Apparatus And Method For Analyzing Crack Path Of Spent Nuclear Fuel Cladding

Abstract

The present invention relates to an apparatus and method for analyzing the crack path of a spent nuclear fuel cladding tube. It comprises a camera for photographing the spent nuclear fuel cladding tube and a processor that preprocesses image data input from the camera, predicts the location where a crack will occur by applying strain energy to the preprocessed image data, and predicts the crack path starting from the crack location. This allows for more accurate prediction of the crack location and crack path of the spent nuclear fuel cladding tube and prevents accidents caused by cracks.

Description

Technology Field

[0001] The present invention relates to an apparatus and method for analyzing crack paths in spent nuclear fuel cladding, which predict cracks occurring in the cladding of spent nuclear fuel. Background Technology

[0002] In nuclear power plant reactors, nuclear fuel rods, with uranium or plutonium as their main components, are used as fuel for nuclear fission reactions. After nuclear power generation is completed, these fuel rods are classified as spent nuclear fuel and are subject to separate management. As nuclear power generation continues, the amount of spent nuclear fuel is continuously increasing.

[0003] Spent nuclear fuel (spent fuel rods) contains some uranium and plutoium that can undergo nuclear fission, and in addition to uranium, highly toxic radioactive materials such as xenon, cesium, and strontium are generated, so management is required. Accordingly, spent nuclear fuel is managed by placing it in deep water to shield it from radiation.

[0004] However, if spent nuclear fuel is dropped or struck during transportation, pinch loads can occur. Pinch loads are problematic because they can induce cracks in the spent fuel cladding. Additionally, cracks can develop in the spent fuel due to hydrides. If cracks occur in the cladding in this manner, it poses a problem as internal fuel or fission products may leak out.

[0005] A related technology is Korean Registered Patent No. 10-2225562, "Inspection device for nuclear fuel rod cladding using ultrasound." The problem to be solved

[0006] The present invention was created out of the aforementioned necessity and aims to provide an apparatus and method for analyzing the crack path of a spent nuclear fuel cladding, which safely manages spent nuclear fuel by predicting a crack path similar to the actual crack path. means of solving the problem

[0007] To achieve the above-mentioned objective, a crack path analysis device for a spent nuclear fuel cladding tube according to one aspect of the present invention comprises: a camera for photographing a spent nuclear fuel cladding tube; and a processor for preprocessing image data input from the camera, predicting a crack location where a crack will occur by applying strain energy to the preprocessed image data, and predicting a crack path starting from the crack location.

[0008] The above processor is characterized by determining the location where the magnitude of the strain energy is greatest as the location where a crack may occur and predicting the crack location.

[0009] The above processor is characterized by distinguishing hydrides generated by radiation from the image data and applying the strain energy to the location of the hydrides.

[0010] The processor is characterized by generating nodes and edges in the image data based on pixels, and calculating the crack path by assigning weights according to the strain energy to the edges.

[0011] The processor is characterized by calculating the path with the smallest sum of weights among a plurality of paths starting from the crack location as the crack path.

[0012] The above processor is characterized by predicting the crack path by applying the Dijkstra algorithm.

[0013] The processor is characterized by calculating similarity by comparing crack data regarding actual cracks occurring in the cladding tube with the crack location and the crack path, and verifying the crack location and the crack path according to the similarity.

[0014] The processor is characterized by performing a ring compression test (RCT) on a specimen simulating the environment of the cladding tube, acquiring data regarding cracks occurring in the specimen as crack data, and comparing it with the crack location and crack path.

[0015] The above processor is characterized by distinguishing between the hydride and the cladding from the image data and removing noise.

[0016] The processor is characterized by identifying the shape of the hydride through morphological operations, distinguishing the hydride and the cladding tube through the Oates algorithm, and determining small hydrides as noise and removing them from the image data.

[0017] The processor is characterized by generating a pixel-based finite element model from the image data, defining material properties and boundary conditions for the finite element model, and analyzing the strain energy distribution by distinguishing the cladding tube and the hydride from the image data.

[0018] A method for analyzing a crack path of a spent nuclear fuel cladding tube according to one aspect of the present invention comprises: a step in which a processor preprocesses image data of a spent nuclear fuel cladding tube; a step in which the processor predicts a crack location where a crack will occur by applying strain energy to the preprocessed image data; and a step in which the processor predicts a crack path starting from the crack location.

[0019] In the step of predicting the crack location, the processor predicts the crack location by determining the location where the magnitude of the strain energy is greatest as the location where a crack may occur.

[0020] In the step of predicting the crack location, the processor distinguishes hydrides generated by radiation from the image data and applies the strain energy to the location of the hydrides.

[0021] The step of predicting the crack location comprises: generating a pixel-based finite element model from the image data; defining material properties and boundary conditions for the finite element model; and analyzing the strain energy distribution by distinguishing the cladding and hydride from the image data.

[0022] The step of predicting the crack path is characterized by comprising: a step of generating nodes and edges in the image data based on pixels; a step of assigning weights according to the strain energy to the edges; and a step of calculating the crack path connecting the starting node to the last node.

[0023] In the step of predicting the crack path, the processor is characterized by calculating the path with the smallest sum of weights among a plurality of paths starting from the crack location as the crack path.

[0024] In the step of predicting the crack path, the processor is characterized by predicting the crack path by applying the Dijkstra algorithm.

[0025] The step of preprocessing the image data comprises: a step of confirming the shape of the hydride from the image data; a step of distinguishing the hydride and the cladding tube from the image data; and a step of determining the hydride with a small size as noise and removing it from the image data.

[0026] After the step of predicting the crack path, the processor further comprises the step of comparing crack data for an actual crack occurring in the cladding tube with the crack location and the crack path; and the step of calculating similarity based on the comparison result and verifying the crack location and the crack path in correspondence with the similarity. Effects of the invention

[0028] According to one aspect, the apparatus and method for analyzing crack paths in spent nuclear fuel cladding according to the present invention can stably manage spent nuclear fuel by predicting crack paths in the cladding of spent nuclear fuel.

[0029] An apparatus and method for analyzing crack paths in spent nuclear fuel cladding tubes according to one aspect of the present invention can improve the accuracy of prediction regarding crack paths by using the Dijkstra algorithm based on strain.

[0030] An apparatus and method for analyzing crack paths in spent nuclear fuel cladding according to one aspect of the present invention can prevent accidents caused by spent nuclear fuel and maintain safety. Brief explanation of the drawing

[0031] FIG. 1 is a block diagram showing spent nuclear fuel and an analysis device according to one embodiment of the present invention. FIG. 2 is a block diagram illustrating the configuration of an analysis device according to one embodiment of the present invention. FIG. 3 is a figure referenced to explain a crack formed in a spent nuclear fuel cladding tube according to one embodiment of the present invention. FIG. 4 is a figure referenced to explain strain energy according to the prediction of a crack path of a spent nuclear fuel cladding tube according to one embodiment of the present invention. FIG. 5 is a figure referenced to explain the Dijkstra algorithm for predicting crack paths in spent nuclear fuel cladding according to one embodiment of the present invention. FIG. 6 is an exemplary diagram showing a predicted crack path of a spent nuclear fuel cladding tube according to one embodiment of the present invention. FIG. 7 is a flowchart illustrating an analysis method for predicting a crack path of a spent nuclear fuel cladding tube according to one embodiment of the present invention. FIG. 8 is a figure referenced to explain a method for processing an image of a spent nuclear fuel cladding tube according to one embodiment of the present invention. FIG. 9 is a figure referenced to explain a finite element model based on an image of a spent nuclear fuel cladding tube according to one embodiment of the present invention. FIG. 10 is a figure referenced to explain a method for predicting a crack path using an algorithm based on strain energy according to one embodiment of the present invention. FIG. 11 is a figure illustrating an example of verification of a predicted crack path according to one embodiment of the present invention. FIG. 12 is an example diagram comparing a predicted crack path and an actual crack path according to one embodiment of the present invention. FIG. 13 is an example diagram comparing a predicted crack path according to one embodiment of the present invention with a crack path according to a different analysis method. Specific details for implementing the invention

[0032] Hereinafter, an embodiment of the present invention will be described with reference to the attached drawings.

[0033] In this process, the thickness of lines or the size of components depicted in the drawings may be exaggerated for the sake of clarity and convenience of explanation. Furthermore, the terms described below are defined considering their functions in the present invention, and these may vary depending on the intent or convention of the user or operator. Therefore, the definitions of these terms should be based on the content throughout this specification.

[0034] FIG. 1 is a block diagram showing spent nuclear fuel and an analysis device according to one embodiment of the present invention.

[0035] Referring to FIG. 1, a crack path analysis device (hereinafter, analysis device) (100) according to one embodiment of the present invention can predict cracks by analyzing image data of spent nuclear fuel (10).

[0036] The analysis device (100) can photograph the cladding of the spent nuclear fuel (10) using a camera (150). The analysis device (100) can preprocess image data of the spent nuclear fuel (10) obtained through the camera (150) and analyze the image data to predict cracks in the cladding of the spent nuclear fuel (10). Additionally, the analysis device (100) can predict crack paths.

[0037] The analysis device (100) can detect locations with low resistance to external forces to predict a crack path similar to an actual crack. For example, the analysis device (100) can predict a crack path using the Dijkstra algorithm based on strain energy. The analysis device (100) can also simulate cracks in a cladding tube by generating actual cracks in a specimen that simulates the environment of the cladding tube through a ring compression test (RCT). The analysis device (100) can acquire crack data regarding the actual cracks generated in the specimen.

[0038] The analysis device (100) can verify the predicted crack path based on the similarity between the predicted crack path and the actual crack path when the crack path is predicted. Additionally, the analysis device (100) can compare the predicted crack path with other crack paths calculated through other analysis methods. Based on the comparison results, the analysis device (100) can verify the accuracy of the predicted crack location and crack path.

[0039] FIG. 2 is a block diagram illustrating the configuration of an analysis device according to one embodiment of the present invention.

[0040] Referring to FIG. 2, the analysis device (100) may include a memory (120), a communication unit (130), a sensor (140), a camera (150), and a processor (110).

[0041] The sensor (140) includes a plurality of sensors and applies the detected data to the processor (110). The sensor (140) can detect the distance to the spent nuclear fuel (10) and detect radiation leakage. Additionally, the sensor (140) can detect the surface condition or cracks of the cladding tube based on infrared or ultrasonic waves. For example, the sensor (140) may include at least one of a distance sensor, an ultrasonic sensor, an infrared sensor, and a radiation detection sensor.

[0042] The camera (150) can photograph the spent nuclear fuel (10). A separate dedicated camera for photographing the spent nuclear fuel (10) may be used as the camera (150). Additionally, the camera (150) may be a camera installed at a facility where the spent nuclear fuel (10) is stored. In some cases, the camera (150) may be an infrared camera or a thermal imaging camera.

[0043] The camera (150) can transmit the generated image data to the processor (110) after photographing the spent nuclear fuel (10). The camera (150) can photograph the cross-section of the cladding tube.

[0044] The communication unit (130) can receive data regarding spent nuclear fuel (10) in response to a control command from the processor (110). Additionally, the communication unit (130) can receive at least one of data, a learning model, and algorithm data for analyzing the cladding of the spent nuclear fuel (10).

[0045] The communication unit (130) can transmit data regarding the predicted crack location and crack path based on the analysis results to an external server or terminal. For example, the communication unit (130) can communicate using a wired or wireless communication module. For example, the communication unit (130) can communicate using a communication method such as serial communication, Ethernet, Wi-Fi, Bluetooth, or power line communication.

[0046] The memory (120) can store data received from the sensor (140) or the communication unit (130). The memory (120) can store image data captured through the camera (150). The memory (120) can store at least one of analysis result data, crack detection result data, crack prediction result data, and crack path data.

[0047] Additionally, the memory (120) can store control data and algorithm data for analyzing the cladding tube. For example, the memory (120) can store data regarding at least one of an image processing algorithm, a strain energy analysis algorithm, Dijkstra algorithm data, a crack prediction algorithm, and a crack path prediction algorithm.

[0048] The memory (120) may include non-volatile memory such as RAM (Random Access Memory), ROM (ROM), EEPROM (Electrically Erased Programmable ROM), flash memory, HDD, SSD, etc.

[0049] The processor (110) may include at least one microprocessor and operate based on data stored in memory (120).

[0050] The processor (110) can determine and analyze the condition of the spent nuclear fuel (10) based on data input from the sensor (140). The processor (110) may stop the analysis if a crack occurs in the cladding of the spent nuclear fuel (10) or if radioactive material leaks out, based on the data input from the sensor (140).

[0051] The processor (110) can preprocess image data of the cross-section of the cladding tube input from the camera (150) and analyze the condition of the cladding tube based on the image data. The processor (110) can detect locations with low resistance to external forces to predict crack occurrence points (crack locations) or crack paths. The processor (110) can detect locations with low resistance to external forces based on strain energy obtained through finite element analysis.

[0052] The processor (110) can calculate the crack path based on strain energy, but based on the Dijkstra algorithm used for path setting. The processor (110) can calculate the crack path by predicting the direction in which the crack occurs and the crack propagates.

[0053] The processor (110) can output the calculated crack path as a predicted crack path. The processor (110) can generate and output analysis result data including the predicted crack location and crack path. The processor (110) can transmit the analysis result data to an external server or terminal through the communication unit (130).

[0054] The processor (110) can verify the predicted crack path. The processor (110) can analyze similarities by comparing the actual crack path with the predicted crack path and verify the accuracy of the predicted crack location and crack path based on the similarities. Additionally, the processor (110) can verify the accuracy of the predicted crack path by comparing it with other crack paths calculated according to different analysis methods.

[0055] FIG. 3 is a figure referenced to explain a crack formed in a spent nuclear fuel cladding tube according to one embodiment of the present invention.

[0056] As illustrated in FIG. 3, cracks may occur in the cladding of the spent nuclear fuel (10). An analysis device (100) analyzes image data of the cladding of the spent nuclear fuel (10) (S10) and can predict the crack path using various methods.

[0057] The analysis device (100) can predict a crack path by using the Dijkstra algorithm based on strain energy (S20). The analysis device (100) can predict a crack path for a location by detecting a part with low resistance to external force before a crack occurs (S30). The analysis device (100) can output the predicted crack path as illustrated.

[0058] Additionally, the analysis device (100) can perform a ring compression test (RCT) (S40) to induce actual cracks in a specimen simulating the environment of a cladding tube (S50). The analysis device (100) can store image data of the actual cracks and calculate the location where the cracks occurred and the crack paths to store them as crack data. The analysis device (100) can verify the predicted values ​​by comparing the crack data from the ring compression test with the predicted crack locations and crack paths.

[0059] The analysis device (100) can analyze similarities by comparing the predicted crack path (predicted crack path) with the actual crack according to the ring compression test (RCT), and verify the accuracy of the predicted crack location and crack path based on the similarities.

[0060] The analysis device (100) can improve prediction accuracy by changing the settings for strain energy and Dijkstra's algorithm based on the calculation of the predicted crack path according to the similarity analysis results.

[0061] Accordingly, the analysis device (100) can predict a crack or crack path before an actual crack occurs in the cladding of the spent nuclear fuel (10), and based on this, the spent nuclear fuel (10) can be managed before a crack occurs to prevent an accident.

[0062] FIG. 4 is a figure referenced to explain strain energy according to the prediction of a crack path of a spent nuclear fuel cladding tube according to one embodiment of the present invention.

[0063] Referring to FIG. 4 (a), the analysis device (100) detects a location with low resistance to external force based on strain energy.

[0064] The analysis device (100) can analyze the energy accumulated in the material when a load is applied by inputting strain energy based on finite element analysis, which analyzes the behavior of the material according to an external load using numerical techniques. Strain is a resistance force against external force and indicates the degree to which deformation occurs due to an external force.

[0065] The analysis device (100) can determine that a location with high strain energy has low resistance to external load. Accordingly, the analysis device (100) can predict that a crack will occur at a location with low resistance to external load, that is, at the location with the highest strain energy.

[0066] Referring to FIG. 4(b), the analysis device (100) can predict the location where a crack will occur by analyzing the strain energy through finite element analysis when a force is applied to the cladding tube. By analyzing the flow of strain energy, the analysis device (100) can predict the location of the crack by detecting the location where stress is concentrated, that is, the location where the strain energy is high.

[0067] The analysis device (100) detects the location where a crack occurs through strain energy to predict a crack path similar to an actual crack.

[0068] FIG. 5 is a figure referenced to explain the Dijkstra algorithm for predicting crack paths in spent nuclear fuel cladding according to one embodiment of the present invention.

[0069] Referring to FIG. 5, the analysis device (100) can predict the location where a crack will occur (crack location) based on strain energy, and predict the path where the crack will proceed, i.e., the crack path, using the Dijkstra algorithm based on the location.

[0070] Dijkstra's algorithm is an algorithm used to find the shortest path, such as in satellite GPS and path planning. Dijkstra's algorithm searches for the shortest path based on the weights of the edges connecting a starting node to a final node.

[0071] The analysis device (100) can calculate weights for multiple paths connecting nodes and calculate the path with the smallest sum of weights as the shortest path. For example, the Dijkstra algorithm can calculate 9 by summing the weights (2, 1, 7) for multiple paths from node 1 to node 9, that is, the first path (1-2-5-9), calculate 6 by summing the weights (2, 1, 1, 1, 1) for the second path (1-2-3-8-9), and calculate 7 by summing the weights (5, 1, 1) for the third path (1-3-8-9). The Dijkstra algorithm can calculate the second path with the smallest weight as the shortest path.

[0072] FIG. 6 is an exemplary diagram showing a predicted crack path of a spent nuclear fuel cladding tube according to one embodiment of the present invention.

[0073] Referring to FIG. 6 (a), the analysis device (100) can detect a crack location (31) that is likely to occur based on strain energy.

[0074] The analysis device (100) can detect a location with high strain energy from image data of the spent nuclear fuel (10) and calculate a hydraulic fracture point, and determine this as a crack location (31).

[0075] Referring to Fig. 6(b), when the analysis device (100) detects a location where a crack may occur through strain energy, it can calculate a path by using the Dijkstra algorithm to connect points with low resistance to load, i.e., points with a high probability of crack occurrence, from the crack location (31).

[0076] The analysis device (100) can calculate the shortest path connecting points with a high probability of crack occurrence with respect to the length and width of the sample crack as the crack path (predicted crack path) (32).

[0077] FIG. 7 is a flowchart illustrating an analysis method for predicting a crack path of a spent nuclear fuel cladding tube according to one embodiment of the present invention.

[0078] Referring to FIG. 7, the analysis device (100) analyzes image data captured by the camera (150) and processes the image data (S110). After image processing, the analysis device (100) performs finite element analysis (S120) and can calculate a crack path using the Dijkstra algorithm based on strain energy (S130).

[0079] In the image processing step (S110), the processor (110) extracts a region of interest from the image data to remove unnecessary parts from the image data (S310).

[0080] The processor (110) performs a morphological operation to enhance the morphological features of the hydride to remove the hydride included in the image data (S320). By emphasizing the features of the hydride so that they appear clearly, the processor (110) makes it possible to distinguish the hydride from the cladding tube.

[0081] The processor (110) divides the image to distinguish between the hydride and the cladding (S330). Additionally, the processor (110) can remove noise from the divided image to improve the quality of the image (S340).

[0082] In the finite element analysis step (S120), the processor (110) generates a pixel-based finite element model based on the segmented image (S350). The processor (110) defines material properties (S360) and defines boundary conditions (S370). At this time, the processor (110) uses the values ​​derived from the nano-indentation test as the elastic modulus and Poisson's ratio of the cladding and the hydride, applies boundary conditions identical to those of the ring compression test (RCT), and can simulate the pinch load of the ring compression test (RCT) to evaluate the integrity of the cladding.

[0083] The processor (110) analyzes the strain energy distribution by distinguishing between the cladding and the hydride based on the set conditions (S380).

[0084] In the step of calculating the crack path (S130), the processor (110) generates pixel-based nodes and edges (S390), applies strain energy to the hydride (S400), and calculates the shortest path connecting points with high strain energy (S410).

[0085] The processor (110) calculates the shortest path as the crack path (predicted crack path) and outputs it as an image (S420). Additionally, the processor (110) can verify the crack path by comparing the crack of the ring compression test with the predicted crack path.

[0086] FIG. 8 is a figure referenced to explain a method for processing an image of a spent nuclear fuel cladding tube according to one embodiment of the present invention.

[0087] Referring to FIG. 8 (a), the processor (110) can process image data for the cladding of spent nuclear fuel (10).

[0088] The processor (110) extracts the cladding region as a region of interest to remove unnecessary parts from the image data. The processor (110) can extract the cladding region using a circle formula passing through three points. At this time, the cladding is displayed in white in the image data, but hydrides are included in the image data as hydrides are generated due to radiation.

[0089] Referring to FIG. 8(b), the processor (110) performs morphological operations on the region of interest. The processor (110) can emphasize the morphological features of the hydride through morphological operations to distinguish between the hydride and the cladding. The processor (110) can repeat expansion and erosion through morphological operations so that the shape of the hydride becomes clearly visible.

[0090] Referring to Fig. 8 (c), the processor (110) divides the image of the hydride and the cladding when the shape of the hydride is clearly visible through morphological operations. The processor (110) can distinguish the cladding and the hydride by binarizing the image using the Otsu algorithm.

[0091] Referring to FIG. 8 (d), the processor (110) can remove noise from segmented image data to improve the quality of the image. The processor (110) identifies small-sized hydrides as noise and removes them.

[0092] FIG. 9 is a figure referenced to explain a finite element model based on an image of a spent nuclear fuel cladding tube according to one embodiment of the present invention.

[0093] Referring to FIG. 9 (a), when the image processing step is completed, the processor (110) generates a pixel-based finite element model from the image data for the cladding tube.

[0094] The processor (110) uses the values ​​derived from the nano-indentation test as the elastic modulus and Poisson's ratio of the cladding tube and the hydride, applies boundary conditions under the same conditions as the ring compression test, and evaluates the integrity of the cladding tube by simulating the pinch load calculated through the ring compression test. The processor (110) can apply a cosine-shaped load condition.

[0095] At this time, the length of the arrow may vary in proportion to the magnitude of the load acting on the cladding tube. The load condition formed by the arrow may be in the form of a cosine. Additionally, the processor (110) can indicate the location where strain energy acts on the image data.

[0096] Referring to FIG. 9 (b) and (c), the processor (110) can display an image in pixel units. The processor (110) can indicate the position of the cladding as 1 and the position of the hydride as 0.

[0097] The processor can generate a finite element model (FE Model) by distinguishing between the hydride and the cladding.

[0098] FIG. 10 is a figure referenced to explain a method for predicting a crack path using an algorithm based on strain energy according to one embodiment of the present invention.

[0099] Referring to FIG. 10 (a) to (c), the processor (110) generates nodes and edges based on pixels from image data for a cladding tube for which image processing is completed. The processor (110) applies strain energy to the hydride. The processor (110) may apply range scaling to SED to be less than 1 to set the shortest path.

[0100] The processor (110) can assign weights by quantifying the probability of an external load being applied based on strain energy. The processor (110) can apply weights to the nodes as shown in Table 1.

[0101] End Node End Node Weight Weight n7 n12 0 (SED5+SED6) / 2 n7 n8 0.5 (1+SED6) / 2 n7 n2 1 1 n7 n6 0.5 (1+SED5) / 2

[0102] The processor (110) can assign weights to the edges connecting to adjacent nodes centered around node 7 (n7), namely node 12 (n12), node 2 (n2), node 8 (n8), and node 6 (n6). For example, the processor (110) can calculate the average by summing SED 5 and SED 6 for node 7 (n7) and node 12 (n12) and dividing by 2, and set the calculated value as the weight. In this case, since SED5 and SED6 are 0, the weight becomes 0. The processor (110) can calculate the average of SED6 and cladding 1 that are in contact with node 7 (n7) and node 8 (n8) and set the weight to 0.5. The processor (110) can set weights 1 and 0.5 for node 7 (n7) and node 2 (n2), and node 7 (n7) and node 6 (n6), respectively, in the same way.

[0103] The processor (110) can set the path with the smallest sum of weights as the shortest path. The processor (110) can search for the shortest path connecting the start node to the end node of the cladding. The processor (110) can search for the shortest path based on the Dijkstra algorithm. The processor (110) can calculate the shortest path as the crack path (predicted crack path).

[0104] FIG. 11 is a figure illustrating an example of verification of a predicted crack path according to an embodiment of the present invention. FIG. 11 shows the test results when pressure is applied to the 12 o'clock position of the cladding tube. Images 41 to 46 of FIG. 11 show the before and after of the test, distinguished by the temperature conditions and hydride content of the ring compression test (RCT).

[0105] Referring to FIG. 11, the processor (110) may simulate cracks in a cladding tube by generating actual cracks in a specimen that simulates the environment of the cladding tube through a ring compression test (RCT). In this case, the specimen may be made of Zircaloy-4 material and may be manufactured by hydrogenating it through a gas diffusion method to simulate the same environment as the cladding tube.

[0106] The processor (110) can obtain crack data from the results of a ring compression test (RCT) performed according to temperature conditions and hydride content. The processor (110) can receive the crack data as input or receive it through the communication unit (130).

[0107] The processor (110) can obtain images (41, 44) of the cladding tube before and after testing when the hydrogen content is 137 ppm at room temperature (RT), images (42, 45) of the cladding tube before and after testing when the hydrogen content is 129 ppm in an environment of 150 degrees Celsius, and images (43, 46) of the cladding tube before and after testing when the hydrogen content is 149 ppm in an environment of 200 degrees Celsius.

[0108] The processor (110) can determine similarity by comparing images before and after testing for each test condition with the predicted crack location and crack path. Since the arrangement of hydrides present in the cladding tube may change due to compression in the ring compression test, the processor (110) can determine similarity by considering the change in the arrangement of hydrides. The processor (110) can verify the predicted crack location and crack path based on the result of the similarity determination.

[0109] FIG. 12 is an example diagram comparing a predicted crack path and an actual crack path according to an embodiment of the present invention. FIG. 12 (a) is image data of a cladding tube before a ring compression test (RCT), FIG. 12 (b) is image data of a cladding tube after a ring compression test (RCT), and FIG. 12 (c) is image data of a cladding tube with a predicted crack path displayed.

[0110] Referring to Fig. 12(a), when pressure is applied to the 12 o'clock position of the specimen in a ring compression test (RCT), the greatest tensile stress occurs at the bottom of the 12 o'clock position. Accordingly, fracture occurs first in the specimen at the bottom of the 12 o'clock position.

[0111] In cases where the hydrogen content is 137 ppm at room temperature (RT) (51), at a temperature of 150 degrees and the hydrogen content is 129 ppm (52), and at a temperature of 200 degrees and the hydrogen content is 149 ppm (53), a crack occurs at any point between B' and C' at the bottom. At this time, the arrangement of hydrides in the specimens under each condition may appear differently.

[0112] Referring to Fig. 12(b), as a result of the ring compression test (RCT), cracks occur in the specimen of the cladding tube from the bottom to the top.

[0113] In the case where the hydrogen content is 137 ppm at room temperature (RT) (54), at a temperature of 150 degrees and a hydrogen content of 129 ppm (55), and at a temperature of 200 degrees and a hydrogen content of 149 ppm (56), a crack occurred from any point between B' and C' at the bottom and toward the top. The crack occurs differently depending on the arrangement of the hydride in the specimen.

[0114] Referring to Figure 12 (c), when comparing the predicted measurement path with the results of the ring compression test, it can be seen that the starting point of the predicted crack path is the point between B' and C' at the bottom of the specimen, indicating that the crack occurred at the same point.

[0115] It can be seen that the predicted crack path and the actual crack starting point are the same in all cases: when the hydrogen content is 137 ppm at room temperature (RT) (57), when the hydrogen content is 129 ppm in an environment with a temperature of 150 degrees (58), and when the hydrogen content is 149 ppm in an environment with a temperature of 200 degrees (59).

[0116] In addition, it can be seen that the predicted crack path when the hydrogen content is 137 ppm at room temperature (RT) differs somewhat from the actual crack, but the predicted crack path when the hydrogen content is 129 ppm (58) in an environment with a temperature of 150 degrees and when the hydrogen content is 149 ppm (59) in an environment with a temperature of 200 degrees is similar to the actual crack.

[0117] In this way, the analysis device (100) can produce a predicted value similar to the actual crack by predicting the crack location and crack path using the Dijkstra algorithm based on strain energy.

[0118] FIG. 13 is an example diagram comparing a predicted crack path according to one embodiment of the present invention with a crack path according to a different analysis method. FIG. 13 (a) is a diagram showing a predicted crack path using the strain energy-based Dijkstra algorithm of the present invention, FIG. 13 (b) is a diagram showing a predicted crack path using the Dijkstra algorithm, and FIG. 13 (c) is a diagram showing a predicted crack path using the genetic algorithm.

[0119] Referring to FIG. 13 (a), in the predicted crack path calculated using the strain energy-based Dijkstra algorithm of the present invention, for the case where the hydrogen content is 137 ppm at room temperature (RT) (61), the case where the hydrogen content is 129 ppm at a temperature of 150 degrees (62), and the case where the hydrogen content is 149 ppm in an environment of 200 degrees (63), it can be seen that the crack location (cracking start point) at the bottom of the specimen is the same as the crack location of the actual crack in FIG. 12 described above.

[0120] In addition, it can be seen that the predicted crack path of the present invention is similar to the actual crack when the hydrogen content is 129 ppm (62) in an environment with a temperature of 150 degrees and when the hydrogen content is 149 ppm (63) in an environment with a temperature of 200 degrees.

[0121] Referring to Fig. 13 (b), in the predicted crack path calculated using the Dijkstra algorithm, the crack location (cracking starting point) and crack path when the hydrogen content at room temperature (RT) is 137 ppm (64) have some similarities to the actual crack.

[0122] However, in the case where the hydrogen content is 129 ppm in an environment with a temperature of 150 degrees (65), the crack location in the predicted crack path is located between A' and B', which is different from the actual crack, and the crack path is also different from the actual crack.

[0123] In addition, in the case where the hydrogen content is 149 ppm in an environment with a temperature of 200 degrees (66), the crack location in the predicted crack path is different from the actual crack, being adjacent to C', and the crack path is also different.

[0124] Referring to Fig. 13 (c), in the predicted crack path calculated using a genetic algorithm, the crack location (cracking starting point) and crack path when the hydrogen content at room temperature (RT) is 137 ppm (66) have some similarities to the actual crack.

[0125] However, in the predicted crack path of the case where the hydrogen content is 129 ppm in an environment with a temperature of 150 degrees (67), the crack location is located between A' and B', which is different from the actual crack, and the crack path is also different from the actual crack.

[0126] In addition, in the predicted crack path of the case where the hydrogen content is 149 ppm in an environment with a temperature of 200 degrees (68), the crack location is located between C' and D', so it can be seen that the location is different from the actual crack and the crack path is also different.

[0127] In this way, the analysis device (100) can predict the crack location and crack path using the Dijkstra algorithm based on strain energy, thereby producing a predicted value that is highly similar to the actual crack compared to other analysis methods.

[0128] Accordingly, the device and method for analyzing the crack path of a spent nuclear fuel cladding according to one aspect of the present invention can more accurately predict the location and path of cracks in the cladding of spent nuclear fuel, prevent accidents caused by cracks, and safely manage spent nuclear fuel.

[0129] Although the present invention has been described with reference to the embodiments illustrated in the drawings, this is merely illustrative, and those skilled in the art will understand that various modifications and equivalent alternative embodiments are possible therefrom.

[0130] Therefore, the true technical scope of protection of the present invention should be determined by the following patent claims. Explanation of the symbols

[0131] 10: Spent nuclear fuel 100: Analysis device 110: Processor 120: Memory 130: Communications Department 140: Sensor 150: Camera

Claims

Claim 1 A crack path analysis device for a spent nuclear fuel cladding tube, characterized by comprising: a camera for photographing the cladding tube of spent nuclear fuel; and a processor for preprocessing image data input from the camera, predicting the location where a crack will occur by applying strain energy to the preprocessed image data, and predicting a crack path starting from the crack location. Claim 2 A crack path analysis device for spent nuclear fuel cladding according to claim 1, characterized in that the processor determines the location where the magnitude of the strain energy is greatest as the location where a crack may occur and predicts the crack location. Claim 3 A crack path analysis device for a spent nuclear fuel cladding according to claim 1, wherein the processor distinguishes hydrides generated by radiation from the image data and applies strain energy to the location of the hydrides. Claim 4 A crack path analysis device for spent nuclear fuel cladding according to claim 1, characterized in that the processor generates nodes and edges in the image data based on pixels and calculates the crack path by assigning weights according to the strain energy to the edges. Claim 5 A crack path analysis device for spent nuclear fuel cladding according to claim 4, characterized in that the processor calculates the crack path as the path with the smallest sum of weights among a plurality of paths starting from the crack location. Claim 6 A crack path analysis device for spent nuclear fuel cladding according to claim 1, characterized in that the processor predicts the crack path by applying the Dijkstra algorithm. Claim 7 A crack path analysis device for spent nuclear fuel cladding according to claim 1, wherein the processor calculates similarity by comparing crack data regarding actual cracks occurring in the cladding with the crack location and the crack path, and verifies the crack location and the crack path according to the similarity. Claim 8 A crack path analysis device for spent nuclear fuel cladding according to claim 7, wherein the processor obtains crack data from the result of performing a ring compression test (RCT) on a specimen simulating the environment of the cladding, and compares the crack data with the crack location and the crack path. Claim 9 A crack path analysis device for spent nuclear fuel cladding according to claim 1, wherein the processor distinguishes between hydrides and cladding from image data and removes noise. Claim 10 A crack path analysis device for spent nuclear fuel cladding according to claim 9, characterized in that the processor identifies the shape of the hydride through morphological operations, distinguishes the hydride and the cladding through an Oates algorithm, and determines small hydrides as noise and removes them from the image data. Claim 11 A crack path analysis device for spent nuclear fuel cladding according to claim 1, characterized in that the processor generates a pixel-based finite element model from the image data, defines physical properties and boundary conditions for the finite element model, and analyzes the strain energy distribution by distinguishing the cladding and the hydride from the image data. Claim 12 A method for analyzing a crack path of a spent nuclear fuel cladding, characterized by comprising: a step of a processor preprocessing image data of a spent nuclear fuel cladding; a step of the processor predicting a crack location where a crack will occur by applying strain energy to the preprocessed image data; and a step of the processor predicting a crack path starting from the crack location. Claim 13 A method for analyzing the crack path of a spent nuclear fuel cladding, characterized in that, in the step of predicting the crack location, the processor determines the location where the magnitude of the strain energy is greatest as the location where a crack may occur and predicts the crack location. Claim 14 A method for analyzing the crack path of a spent nuclear fuel cladding, characterized in that, in the step of predicting the crack location, the processor distinguishes hydrides generated by radiation from the image data and applies the strain energy to the location of the hydrides in claim 13. Claim 15 A method for analyzing a crack path of a spent nuclear fuel cladding, characterized in that, in claim 12, the step of predicting the crack location comprises: generating a pixel-based finite element model from the image data; defining material properties and boundary conditions for the finite element model; and analyzing the strain energy distribution by distinguishing the cladding and the hydride from the image data. Claim 16 A method for analyzing a crack path of a spent nuclear fuel cladding, characterized in that, in claim 12, the step of predicting the crack path comprises: a step of generating nodes and edges in the image data based on pixels; a step of assigning weights according to the strain energy to the edges; and a step of calculating the crack path connecting the start node to the last node. Claim 17 A method for analyzing a crack path of a spent nuclear fuel cladding, characterized in that, in the step of predicting the crack path, the processor calculates the crack path having the smallest sum of weights among a plurality of paths starting from the crack location. Claim 18 A method for analyzing a crack path of a spent nuclear fuel cladding, characterized in that, in the step of predicting the crack path, the processor predicts the crack path by applying the Dijkstra algorithm. Claim 19 A method for analyzing crack paths in spent nuclear fuel cladding according to claim 12, wherein the step of preprocessing the image data comprises: a step of confirming the shape of the hydride from the image data; a step of distinguishing the hydride and the cladding from the image data; and a step of determining small hydrides as noise and removing them from the image data. Claim 20 A method for analyzing a crack path of a spent nuclear fuel cladding according to claim 12, further comprising: a step of, after the step of predicting the crack path, a step in which the processor compares crack data for an actual crack occurring in the cladding with the crack location and the crack path; and a step of calculating similarity according to the comparison result and verifying the crack location and the crack path in correspondence with the similarity.

Images