dissipated into the surroundings as it propagated inward, thus dissipating the effect of the vertical load extending to the deeper parts. Therefore, the displacement amounts in each direction of Subset C and Subset D were considered to be small compared to the subset of the rubber test specimen surface layer in both directions.

[0125] Figures 21 and 22 show the time-series changes in the positions of each subset in the xz plane. Subset A and Subset C were displaced in the positive x-axis direction as the sliding began, and were further displaced significantly in the negative z-axis direction as the resin sphere approached the metal marker layer. Subsequently, they folded back and returned to their initial positions along almost the same path, but while the x-axis displacement recovered almost completely to near the initial position, the z-axis displacement remained at a deeper position than the initial position, meaning that the vertical displacement did not fully recover.

[0126] On the other hand, in Subset B and Subset D, similar to Subset A and Subset C, a displacement in the positive x-axis direction occurred at the start of slip, followed by a large displacement in the z-axis direction. In both cases, the displacement progressed in a clockwise arc and converged towards the starting point relative to the z-axis. A common phenomenon in these results is the displacement in the positive x-axis direction accompanying the start of slip.

[0127] The likely cause of these phenomena is the shear effect on the entire specimen due to friction. Specifically, it is thought that shear occurred in the tangential direction of the contact interface when the resin sphere and the specimen came into contact and rubbed together, causing the specimen itself to deform in the positive x-axis direction. In addition, in the yz plane (Figures 21, 22), unlike in the xy plane (Figures 19, 20), the trajectories of the subset took on a loop shape.

[0128] These results suggest that the friction factors contributing to the displacement of the subsets differed between the yz-plane and the xy-plane. Furthermore, for the subsets on the surface of the specimen (Subset A and Subset B), although the displacement was larger than in the interior of the rubber specimen due to a greater influence of the adhesion term, the projected area of ​​the trajectory loop was smaller. This result is presumed to be because the forward and return paths were fixed by constraint due to contact with the resin spheres.

[0129] On the other hand, the subsets within the specimen (Subset C and Subset D) tended to have smaller displacements and larger projected areas of the trajectory loops. Furthermore, the larger projected area of ​​the trajectory loops suggests that the difference between the compression and recovery pathways was more pronounced, and therefore may have contributed more significantly to the emergence of the hysteresis term.

[0130] Figures 23 and 24 show the time-series changes in the positions of each subset in the yz plane. Subsets A and C show that the resin spheres were significantly displaced in the negative z-axis direction as the metal marker layer approached, tracing trajectories that approached their initial positions along a path almost identical to that of the compression process.

[0131] In contrast, Subset B and Subset D, upon the onset of sliding, were displaced in the negative direction of the y-axis and simultaneously in the negative direction of the z-axis, returning to their initial positions while tracing an arc counterclockwise. This revealed that, similar to the xy and xz planes, differences in the general shape and deformation of the subset's trajectory occur in the yz plane depending on the relative positional relationship with the contact point.

[0132] In particular, within the yz plane, a tendency for radial displacement relative to the contact point with the resin sphere was observed. Subsets A and C, located directly below the contact point, were displaced linearly in the vertical direction. In contrast, in Subsets B and D, located at a distance from the contact point, it is presumed that the stress within the material propagated horizontally (y-axis direction), causing the rubber test piece to be pushed out radially, resulting in an increase in the displacement component in the y-axis direction. This occurrence of displacement in the y-axis direction is thought to have manifested as a difference in the trajectory due to the asymmetry of stress during the compression and recovery processes.

[0133] These results suggest that the asymmetrical distribution of trajectories in each subset with respect to each axis influences the distribution of contributions to the hysteresis term.

[0134] [Relationship between distance from contact point and sliding direction on strain inside a test specimen during sliding friction between a rubber test specimen and a resin ball] This section focuses on the three-dimensional strain distribution in the analysis region within the frame shown in Figure 15. Figures 25 to 33 show the three-dimensional displacement and strain distribution in the analysis region before and after the resin ball passes over the metal marker layer.

[0135] In the graphs in Figures 25-33, the time at which the sliding distance of the resin ball reaches 20 mm is defined as t = 0.000 s, and the graphs show the normal strain and shear strain in each axial direction at t = 0.3000 s, 0.000 s, and 0.300 s, respectively. For each axis, the contact point between the resin ball and the rubber test piece at t = 0.000 s was set to (x, y, z) = (0, 0, 0). In addition, the index bars in Figures 25-33 represent the displacement amount for each subset for displacement, the magnitude of the strain for normal strain, and the direction and degree of shear strain for shear strain.

[0136] Figure 25(a) shows that the entire analysis region was displaced in the positive x-axis direction. The displacement in the x-axis direction showed a tendency to gradually increase from the start of sliding. From these results, it can be inferred that shear strain occurred throughout the rubber test specimen upon the start of frictional force application.

[0137] Furthermore, in the y-direction (Figure 25(a)) and z-direction (Figure 25(c)), displacements in the positive and negative directions were observed at one end of the analysis region, respectively. This is presumed to be the result of the bulk of the rubber test piece being extruded radially along all axes due to the penetration of the resin sphere as it approached the contact point.

[0138] Furthermore, Figures 26(d) and 26(f) can be seen to confirm that perpendicular strain in the x-direction and compressive strain in the z-direction occurred near the surface of the test specimen. This result is thought to be due to the localized application of frictional force and perpendicular load directly beneath the contact point.

[0139] Furthermore, as shown in Figure 34(c) and the time-series change of the coefficient of friction during the friction test (Figure 17), the metal marker layer had a higher elastic modulus compared to the rubber base material. As a result, the low-elasticity rubber base material was sandwiched between the resin ball and the metal marker layer during friction, causing slightly asymmetrical displacement and strain in the y-axis direction. In some areas near the surface, a phenomenon of the rubber material surface bulging could be observed at -0.300 s ≤ t ≤ 0.000 s.

[0140] As shown in Figure 28(a), a positive displacement in the x-axis direction was observed throughout the entire analysis region, similar to Figures 25-27. Regarding the y-axis direction, as shown in Figures 25(a), 25(b), and 25(c), the displacement was negative when y≧0μm and positive when y≦0μm, while the displacement in the z-axis direction was significantly negative. Regarding the normal strain in each direction, as shown in Figures 26(d), 26(e), and 26(f), tension and compression occurred at the front and rear ends of the sphere in the x-axis direction, and tension and compression occurred in the y-axis and z-axis directions, respectively. These results are presumed to be strains resulting from displacement along the surface shape of the sphere.

[0141] Furthermore, the results in Figures 27(g) and (h) show that in the surface layer of the specimen in the analysis region, shear strain occurred in opposite directions at the front and rear ends of the resin sphere surface in both the xy and yz planes. This is thought to be due to the deformation of the rubber specimen along the surface of the resin sphere. In contrast, asymmetric shear strain was observed in the zx plane. This is thought to be because the stress state at the contact interface as the resin sphere passed through became asymmetrical at the front and rear of the sphere. At the front end of the sphere, deformation of the specimen due to vertical load was dominant, while at the rear end of the sphere, the surface layer of the specimen was dragged horizontally due to adhesion. Therefore, it is inferred that a significant shear strain distribution in that direction was formed as a result of the combined action of these different deformation modes.

[0142] Furthermore, in Figures 31(a), 31(b), and 31(c), a continuous positive displacement in the x-axis direction was observed within the analysis region, while positive displacement in the y-axis direction and negative displacement in the z-axis direction were observed moving towards the opposite end of the analysis region. Similar trends were observed for the vertical strain in each axial direction and the shear strain within each plane as in Figures 28-30, but their absolute values ​​decreased, suggesting that the distance from the contact point between the resin sphere and the rubber test piece had a significant influence on the strain.

[0143] Figures 34-41 show the time-series changes in vertical strain and shear strain in Subsets A-D. The vertical axis represents the magnitude of each strain, and the horizontal axis represents time. In addition to the absolute values ​​of the strains during the compression and recovery processes in each subset, a clear positional dependence can be observed in the slope of the graph, i.e., the strain rate.

[0144] This was particularly noticeable in the z-axis strain of the subsets (Subset A, Subset B) directly below the contact point. Furthermore, they exhibited different behavior before and after passing through the contact area. This result is consistent with the "asymmetry of the subset trajectories during the compression and recovery processes" described in the previous section. This asymmetry suggests the emergence of a hysteresis term due to viscoelastic properties and supports the occurrence of energy dissipation during the compression and recovery processes.

[0145] Figures 42 to 45 show the displacement distribution in the three axes, the vertical strain distribution in the three axes, and the shear strain distribution within the three planes in the xy plane, including Subset A, Subset B, and Subset C, Subset D, respectively. Figures 46 to 49 show the displacement distribution in the three axes, the vertical strain distribution in the three axes, and the shear strain distribution within the three planes in the zx plane, including Subset A, Subset C, and Subset B, Subset D, respectively.

[0146] These results were calculated based on the relative positional relationship between the resin sphere and the metal marker layer, after applying an offset process equivalent to the slip distance to the strain distribution of the metal marker layer obtained in each frame. As in the previous section, the coordinates of the contact point between the resin sphere and the rubber test piece at t = 0.000 s are set to (x, y, z) = (0, 0, 0). Furthermore, the index scales in each figure represent the amount of movement in each subset for displacement, the magnitude of strain for normal strain, and the magnitude and direction of deformation for shear strain.

[0147] In Figures 42-45, the absolute values ​​of each displacement and strain consistently showed larger values ​​at the surface compared to the interior of the specimen. This indicates that the effects of the vertical load from the resin spheres and the frictional force (horizontal force) at the contact interface are significantly more pronounced in the surface layer of the specimen.

[0148] On the other hand, focusing on the distribution of each parameter, it was observed that the displacement and strain regions extended diagonally with respect to the direction of motion of the resin sphere. This is presumed to be due to the 45° angle between the resin sphere and the test specimen, as well as the difference in elastic modulus between the metal marker layer and the rubber matrix. In other words, the difference in stiffness between the metal marker layer and the rubber matrix caused a change in the strain distribution, which would have been parallel to the direction of motion if the stiffness distribution were uniform. Therefore, it is suggested that matching the stiffness of the metal marker layer and the rubber matrix may allow for obtaining analytical results for friction in uniform materials.

[0149] Furthermore, as shown in Figures 46-49, the absolute values ​​of displacement and strain in each axial direction were largest directly below the contact point between the resin sphere and the rubber test piece. While these spatial distributions reflect the shape of the sphere, they were not perfectly concentric, but rather showed an elliptical distribution elongated in the sliding direction. This suggests that a stress difference occurred between the compression process at the leading end of the sliding direction and the delamination and recovery process at the trailing end. In other words, this elliptical distribution of parameters is likely to most clearly determine the emergence of the hysteresis term.

[0150] <Summary of Examples> Based on the results of the above-described examples, the strain distribution inside the rubber material during friction is asymmetrical with respect to all axial directions, and it is thought that the hysteresis term can be determined by geometric and positional factors such as the distance from the contact point and the sliding direction.

[0151] Rubber materials are used in a wide range of products, such as automobile tires and shoe soles. During these uses, the rubber material comes into contact with the mating material, generating friction at the interface. The resulting frictional force is generally expressed as the sum of the adhesion term and the hysteresis term. However, the proportion of each term's contribution varies depending on the surface properties of the mating material, sliding conditions, lubrication conditions, and other factors.

[0152] It is known that rubber materials exhibit high friction due to an increase in the true contact area and hysteresis loss, respectively, resulting in increased adhesion and hysteresis terms. In particular, because rubber materials have low elasticity and viscoelasticity, the contribution of the hysteresis term, which is caused by the stress difference between the compression and recovery processes, is considered to be larger compared to hard materials such as metals and ceramics.

[0153] Therefore, in order to elucidate the friction mechanism in rubber, it is essential to clarify the strain behavior during friction, and for that purpose, the quantification of the three-dimensional strain distribution is required. Accordingly, in the embodiments of the present invention, we focused on the friction between a rubber block and a resin ball and attempted to quantify the three-dimensional strain distribution of the rubber material during friction.

[0154] First, we verified a method for quantifying the spatial position of metal markers by applying digital image correlation (DICM) to X-ray transmission images. In addition, we conducted friction tests between a rubber test piece with a marker layer inside and a resin sphere using this method. At this time, we attempted to experimentally quantify the three-dimensional distribution of displacement and strain in the marker layer by X-ray imaging. The findings obtained in this embodiment are described below.

[0155] (1) X-ray imaging was performed from two directions, and the center coordinates of the stainless steel sphere calculated by DICM showed a strong positive correlation (r = 0.99) in the x and y directions with the center coordinates of the stainless steel sphere calculated from CT based on X-ray imaging results from 360 degrees, with RMSE values ​​ranging from 2.18 to 4.02 μm. In the z direction, the RMSE values ​​ranged from 2.97 to 7.82 μm.

[0156] Regardless of the axial direction, the positional measurement accuracy of the stainless steel sphere was less than or equal to the resolution (voxel size) of the X-ray imaging, which is 8.81 μm. Furthermore, the fixing error in each axial direction was found to be 1.62 to 7.51 μm, as shown by the Bland-Altman plot created using the center coordinates of the stainless steel sphere calculated by DICM and the center coordinates of the stainless steel sphere calculated by CT.

[0157] (2) Metal markers placed on the cross-section of a rubber test piece that rubbed against a resin sphere were observed by X-ray imaging, and the spatial positions (3D positions) of 752 minute regions (subsets) within each marker layer were quantified based on DICM, and the 3D strain distribution was calculated from the relative position changes.

[0158] As a result, the displacement and strain in each subset changed asymmetrically in each axial direction and within each plane before and after the resin sphere passed over the metal marker layer, and the degree of this change depended on the position in the rubber specimen.

[0159] (3) The three-dimensional asymmetric strain distribution obtained from these results is considered to determine the appearance of the hysteresis term. Therefore, it has been confirmed that the strain distribution testing apparatus of the present invention is capable of measuring a strain distribution that allows observation of the deformation behavior due to strain inside the test specimen.

[0160] 10...Strain distribution testing apparatus 11...Stage 12...Pressing section 12a...Indenter 13...Imaging section 13a, 13b...X-ray source 14...Control unit D...Observation particle F...Measurement surface M...Moving surface S...Test piece

Claims

1. A strain distribution testing apparatus capable of observing deformation behavior due to strain inside a test specimen, wherein a plurality of observation particles that do not transmit X-rays are arranged inside the test specimen so as to fit within a measurement surface which is an arbitrary plane, and the apparatus comprises: a pressing unit that is movable along a moving surface and presses the measurement surface, an imaging unit capable of capturing a transmitted image of the observation particles arranged in the test specimen, and a control unit that controls the pressing unit and the imaging unit, wherein the measurement surface is arranged perpendicular to the moving surface, and the control unit presses the measurement surface so that the pressure applied by the pressing unit is perpendicular to the moving surface, moves the pressing unit, irradiates the test specimen with X-rays from two intersecting directions, deforms the test specimen by the pressing force of the pressing unit, and measures the change in position of the observation particles inside the test specimen caused by the deformation of the test specimen.

2. The strain distribution testing apparatus according to claim 1, wherein the test specimen includes a material that transmits X-rays.

3. The strain distribution testing apparatus according to claim 1 or 2, wherein the observed particles include a metal with an atomic number of 26 or higher.

4. The strain distribution testing apparatus according to claim 1 or 2, wherein the material constituting the test specimen includes an element with an atomic number of 14 or less.

5. The strain distribution testing apparatus according to claim 1 or 2, wherein the X-rays are white X-rays.