Test piece analysis device and test piece analysis method

Abstract

To analyze a test piece on the basis of a result of analysis from an initial crack generated in a surface of the test piece to a crack immediately before the test piece ruptures in fatigue testing.SOLUTION: An imaging unit creates picked-up images by picking up an image of a surface of a test piece at the timing when the surface of the test piece is irradiated with light from one direction and the timing when the surface is irradiated with light from the other direction. An arithmetic processing unit specifies a crack in the picked-up images, extracts areas in the picked-up images including the specified crack as crack observation images, pays attention to, of the extracted crack observation images, a crack observation image immediately before rupture capturing a crack that can be considered as a crack immediately before the test piece ruptures, goes back to find a past crack observation image obtained by picking up an image of a position corresponding to a position on the surface of the test piece captured in the crack observation image immediately before rupture, from the crack observation image immediately before rupture to an initial crack observation image capturing an initial crack that is the starting point from which the crack is generated, and thereby analyzes the crack generated in the surface of the test piece.SELECTED DRAWING: Figure 16

Description

[Technical Field]

[0001] The present invention relates to a test specimen analysis device and a test specimen analysis method for analyzing a test specimen based on cracks that occur on the surface of the test specimen during a fatigue test in which the temperature of the test specimen is maintained within a set temperature range below 700°C. [Background technology]

[0002] During fatigue testing, cracks that develop on the surface of a specimen are observed by stopping the testing machine after each cycle, both visually by an operator and by microscopic observation using the replica method. However, cracks are typically only a few tens of micrometers in size immediately after initiation, making them difficult to detect even with a microscope. In the early stages of fatigue testing, it is difficult to determine the location and presence of cracks. When replicas of the specimen surface are taken after each cycle, cracks that are visible under a microscope using the replica method only grow to a size of several hundred micrometers to a few millimeters, making it possible to identify their location. By reexamining the initial replicas taken from the position of visible cracks, the initial crack and its subsequent growth rate can be investigated. However, microcracks such as initial cracks are difficult to transfer to the replica, making it difficult to identify them. Furthermore, air bubbles may appear in the replica, making observation of the specimen surface itself difficult.

[0003] The replica method requires advanced technology to accurately transfer initial cracks to the replica, and replica fabrication is time-consuming. Furthermore, the process requires careful checking for air bubbles in the replica, which can take more than an hour in some cases. Furthermore, replica fabrication is a part of the fatigue testing process, so replica fabrication after each cycle is a post-cycle operation. After replica fabrication, fatigue testing is continued until the next cycle. This requires repeated cycles and replica fabrication, resulting in a significant amount of time required for fatigue testing of a single specimen. Furthermore, even if large cracks are identified in the later stages of fatigue testing, failure to transfer the initial cracks often renders the fatigue test itself useless. While the replica method is extremely difficult, identifying cracks on the surface of a specimen after each cycle and evaluating their propagation is crucial for accurately determining the lifespan and strength of the product corresponding to the specimen. Therefore, underestimating crack size or overlooking cracks is essential.

[0004] The following methods have been proposed as techniques for observing cracks in a test specimen. Patent Document 1 discloses a method in which a dot pattern or grid pattern is applied to the surface of a test specimen in advance, and the strain on the surface of the test specimen is measured using a moire method and an image correlation method based on changes in the pattern during a fatigue test. Patent Document 2 discloses a method in which a one-dimensional or two-dimensional pattern (grating) is applied to the surface of a test specimen in advance, and the strain on the surface of the test specimen is measured using a Fourier transform and phase change based on changes in the pattern during a fatigue test.

[0005] Furthermore, the following method has been proposed as a method for measuring the length of a crack: Patent Document 3 proposes a method in which the tip of a crack is detected based on an image captured by an imaging device, at least one of the imaging device and the test piece support device is moved relative to the other by a moving device so that the field of view of the imaging device moves relative to the test piece in the direction of crack growth, and the growth dimension of the crack is calculated based on the amount of relative movement of the imaging device and the test piece support device. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Laid-Open No. 2011-27526 [Patent Document 2] Japanese Patent Application Publication No. 2019-66369 [Patent Document 3] Japanese Patent Application Publication No. 5-223717 Summary of the Invention [Problem to be solved by the invention]

[0007] However, the methods described in Patent Documents 1 and 2 have a problem in that the pattern peels off from the surface of the test specimen during fatigue testing. Furthermore, the method described in Patent Document 3 is premised on detecting the tip of a crack, and therefore has a problem in that it is not possible to detect the tip of an initial crack. Therefore, there is room for improvement in order to be able to analyze a test specimen based on the results of analyzing the initial crack that appears on the surface of the test specimen during fatigue testing, up to the crack that occurs immediately before the test specimen breaks.

[0008] The present invention aims to provide a test specimen analysis device and a test specimen analysis method that can analyze a test specimen based on the results of analyzing the initial crack that appears on the surface of the test specimen during fatigue testing, from the crack that appears just before the test specimen breaks. [Means for solving the problem]

[0009] A first aspect of the present invention is a test piece analyzing device that analyzes a test piece based on cracks that occur on the surface of the test piece in a fatigue test that is conducted while maintaining the temperature of the test piece within a set temperature range below 700°C, the device comprising: a stress applying unit that repeatedly applies tensile stress and compressive stress in a predetermined cycle in the axial direction of the test piece as a load in the fatigue test; an illumination unit that alternately irradiates the surface of the test piece with light from one direction and light from another direction across a plane perpendicular to the axis of the test piece; an imaging unit that uses one or more area cameras provided opposite the surface of the test piece to capture images of the surface of the test piece from a normal direction to the surface of the test piece and generate captured images; and a calculation processing unit that uses the captured images generated by the imaging unit to analyze cracks that occur on the surface of the test piece and analyzes the test piece based on the analysis results, and the imaging unit is configured to capture images of the surface of the test piece from a normal direction to the surface of the test piece when the application of stress by the stress applying unit is stopped during the fatigue test. The illumination unit is disposed between the imaging unit and the test piece, and generates the captured image by imaging the surface of the test piece at each of the timings when light from the one direction is irradiated onto the surface of the test piece and the timing when light from the other direction is irradiated onto the surface of the test piece. The calculation processing unit has: a crack observation image extraction unit that identifies cracks in the captured image and extracts an area in the captured image that includes the identified crack as a crack observation image; and a crack analysis unit that focuses on a crack observation image just before fracture, among the crack observation images extracted by the crack observation image extraction unit, that captures a crack that can be considered to be a crack that occurs just before the test piece fractures, and analyzes cracks that occur on the surface of the test piece by tracing back from the crack observation image just before fracture to an initial crack observation image that captures an initial crack that is the starting point for crack generation.

[0010] A second aspect of the present invention is a test specimen analysis method for analyzing a test specimen based on cracks that occur on the surface of the test specimen in a fatigue test that is conducted while maintaining the temperature of the test specimen within a set temperature range below 700°C, the method including: a stress application unit that repeatedly applies tensile stress and compressive stress in a predetermined cycle in the axial direction of the test specimen as a load in the fatigue test; an illumination unit that alternately irradiates the surface of the test specimen with light from one direction and light from another direction across a plane perpendicular to the axis of the test specimen; an imaging unit that uses one or more area cameras provided opposite the surface of the test specimen to capture images of the surface of the test specimen from a normal direction to the surface of the test specimen and generate captured images; and a calculation processing unit that uses the captured images generated by the imaging unit to analyze cracks that occur on the surface of the test specimen and analyze the test specimen based on the analysis results, and the method includes using the imaging unit to analyze cracks that occur on the surface of the test specimen based on the analysis results when stress application by the stress application unit is stopped in the fatigue test. the illumination unit is disposed between the imaging unit and the test piece, and the surface of the test piece is imaged at each of the timings when light from the one direction is irradiated onto the surface of the test piece and the timing when light from the other direction is irradiated onto the surface of the test piece, thereby generating the image; a crack observation image extraction step of identifying a crack in the image and extracting an area in the image containing the identified crack as a crack observation image; and a crack analysis step of analyzing a crack occurring on the surface of the test piece by focusing on a crack observation image immediately before fracture that captures a crack that can be considered to be a crack that occurs immediately before the test piece fractures, among the crack observation images extracted in the crack observation image extraction step, and tracing back from the crack observation image immediately before fracture to an initial crack observation image that captures an initial crack that is a starting point for crack generation. [Effects of the Invention]

[0011] According to the present invention, it is possible to provide a test specimen analysis device and a test specimen analysis method that can analyze a test specimen based on the results of analyzing the initial crack that appears on the surface of the test specimen during fatigue testing, from the crack that occurs just before the test specimen breaks. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a block diagram showing an example of the configuration of a test piece analyzing device according to an embodiment of the present invention. [Figure 2] FIG. 4 is a front view showing an example of the configuration of a stress applying portion. [Figure 3] FIG. 2 is a perspective view showing an example of the configuration of an imaging unit. [Figure 4] FIG. 2 is a plan view showing an example of the configuration of an imaging unit. [Figure 5] FIG. 2 is a perspective view showing an example of the configuration of an illumination unit and an imaging unit. [Figure 6] FIG. 2 is a side view showing an example of the configuration of a lighting device and a camera. [Figure 7] 10A and 10B are explanatory diagrams showing examples of captured images captured at a timing when light is irradiated from one direction and at a timing when light is irradiated from another direction, in comparison with each other; [Figure 8] FIG. 2 is a block diagram showing an example of the configuration of a calculation processing unit. [Figure 9] FIG. 10 is an explanatory diagram showing an example of captured images stored for each predetermined cycle of a fatigue test. [Figure 10] FIG. 10 is a block diagram showing an example of a processing flow from a multi-resolution processing unit to a crack analysis unit. [Figure 11] 10A and 10B are explanatory diagrams showing an example of multi-resolution processing by a multi-resolution processing unit and enhancement processing by an enhancement processing unit; [Figure 12] FIG. 10 is an explanatory diagram showing an example of an enhanced image generated immediately before the test specimen breaks. [Figure 13] FIG. 10 is an explanatory diagram showing an example of tracing back past crack observation images by the crack analysis unit. [Figure 14] FIG. 10 is an explanatory diagram showing an example of a manner in which the crack extraction region is updated in sequence. [Figure 15]FIG. 10 is an explanatory diagram showing an example of an enhanced image generated at the beginning of a fatigue test. [Figure 16] 1 is a flowchart showing an example of the flow of a test piece analysis method according to an embodiment of the present invention. [Figure 17] 10 is a graph showing an example of the results of detecting the length of a main crack using a test specimen analysis method according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0013] FIG. 1 shows an example of the configuration of a test specimen analyzing apparatus 10 according to one embodiment of the present invention. The test specimen analyzing apparatus 10 is an apparatus that analyzes a test specimen 80 based on cracks that appear on the surface of the test specimen 80 in a fatigue test that is performed by maintaining the temperature of the test specimen 80 within a set temperature range below 700°C (i.e., a temperature range corresponding to cold treatment). The test specimen 80 is made of a metal such as steel or an aluminum alloy, and is formed in a cylindrical shape. However, the test specimen 80 may be formed in a shape other than a cylindrical shape, such as a square pillar. The surface of the test specimen 80 refers to the surface around the axial direction of the test specimen 80 (i.e., the outer circumferential surface).

[0014] More specifically, the fatigue test is a test in which one cycle consists of a tension process in which the tensile stress acting in the axial direction of the test specimen 80 is gradually increased, and a compression process in which the compressive stress acting in the axial direction of the test specimen 80 is gradually increased, and this cycle is repeated multiple times. The time for the tension process, the time for the compression process, the tensile load, and the compressive load are predetermined to be the same for each cycle. The fatigue test may be terminated when the number of cycles reaches a predetermined number, or may be terminated when the test specimen 80 breaks due to the tensile stress or buckles due to the compressive stress.

[0015] The specimen analysis device 10 includes a stress applying unit 12, an illumination unit 14, an imaging unit 16, a control unit 18, and a calculation processing unit 20. The stress applying unit 12 repeatedly applies tensile stress and compressive stress in the axial direction of the specimen 80 in the predetermined cycle as a load in a fatigue test. The illumination unit 14 illuminates the surface of the specimen 80 when the stress applying unit 12 stops applying stress for each predetermined cycle. The imaging unit 16 captures images of the entire circumference and length of the surface of the specimen 80 while the surface of the specimen 80 is illuminated by the illumination unit 14, and generates the captured images. The calculation processing unit 20 uses the captured images generated by the imaging unit 16 to analyze cracks occurring on the surface of the specimen 80 and analyzes the specimen 80 based on the analysis results. The timing for capturing images of the surface of the specimen 80 under illumination may be after each cycle or after a preselected cycle from among multiple cycles.

[0016] FIG. 2 shows an example of the configuration of the stress applying unit 12. The stress applying unit 12 has a first gripping unit 26, a second gripping unit 28, and a driving unit 30. The first gripping unit 26 and the second gripping unit 28 are configured as chuck mechanisms. The first gripping unit 26 grips one axial end of the test specimen 80, and the second gripping unit 28 grips the other axial end of the test specimen 80. The first gripping unit 26 is connected to the driving unit 30, and the second gripping unit 28 is fixed to a test table 32. The driving unit 30 moves the first gripping unit 26 in the axial direction of the test specimen 80, thereby repeatedly applying tensile stress and compressive stress in the axial direction of the test specimen 80. The driving unit 30 is configured with, for example, a reduction mechanism, a motor, etc.

[0017] 3 and 4 show an example of the configuration of the imaging unit 16. The imaging unit 16 has multiple cameras 34. Each camera 34 is an area camera (e.g., a monochrome area camera). Each camera 34 is installed, for example, so that the horizontal direction of the camera 34 is parallel to the axial direction of the test piece 80. The number of multiple cameras 34 is, for example, four. The four cameras 34 are provided facing the surface of the test piece 80 and are arranged at 90° intervals around the periphery of the test piece 80. The four cameras 34 capture images of the surface of the test piece 80 from the normal direction of the surface of the test piece 80, and each generate a captured image. The field of view of each camera 34 is set so that the four cameras 34 can capture images of the entire periphery and entire length of the surface of the test piece 80.

[0018] Note that, as long as the four cameras 34 can capture images of the entire circumference and entire length of the surface of the test piece 80, the four cameras 34 may be arranged around the test piece 80 at angular intervals other than 90°. The number of cameras 34 may be one or may be any number other than four. Furthermore, each camera 34 may be installed, for example, so that the horizontal direction of the camera 34 is perpendicular to the axial direction of the test piece 80. Hereinafter, when it is necessary to distinguish between the multiple cameras 34 in the description, the multiple cameras 34 will be referred to as cameras 34A to 34D.

[0019] FIG. 5 shows an example of the configuration of the illumination unit 14 and the imaging unit 16. The illumination unit 14 has a pair of illumination devices 110. The pair of illumination devices 110 are arranged opposite each other across the test specimen 80 in a direction perpendicular to the axial direction of the test specimen 80. Each illumination device 110 has a housing 112 and a pair of buffer materials 114. The housing 112 is formed in a rectangular frame shape. The housing 112 has a window portion 116 that penetrates in a direction perpendicular to the axial direction of the test specimen 80. The pair of buffer materials 114 are arranged side by side in a direction parallel to the axial direction of the test specimen 80. The pair of buffer materials 114 are provided on the outside of the housing 112 (for example, on the surface of the housing 112 facing the test specimen 80). Each buffer material 114 is made of, for example, a polyurethane sponge or a rubber elastic material.

[0020] The lighting unit 14 (pair of lighting devices 110) is supported by a support mechanism (not shown) and is configured to be movable around the test specimen 80 while maintaining a constant distance from the test specimen 80 by contacting the test specimen 80 via a buffer material 114. The lighting unit 14 is also disposed between the imaging unit 16 (plurality of cameras 34) and the test specimen 80, and is configured to be movable between the imaging unit 16 and the test specimen 80. A drive unit 118 is connected to the lighting unit 14. The drive unit 118 is configured to move the lighting unit 14 along the periphery of the test specimen 80. The drive unit 118 is configured to have, for example, a reduction mechanism, a motor, etc.

[0021] The driving unit 118 is controlled by the control unit 18 (see FIG. 1 ) to first move the pair of lighting devices 110 to a position where the pair of housings 112 directly face two of the four cameras 34, 34A and 34C, which face each other, and then move the pair of lighting devices 110 to a position where the pair of housings 112 directly face two of the four cameras 34, 34B and 34D, which face each other. When the housings 112 directly face each camera 34, the cameras 34 are able to capture an image of the surface of the test piece 80 through a window 116 inside the housings 112. Note that although the lighting unit 14 includes a pair of lighting devices 110, it may also include a single lighting device 110.

[0022] FIG. 6 shows an example of the lighting device 110 and the camera 34. The lighting device 110 has a pair of optical systems 120 and a pair of light sources 122. The pair of optical systems 120 are arranged side by side in a direction parallel to the axial direction of the test piece 80. The pair of optical systems 120 are provided inside the housing 112. Each optical system 120 has a prism 124 or the like. Each optical system 120 is configured to emit light emitted from the light source 122 toward the surface of the test piece 80. The pair of light sources 122 are arranged on the opposite side of the pair of optical systems 120 from the test piece 80. One light source 122 is arranged at a position where it emits light toward one optical system 120, and the other light source 122 is arranged at a position where it emits light toward the other optical system 120. Each light source 122 is, for example, an LED (Light Emitting Diode) bulb or an incandescent bulb.

[0023] The lighting device 110 is configured to be switchable between a state in which one light source 122 emits light and a state in which the other light source 122 emits light. By switching between a state in which one light source 122 emits light and a state in which the other light source 122 emits light, the lighting device 110 alternately irradiates the surface of the test piece 80 with light from one direction and light from another direction across a plane perpendicular to the axis of the test piece 80. While facing the housing 112, each camera 34 captures an image of the surface of the test piece 80 at the timing when light is irradiated from one direction and the timing when light is irradiated from the other direction, and generates each captured image.

[0024] 7A and 7B show examples of images captured when light is irradiated from one direction and when light is irradiated from another direction, for comparison. The image shown in FIG. 7A is an image captured when light is irradiated from one direction (the right side of the paper), and the image shown in FIG. 7B is an image captured when light is irradiated from the other direction (the left side of the paper). Each image includes an image of a crack 82 that has occurred on the surface of the test piece 80. The direction of the shadow caused by the crack 82 also differs between the image captured when light is irradiated from one direction and the image captured when light is irradiated from the other direction.

[0025] The control unit 18 (see FIG. 1) is a device configured by a computer that controls the stress applying unit 12, the illumination unit 14, and the imaging unit 16. The control unit 18 has a function of controlling the stress applying unit 12 to carry out a fatigue test, and a function of controlling the illumination unit 14, the drive unit 118 (see FIG. 5), and the imaging unit 16 so that the surface of the test piece 80 is imaged under illumination when the application of stress by the stress applying unit 12 stops for each predetermined cycle of the fatigue test.

[0026] Specifically, in the tension process of the fatigue test, the control unit 18 controls the drive unit 30 of the stress application unit 12 so that the first gripping unit 26 gradually moves away from the second gripping unit 28, and in the compression process of the fatigue test, the control unit 18 controls the drive unit 30 of the stress application unit 12 so that the first gripping unit 26 gradually moves closer to the second gripping unit 28. The control unit 18 stops the application of stress by the stress application unit 12 at each predetermined cycle of the fatigue test and switches to a mode for capturing images of the surface of the test piece 80.

[0027] When the control unit 18 transitions to a mode for capturing an image of the surface of the test piece 80, it first controls the drive unit 118 to move the pair of lighting devices 110 to a position where the pair of housings 112 directly face two of the four cameras 34, namely, cameras 34A and 34C, which face each other. Next, the control unit 18 controls each lighting device 110 so that light is irradiated onto the surface of the test piece 80 from one direction, and controls the two cameras 34A and 34C so that the surface of the test piece 80 is captured at the timing when the light from the one direction is irradiated. Next, the control unit 18 controls each lighting device 110 so that light is irradiated onto the surface of the test piece 80 from another direction, and controls the two cameras 34A and 34C so that the surface of the test piece 80 is captured at the timing when the light from the other direction is irradiated.

[0028] Next, the control unit 18 controls the drive unit 118 to move the pair of lighting devices 110 to a position where the pair of housings 112 directly faces two of the four cameras 34, namely, cameras 34B and 34D, which face each other. Next, the control unit 18 controls each lighting device 110 to irradiate the surface of the test specimen 80 with light from one direction, and controls the two cameras 34B and 34D to capture an image of the surface of the test specimen 80 at the timing when the light from the one direction is irradiated. Next, the control unit 18 controls each lighting device 110 to irradiate the surface of the test specimen 80 with light from another direction, and controls the two cameras 34B and 34D to capture an image of the surface of the test specimen 80 at the timing when the light from the other direction is irradiated. In this way, at each predetermined cycle of the fatigue test, when stress application by the stress application unit 12 stops, the surface of the test specimen 80 is imaged under illumination.

[0029] 8 shows an example of the configuration of the arithmetic processing unit 20. The arithmetic processing unit 20 is configured by a computer having a CPU (Central Processing Unit) 40, a memory 42, and a storage device 44. The CPU 40, the memory 42, and the storage device 44 are connected to each other so as to be able to communicate with each other via a bus 46 or the like. The computer that configures the arithmetic processing unit 20 and the computer that configures the above-mentioned control unit 18 may be the same computer or may be separate computers.

[0030] The CPU 40 is a central processing unit that reads programs from the storage device 44 and executes the programs using the memory 42 as a work area. The memory 42 is composed of RAM (Random Access Memory) or the like, and temporarily stores programs and data as a work area. The storage device 44 is composed of ROM (Read Only Memory), HDD (Hard Disk Drive), SSD (Solid State Drive), or the like, and stores various programs including the operating system and various data.

[0031] The storage device 44 stores a program for analyzing the test specimen 80. The CPU 40 reads the program for analyzing the test specimen 80 from the storage device 44, and loads and executes the read program in the memory 42, thereby functioning as each functional unit of the arithmetic processing unit 20. Specifically, the CPU 40 functions as a captured image storage control unit 50, a crack observation image extraction unit 52, a multi-resolution processing unit 54, an enhancement processing unit 56, and a crack analysis unit 58.

[0032] The captured image storage control unit 50 stores, in the storage device 44, one captured image obtained by each camera 34 capturing the surface of the test piece 80 when light is irradiated from one direction, and another captured image obtained by each camera 34 capturing the surface of the test piece 80 when light is irradiated from another direction. Each time one captured image and another captured image are generated by each camera 34, the captured image storage control unit 50 may generate a captured image (hereinafter referred to as a "composite captured image") by combining the one captured image and the other captured image. When generating a composite captured image, the captured image storage control unit 50 may then combine the composite captured images obtained by the multiple cameras 34 in a direction corresponding to the circumferential direction of the test piece 80. This results in a single captured image 100 capturing the entire circumference and entire length of the surface of the test piece 80. The captured image storage control unit 50 then stores each composite captured image or the combined composite captured image in the storage device 44. As a result, the captured image 100 obtained by capturing the entire circumference and entire length of the surface of the test piece 80 may be stored in the storage device 44 for each predetermined cycle of the fatigue test.

[0033] The crack observation image extraction unit 52 identifies a crack in the captured image 100 from among the captured images 100 generated by the imaging unit 16, and extracts the captured image 100 in which the identified crack appears as a crack observation image 102. That is, the crack observation image extraction unit 52 determines the presence or absence of a crack based on the difference in brightness that occurs in the captured image 100 according to the shadow of the crack 82, and extracts the captured image 100 that includes the crack as the crack observation image 102.

[0034] 9 shows an example of captured images 100 stored for each predetermined cycle of a fatigue test. In the captured image 100 acquired immediately before the fracture of the test specimen 80, the presence, location, and size of a crack 82 can be easily confirmed. However, in the captured image 100 acquired at the beginning of the fatigue test, it is difficult to confirm the presence, location, and size of a crack 82. However, identifying the crack 82 that occurs on the surface of the test specimen 80 for each predetermined cycle and evaluating the growth rate of the crack 82 are very important for accurately determining the life and strength of the product corresponding to the test specimen 80. Therefore, underestimating the size of the crack 82 or overlooking the crack 82 must be avoided.

[0035] Therefore, in the calculation processing unit 20, the multi-resolution processing unit 54, the emphasis processing unit 56, and the crack analysis unit 58 perform the respective processes described below so that the initial cracks 82 that appear on the surface of the test piece 80 during the fatigue test can be analyzed, from the cracks 82 that appear just before the test piece 80 breaks.

[0036] 10 shows an example of the processing flow from the multi-resolution processing unit 54 to the crack analysis unit 58. In the arithmetic processing unit 20, for each crack observation image 102 extracted by the crack observation image extraction unit 52, multi-resolution processing by the multi-resolution processing unit 54, emphasis processing by the emphasis processing unit 56, and crack analysis processing by the crack analysis unit 58 are performed in this order. At this time, in the arithmetic processing unit 20, for each crack observation image 102 extracted by the crack observation image extraction unit 52, the processing from the multi-resolution processing unit 54 to the crack analysis unit 58 is repeatedly performed by tracing back the crack observation images 102 one by one, starting from the crack observation image 102 acquired immediately before the fracture of the test piece 80 to the crack observation image 102 acquired at the beginning of the fatigue test. The processing flow will be specifically described below.

[0037] The crack analysis unit 58 focuses on the crack observation image 102 extracted by the crack observation image extraction unit 54, which is a crack 82 that can be considered to be a crack that occurred just before the test specimen 80 broke, and analyzes the crack that occurs on the surface of the test specimen by tracing back past crack observation images 102 that were taken at positions corresponding to the position of the surface of the test specimen 80 that was captured in the crack observation image 102 just before the break, from the crack observation image 102 just before the break to the initial crack observation image 102 that captured the starting point of the crack. Specifically, the crack analysis unit 58 first focuses on the crack observation image 102 immediately before fracture, among the crack observation images 102 extracted by the crack observation image extraction unit 54, which captures a crack 82 that can be considered to be the crack immediately before fracture of the test specimen 80. That is, among the captured images 100 captured immediately before the end of the fatigue test, the crack analysis unit 58 focuses on the crack observation image 102 that captures the crack 82 that can be considered to be the crack that caused the fracture of the test specimen 80. Next, the image capture unit 16 identifies a past captured image 102 captured after the start of the fatigue test at the same position where the focused crack observation image 102 immediately before fracture was captured. The crack analysis unit 58 traces back the multiple past crack observation images 102 acquired in chronological order, from the crack observation image 102 immediately before fracture to the initial crack observation image 102. By performing the above-described process, even minute initial cracks immediately after occurrence can be identified. Details of the process of tracing back the past crack observation images 102 will be described later.

[0038] FIG. 11 shows an example of multi-resolution processing by the multi-resolution processor 54 and enhancement processing by the enhancement processor 56. Performing multi-resolution processing by the multi-resolution processor 54 and enhancement processing by the enhancement processor 56 is preferable because it facilitates crack observation. As described above, the crack observation image 102 acquired immediately before the fracture of the specimen 80 is an image that makes it easy to confirm the presence, position, and size of the crack 82. The multi-resolution processor 54 then performs multi-resolution processing to convert the crack observation image 102 extracted by the crack observation image extractor 52 into captured images 100A at multiple different resolutions. For example, the number of resolution levels is set to 20. The number of resolution levels may be any number. The high-resolution captured image 100A is an image for observing minute cracks, and the low-resolution captured image 100A is an image for observing extremely large cracks. Generating the captured images 100A at multiple different resolutions makes it possible to observe everything from minute initial cracks immediately after their occurrence to extremely large cracks immediately before the fracture of the specimen 80.

[0039] The enhancement processing unit 56 performs enhancement processing on the captured images 100A at multiple resolutions generated by the multi-resolution processing unit to generate enhanced images. The enhancement processing includes first, second, and third processes. The first process includes Gabor filtering and binarization. The Gabor filtering process applies a first Gabor filter corresponding to the axial direction of the test object 80 and a second Gabor filter corresponding to a direction perpendicular to the axial direction of the test object 80 to the captured images 100A at each resolution to enhance the crack 82 in the captured images 100A. By applying the first and second Gabor filters corresponding to the horizontal and vertical directions of the captured images 100A, it is possible to enhance the boundary portion with a difference in brightness corresponding to the opening edge of the crack 82 in both the horizontal and vertical directions of the captured images 100A. The binarization process binarizes the captured images 100A at each resolution after the Gabor filtering process. As a result, a new captured image 100B is generated for each resolution, on which Gabor filtering and binarization have been performed. The Gabor filtering process is not limited to the horizontal and vertical directions, and emphasis processes in other directions can also be added.

[0040] The second process is an averaging process that adds together the new captured images 100B generated for each resolution in the first process. This generates an averaged captured image 100C. The third process is a process that further binarizes the captured image 100C averaged in the second process. This generates an enhanced image 100D in which the cracks 82 are enhanced. The enhanced image 100D in which the cracks 82 are enhanced is an example of the "enhanced image" of the present invention.

[0041] Fig. 12 shows an example of an enhanced image 100D generated immediately before fracture of the test piece 80. The enhanced image 100D shown in Fig. 12 is a captured image generated from a crack observation image 102 acquired immediately before fracture of the test piece 80. By performing multi-resolution processing and enhancement processing on the crack observation image 102 acquired immediately before fracture of the test piece 80, the difference in brightness according to the shadow of the crack 82 becomes clearer in the enhanced image 100D, and therefore the crack 82 can be imaged more clearly.

[0042] The crack analysis unit 58 chronologically traces past crack observation images 102, starting with the crack observation image 102 immediately before fracture as the first image, from the crack observation image 102 immediately before fracture to the initial crack observation image 102. FIG. 13 shows an example of the process of tracing past crack observation images 102 by the crack analysis unit 58. FIG. 13 shows an enhanced image 100D obtained by performing multi-resolution processing and enhancement processing on the ith (variable i is an integer equal to or greater than 1) and (i+1)th crack observation images 102, where the crack observation image 102 immediately before fracture is the first image. That is, the enhanced image 100D1 shown in FIG. 13 is a captured image generated from the ith crack observation image 102. The enhanced image 100D2 shown in FIG. 13 is a captured image generated from the (i+1)th crack observation image 102.

[0043] The crack analysis unit 58 sets, as a first crack extraction region 104 in the ith crack observation image 102, a rectangular region having approximately the same height and width as the height of a crack portion that has occurred due to an initial crack among the cracks captured in the ith crack observation image 102. That is, the crack analysis unit 58 sets the first crack extraction region 104 so as to surround the crack 82 in the ith enhancement-processed image 100D1 shown in FIG. Next, the crack analysis unit 58 sets, in the (i+1)th crack observation image 102, a region representing the same position as the first crack extraction region 104 in the i-th crack observation image 102 as the second crack extraction region 106, and for a portion of the crack included in the second crack extraction region 106 that has occurred due to an initial crack, sets a rectangular region having approximately the same height and width as the height of the crack portion as the first crack extraction region 104 in the (i+1)th crack observation image 102. That is, in the (i+1)th enhancement-processed image 100D2 shown in FIG. 11 , the crack analysis unit 60 sets, as the second crack extraction region 106, a region representing the same position as the first crack extraction region 104 in the i-th enhancement-processed image 100D1, and identifies, among the cracks included in the second crack extraction region 106, a portion of the crack 82 that has occurred due to an initial crack. Then, a rectangular area having approximately the same height as the height of the identified crack 82 portion and approximately the same width as the width of the crack portion is set as the first crack extraction area 104 in the (i+1)th enhanced processed image 100D2. In this way, in the (i+1)th enhanced processed image 100D2, by identifying cracks that have arisen due to an initial crack within the second crack extraction region 106 and setting the first crack extraction region 104, it is possible to avoid erroneously extracting noise (such as cracks that have not arisen due to the initial crack). Here, the crack that has occurred due to the initial crack can be the largest crack in the second crack extraction region 106. By detecting the length, depth, etc. of the crack, it is possible to identify the crack that has occurred due to the initial crack. The crack analysis unit 58 repeats the above process to chronologically trace past crack observation images 102, from the crack observation image 102 immediately before fracture to the initial crack observation image 102. If the second crack extraction region 106 can be set in the (i+1)th crack observation image 102 but the first crack extraction region 104 cannot be set, the (i+1)th crack observation image 102 is determined to be an initial crack observation image, and a starting point for crack generation is present within the second crack extraction region 106 of the initial crack observation image. Specifically, FIG. 14 shows an example of how the first crack extraction region 104 is sequentially set when tracing back past crack observation images in chronological order. As shown in FIG. 14, in the (n+1)th enhanced image 100D(n+1), the region representing the same position as the first crack extraction region 104 in the (n)th enhanced image 100D(n) can be set as the second crack extraction region 106. However, a minute initial crack cannot be detected, and the first crack extraction region 104 cannot be set. Therefore, the (n+1)th enhanced image 100D(n+1) is determined to be an initial crack observation image, and it is determined that the starting point of a crack is present within the second crack extraction region 106 of the initial crack observation image. Through the above process, the crack analysis unit 58 can analyze the progression (progression amount and propagation speed) of the crack that caused the fracture from the starting point of the crack.

[0044] The crack analysis unit 58 sets a first crack extraction region 104 for the ith enhanced image 100D1 generated using the multi-resolution processing unit 54 and the enhancement processing unit 56. That is, the multi-resolution processing by the multi-resolution processing unit 54 and the enhancement processing by the enhancement processing unit 56 described above are performed when tracing back past crack observation images 102 in chronological order, starting with the crack observation image 102 immediately before fracture as the first image, from the crack observation image 102 immediately before fracture to the initial crack observation image 102.

[0045] FIG. 15 shows an example of an enhanced image 100D generated at the beginning of a fatigue test. The enhanced image 100D shown in FIG. 15 is a captured image generated from a crack observation image 102 acquired at the beginning of a fatigue test. It is difficult to confirm the presence, position, and size of a crack 82 in the crack observation image 102 acquired at the beginning of a fatigue test. However, in this embodiment, multi-resolution processing and enhancement processing are performed on the crack observation image 102 acquired at the beginning of a fatigue test to generate the captured image 100D. As a result, the difference in brightness according to the shading of the crack 82 becomes clear in the enhanced image 100D, making it possible to confirm the presence, position, and size of the crack 82.

[0046] Furthermore, because an initial crack is small, it is difficult to detect the initial crack from the beginning based on a crack observation image 102 acquired at the beginning of a high-temperature fatigue test. However, in this embodiment, first, attention is focused on a crack observation image 102 for observing a crack that can be considered to be a crack immediately before the test piece 80 breaks, and then past crack observation images 102 captured at a position corresponding to the position on the surface of the test piece 80 captured in the focused crack observation image 102 are traced back to the initial crack observation image 102 in which the initial crack that is the starting point for the crack 82 to occur was captured. This makes it possible to predict the position of the initial crack, thereby making it possible to detect the initial crack.

[0047] An example of the flow of the test piece analyzing method according to this embodiment is shown in Fig. 16. The test piece analyzing method according to this embodiment is executed using a test piece analyzing apparatus 10.

[0048] When the test piece analysis method is started, first, in step S10, the captured image storage control unit 50 stores, in the storage device 44, one captured image of the surface of the test piece 80 captured by each camera 34 when light is irradiated from one direction and another captured image of the surface of the test piece 80 captured when light is irradiated from another direction. Each time one captured image and another captured image are generated by each camera 34, the captured image storage control unit 50 may generate a composite captured image by combining the one captured image and the other captured image. When generating a composite captured image, the captured image storage control unit 50 may then combine the composite captured images obtained by the multiple cameras 34 in a direction corresponding to the circumferential direction of the test piece 80. This results in a single captured image 100 capturing the entire circumference and entire length of the surface of the test piece 80. The captured image storage control unit 50 may then store each composite captured image or the combined composite captured image in the storage device 44. As a result, captured images 100 of the entire circumference and entire length of the surface of the test piece 80 are stored in the storage device 44 for each predetermined cycle of the fatigue test. Step S10 is an example of the "captured image generating step" in the present invention.

[0049] In step S12, the crack observation image extraction unit 52 identifies a crack in the captured image 100, and extracts the captured image 100 that shows the identified crack as the crack observation image 102. That is, the crack observation image extraction unit 52 determines the presence or absence of a crack based on the difference in brightness that occurs in the captured image 100 according to the shadow of the crack 82, and extracts the captured image 100 that includes the crack as the crack observation image 102. Step S12 is an example of the "crack observation image extraction step" in the present invention.

[0050] In step S14, the crack analysis unit 58 focuses on the crack observation image 102 immediately before fracture that captures a crack 82 that can be considered to be a crack immediately before fracture of the test piece 80, among the crack observation images 102 extracted by the crack observation image extraction unit 54. That is, the crack analysis unit 58 performs processing using the crack observation image immediately before fracture as the first image.

[0051] In step S16, the multi-resolution processing unit 54 performs multi-resolution processing and the enhancement processing unit 56 performs enhancement processing on the crack observation image 102 immediately before fracture, thereby generating an enhanced image 100D. Specifically, the multi-resolution processing unit 54 converts the crack observation image 102 immediately before the fracture into captured images 100A at a plurality of different resolutions. Next, the enhancement processing unit 56 performs Gabor filtering on the captured image 100A at each resolution, and performs binarization on the captured image 100A at each resolution that has been subjected to the Gabor filtering process, thereby generating a new captured image 100B for each resolution that has been subjected to the Gabor filtering process and binarization process. Furthermore, the enhancement processing unit 56 performs averaging processing to add together the new captured images 100B for each resolution that have been binarized, thereby generating an averaged captured image 100C. Finally, the enhancement processing unit 56 performs binarization processing to further binarize the averaged captured image 100C, thereby generating an enhancement-processed image 100D in which the cracks 82 are enhanced.

[0052] In step S18, the crack analysis unit 58 sets a rectangular area in the enhanced processed image 100D, which is generated in step S16 by processing the crack observation image 102 just before fracture, having a height approximately equal to the height of the crack just before fracture and a width approximately equal to the width of the crack just before fracture as the first crack extraction area 104 in the crack observation image 102 just before fracture.

[0053] In step S20, the crack analysis unit 58 performs multi-resolution processing by the multi-resolution processing unit 54 and enhancement processing by the enhancement processing unit 56 on the next (i+1th) crack observation image 102 when tracing back past crack observation images 102 in chronological order, from the crack observation image 102 just before fracture to the initial crack observation image 102, to generate an enhanced processed image 100D.

[0054] In step S22, the crack analysis unit 58 sets, in the (i+1)th enhanced image 100D, an area that represents the same position as the first crack extraction area 104 in the i-th crack observation image 102 as the second crack extraction area 106. That is, when the crack observation image 102 immediately before fracture is set as the i-th crack observation image (i=1), in the (i+1)th enhanced image 100D, an area that represents the same position as the first crack extraction area 104 in the crack observation image 102 immediately before fracture set in step S18 is set as the second crack extraction area 106.

[0055] In step S24, the crack analysis unit 58 sets a rectangular area having approximately the same height and width as the height of the crack portion that was caused by an initial crack among the cracks contained in the second crack extraction area 106 in the (i+1)th enhanced processed image 100D as the first crack extraction area 104 in the (i+1)th crack observation image 102.

[0056] In step S26, the crack analysis unit 58 determines whether the first crack extraction region 104 was set in the (i+1)th enhanced image 100D. If the second crack extraction region 106 was set but the first crack extraction region 104 was not set in the (i+1)th enhanced image 100D, the process proceeds to step S28. If the second crack extraction region 106 was set and the first crack extraction region 104 was set in the (i+1)th enhanced image 100D, the process proceeds to step S32. Specifically, i is updated. That is, past crack observation images 102 are traced back in time series, and the process proceeds to step S20 to generate an enhanced image 100D for the next crack observation image 102.

[0057] If it is determined in step S26 that the second crack extraction region 106 can be set in the (i+1)th enhanced processed image 100D but the first crack extraction region 104 cannot be set, then in step S28 the crack analysis unit 60 determines that the (i+1)th crack observation image 102 is an initial crack observation image, and determines that the starting point of a crack is present within the second crack extraction region 106 of the initial crack observation image.

[0058] In step S30, the crack analysis unit 60 analyzes the progress (amount of progress and progress rate) of the crack from the crack initiation point determined in step S28. Then, the test specimen analysis method ends. Steps S26, S30, and S32 are an example of a "crack analysis step" in the present invention.

[0059] As described above in detail, in this embodiment, first, attention is paid to the crack observation image 102 for observing a crack that can be considered to be a crack immediately before the test piece 80 breaks, and then past crack observation images 102 that captured a position corresponding to the position of the surface of the test piece 80 captured in the focused crack observation image 102 are traced back to the initial crack observation image 102 that captured an initial crack that is the starting point for the crack 82 to occur. Therefore, the position of the initial crack can be predicted, and therefore the initial crack can be detected.

[0060] In this embodiment, multi-resolution processing and enhancement processing are performed on the crack observation image 102 acquired during the fatigue test to generate an enhanced image 100D. Therefore, in the enhanced image 100D, the difference in brightness that occurs depending on the shading of the crack 82 becomes clear, so that the presence, position, and size of the crack 82 can be confirmed from the captured image 100 acquired at the beginning of the fatigue test to the captured image 100 acquired immediately before the fracture of the test piece 80.

[0061] In addition, in this embodiment, when tracing back past crack observation images 102 in chronological order from the crack observation image 102 immediately before fracture to the initial crack observation image 102, with the crack observation image 102 immediately before fracture as the first image, in the (i+1)th crack observation image 102, a region representing the same position as the first crack extraction region 104 in the i-th crack observation image 102 is set as a second crack extraction region 106, and among the cracks included in the second crack extraction region 106, a crack portion caused by the initial crack is extracted. A rectangular area having approximately the same height as the height of the crack portion and approximately the same width as the width of the crack portion is set as the first crack extraction area 104 in the (i+1)th crack observation image 102, and if the second crack extraction area 106 can be set in the (i+1)th crack observation image 102 but the first crack extraction area 104 cannot be set, the (i+1)th crack observation image 102 is determined to be an initial crack observation image, and it is determined that the starting point of a crack is present in the second crack extraction area 106 of the initial crack observation image. Therefore, it is possible to prevent noise from being included around the crack 82 in the crack observation image 102, and it is possible to prevent the maximum length of the crack 82 in the crack observation image 104 from being erroneously detected as being too long.

[0062] Furthermore, in this embodiment, the crack 82 is analyzed based on the captured image 100. Therefore, the time required to analyze the crack 82 can be shortened compared to the replica method.

[0063] Furthermore, in this embodiment, the illumination unit 14 has fewer illumination devices 110 than the multiple cameras 34 that make up the imaging unit 16, and is configured to be movable around the test object 80. Therefore, compared to a configuration in which the illumination unit 14 has the same number of illumination devices 110 as the multiple cameras 34 that make up the imaging unit 16 and is fixed to the test object 80, it is possible to save space, and therefore it is possible to image the surface of the test object 80 under illumination even when the area around the test object 80 is small.

[0064] FIG. 17 shows an example of the results of detecting the length of a main crack using the test specimen analysis method according to this embodiment. In the graph of FIG. 17, the horizontal axis represents the test time [sec], which is the time elapsed since the start of the fatigue test, and the vertical axis represents the length of the main crack. As shown in FIG. 17, the test specimen analysis method according to this embodiment makes it possible to detect the length of a main crack immediately after its occurrence. This makes it possible to evaluate the amount and rate of growth of the main crack, which change with each cycle, and therefore makes it possible to accurately determine the lifespan, strength, etc. of the product corresponding to test specimen 80.

[0065] The above describes one embodiment of the present invention, but the present invention is not limited to the above, and it goes without saying that the present invention can be implemented in various modified forms within the scope of the gist of the present invention. [Explanation of symbols]

[0066] 10. Test specimen analysis equipment 12 Stress applying section 14 Lighting Department 16 Imaging unit 18 Control Unit 20 Processing unit 26 1st grip part 28 Second grip part 30 Drive unit 32 Test bench 34 Camera 40 CPU 42 memory 44 Storage device 46 Bus 50 Captured image storage control unit 52 Crack observation image extraction unit 54 Multi-resolution processing unit 56 Emphasis processing unit 58 Crack Analysis Department 80 test specimens 82 Crack 100 captured images 102 Crack observation image 104 First crack extraction area 106 Second crack extraction area 110 Lighting equipment 112 Case 114 Cushioning material 116 Window 118 Drive Unit 120 Optical system 122 Light source 124 Prism

Claims

1. 1. A test specimen analysis device that analyzes a test specimen based on cracks that occur on the surface of the test specimen during a fatigue test in which the temperature of the test specimen is maintained within a set temperature range of less than 700°C, a stress applying unit that repeatedly applies tensile stress and compressive stress in a predetermined cycle in the axial direction of the test specimen as a load in the fatigue test; an illumination unit that switches between irradiating the surface of the test piece with light from one direction and light from another direction across a plane perpendicular to the axis of the test piece; an imaging unit that captures an image of the surface of the test object from a normal direction of the surface of the test object using one or more area cameras provided opposite the surface of the test object, and generates a captured image; a calculation processing unit that analyzes cracks occurring on the surface of the test specimen using the captured image generated by the imaging unit and analyzes the test specimen based on the analysis results; and the imaging unit generates the captured image by positioning the illumination unit between the imaging unit and the test piece during the fatigue test when stress application by the stress application unit is stopped, and capturing images of the surface of the test piece at each of a timing when the light from the one direction is irradiated onto the surface of the test piece and a timing when the light from the other direction is irradiated onto the surface of the test piece; The arithmetic processing unit a crack observation image extraction unit that identifies a crack in the captured image and extracts an area including the identified crack in the captured image as a crack observation image; a crack analysis unit that focuses on a crack observation image immediately before fracture, which captures a crack that can be considered to be a crack that occurs immediately before fracture of the test piece, among the crack observation images extracted by the crack observation image extraction unit, and analyzes cracks that occur on the surface of the test piece by tracing back past crack observation images that were taken at positions corresponding to the positions on the surface of the test piece captured in the crack observation image immediately before fracture from the crack observation image immediately before fracture to an initial crack observation image that captures an initial crack that is the starting point for crack generation; A test specimen analysis device having the above.

2. The crack analysis unit When tracing back the past crack observation images in chronological order from the crack observation image immediately before the fracture to the initial crack observation image, with the crack observation image immediately before the fracture being the first image, In an i-th crack observation image (variable i is an integer of 1 or more), for a portion of a crack that has occurred due to the initial crack among the cracks captured in the i-th crack observation image, a rectangular area having approximately the same height as the height of the portion of the crack and approximately the same width as the width of the portion of the crack is set as a first crack extraction area in the i-th crack observation image; In the (i+1)th crack observation image, a region representing the same position as the first crack extraction region in the i-th crack observation image is set as a second crack extraction region, and for a portion of a crack included in the second crack extraction region that has occurred due to the initial crack, a rectangular region having approximately the same height as the height of the portion of the crack and approximately the same width as the width of the portion of the crack is set as the first crack extraction region in the (i+1)th crack observation image; In the (i+1)th crack observation image, when the second crack extraction region can be set and the first crack extraction region cannot be set, A test specimen analysis device as described in claim 1, which determines that the (i+1)th crack observation image is the initial crack observation image and determines that the starting point of a crack exists within the second crack extraction area of ​​the initial crack observation image.

3. The arithmetic processing unit a multi-resolution processing unit that converts the captured image generated by the imaging unit into captured images with a plurality of different resolutions; an enhancement processing unit that applies a first Gabor filter corresponding to a first direction and a second Gabor filter corresponding to a second direction different from the first direction to the captured images of the plurality of resolutions generated by the multi-resolution processing unit to generate new captured images, and adds and averages the new captured images to generate an enhancement-processed image; and When tracing back the past crack observation images in chronological order from the crack observation image immediately before the fracture to the initial crack observation image, with the crack observation image immediately before the fracture being the first image, A test specimen analysis device as described in claim 2, wherein the i+1th enhanced processed image is generated for the i+1th crack observation image using the multi-resolution processing unit and the enhancement processing unit, and a first crack extraction region is set in the i+1th enhanced processed image.

4. 1. A test specimen analysis method for analyzing a test specimen based on cracks that occur on a surface of the test specimen during a fatigue test in which the temperature of the test specimen is maintained within a set temperature range of less than 700°C, comprising: a stress applying unit that repeatedly applies tensile stress and compressive stress in a predetermined cycle in the axial direction of the test specimen as a load in the fatigue test; an illumination unit that switches between irradiating the surface of the test piece with light from one direction and light from another direction across a plane perpendicular to the axis of the test piece; an imaging unit that captures an image of the surface of the test object from a normal direction of the surface of the test object using one or more area cameras provided opposite the surface of the test object, and generates a captured image; a calculation processing unit that analyzes cracks occurring on the surface of the test specimen using the captured image generated by the imaging unit and analyzes the test specimen based on the analysis results; Using a test specimen analysis device having an image generating step of generating the captured image by using the imaging unit during the fatigue test, when the stress application by the stress application unit is stopped, by placing the illumination unit between the imaging unit and the test piece, and capturing images of the surface of the test piece at each of the timings when the light from the one direction is irradiated onto the surface of the test piece and the timing when the light from the other direction is irradiated onto the surface of the test piece; a crack observation image extraction step of identifying a crack in the captured image and extracting an area including the identified crack in the captured image as a crack observation image; a crack analysis step of analyzing cracks occurring on the surface of the test piece by focusing on a crack observation image immediately before fracture that captures a crack that can be considered to be a crack that occurs immediately before fracture of the test piece among the crack observation images extracted in the crack observation image extraction step, and tracing back past crack observation images captured at positions corresponding to the positions on the surface of the test piece captured in the crack observation image immediately before fracture from the crack observation image immediately before fracture to an initial crack observation image that captures an initial crack that is the starting point for crack generation; A test specimen analysis method comprising:

Images