Method and apparatus for evaluating the remaining lifespan of polymer materials.

Abstract

To provide a remaining life evaluation method of a polymer material capable of evaluating the remaining life of a polymer material more precisely than before, and a remaining life evaluation device.SOLUTION: The remaining life evaluation method includes: a step of creating a deterioration prediction model in which an FT-IR spectrum of a polymer material and mechanical properties thereof are associated with each other; and a step of acquiring the FT-IR spectrum and mechanical properties of an actual used polymer material. The remaining life evaluation is obtained by correlating the deterioration prediction model with the mechanical properties of the actually used polymer material from multiple regression analysis using the intensity or area of multiple peaks in the FT-IR spectrum which change due to polymer material deterioration but do not affect each other.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention preferably relates to a method and apparatus for evaluating the remaining life of polymer materials used in nuclear power plants. [Background technology]

[0002] As an example of a method for determining the remaining lifespan of a polymer material that can easily and accurately diagnose the remaining lifespan of a polymer material due to external factors such as radiation, Patent Document 1 describes a method for determining the remaining lifespan of a polymer material, which includes the steps of: irradiating a polymer material of the same material as the polymer material to be determined with gamma rays to determine the correlation between the change in the infrared absorption spectrum of the surface and the mechanical properties or absorbed dose; measuring the infrared absorption spectrum of the surface of the polymer material to be determined; and comparing the measurement results of the infrared absorption spectrum with the limit values ​​of the mechanical properties or absorbed dose by applying the correlation relationship. The method for determining the remaining lifespan of the polymer material to be determined is based on the relationship between the estimated value estimated from the measurement results based on the correlation and the limit values ​​of the mechanical properties or absorbed dose, or the ratio of the estimated value to the limit value. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-000614 [Overview of the project] [Problems that the invention aims to solve]

[0004] In nuclear power plants, polymer materials are used in piping, joints, packings, gaskets, O-rings, and other applications. It is known that these polymer materials deteriorate when exposed to heat and radiation for extended periods, as their chemical structure changes.

[0005] Such polymer materials are typically maintained based on time, being replaced after a certain period of use. However, if the lifespan of polymer materials can be properly evaluated, it will be possible to reduce the frequency of replacement, and thus reduce maintenance work, while ensuring safety and reducing the radiation exposure of workers. For this reason, many methods for diagnosing the degradation and remaining lifespan of polymer materials have been disclosed.

[0006] Patent Document 1 discloses that by using a single peak (carbonyl group) that changes due to degradation of a polymer material by gamma-ray irradiation, it is possible to obtain a correlation between the change in that peak and mechanical properties such as permanent compressive strain and tensile strength. Specifically, it shows the correlation between the amount of gamma-ray irradiation and the peak intensity or area, and the correlation between the peak area or area and mechanical properties. Based on these correlations, it is possible to evaluate the degradation and remaining life of the polymer material.

[0007] The aforementioned Patent Document 1 is a technique for obtaining a correlation between one or two peaks that change due to degradation in the FT-IR spectrum and mechanical properties used as indicators of degradation.

[0008] However, Patent Document 1 suggests that when the main component of a polymer material has multiple degradation mechanisms or when the polymer material contains multiple additives, it is difficult to adequately represent the degradation state with data analysis targeting one or two peaks on FT-IR, indicating room for improvement.

[0009] Furthermore, when dealing with three or more peaks, methods such as calculating the peak ratio are difficult, and there is room for improvement.

[0010] This invention provides a method and apparatus for evaluating the remaining lifespan of polymer materials, which can evaluate the remaining lifespan of polymer materials with higher accuracy than conventional methods. [Means for solving the problem]

[0011] The present invention includes several means for solving the above problems, but one example is: Contains main ingredients and additives A method for diagnosing the remaining life of polymer materials, A first step of obtaining the FT-IR spectrum and mechanical properties of the polymer material, and a second step of applying degradation factors expected at the application site in a nuclear power plant to the polymer material, and obtaining the FT-IR spectrum and mechanical properties of the polymer material in a degraded state. A step of creating a deterioration prediction model that correlates the FT-IR spectrum of the polymer material with its mechanical properties, of comprising, the deterioration prediction model Acquired in the first step of the polymer material From the FT-IR spectral results and the FT-IR spectral results of the degraded polymer material obtained in the second step, which changes due to deterioration and and is obtained by correlating the mechanical properties of the polymer material with the intensities or areas of a plurality of peaks in the FT-IR spectrum that do not affect each other, This is due to the chemical structure of the main ingredient and additives. from multiple regression analysis using the intensities or areas of a plurality of peaks in the FT-IR spectrum before of the polymer material machine shall be characterized in that it is obtained by correlating with the mechanical properties.

Advantages of the Invention

[0012] According to the present invention, the remaining life of a polymer material can be evaluated with higher accuracy than in the prior art. Other problems, configurations, and effects than those described above will be clarified by the description of the following examples.

Brief Description of the Drawings

[0013] [Figure 1] Schematic configuration diagram of the remaining life evaluation apparatus for the polymer material of the embodiment. [Figure 2] Process diagram of the remaining life evaluation method for the polymer material of the embodiment. [Figure 3] A diagram showing an example of the FT-IR spectrum of an O-ring made of acrylonitrile nitrile butadiene rubber (NBR) before and after deterioration in Example 1. [Figure 4] Deterioration prediction model (linear regression) using changes in multiple peaks of the FT-IR spectrum in compression set in Example 1. [Figure 5] Deterioration prediction model (linear regression) using changes in multiple peaks of the FT-IR spectrum in hardness change rate in Example 1. [Figure 6] Deterioration prediction model (linear regression) adding environmental factors as explanatory variables in Example 2. [Figure 7] Deterioration prediction model (non-linear regression) using changes in multiple peaks of the FT-IR spectrum in compression set in Example 3. [Figure 8] A degradation prediction model (nonlinear regression) using the change in multiple peaks of the FT-IR spectrum in the hardness change rate in Example 3. [Figure 9] A degradation prediction model (linear regression) using a single peak (derived from the butadiene group) in the FT-IR spectrum at compression set in the comparative example. [Figure 10] A degradation prediction model (linear regression) using a single peak (derived from the butadiene group) in the FT-IR spectrum of the hardness change rate in the comparative example. [Modes for carrying out the invention]

[0014] Examples of the remaining life evaluation method and remaining life evaluation apparatus for polymer materials of the present invention will be described with reference to Figures 1 to 10. In the drawings used herein, the same or corresponding components are denoted by the same or similar reference numerals, and repeated descriptions of these components may be omitted.

[0015] First, the overall configuration of the polymer material remaining life evaluation device according to an embodiment of the present invention will be described with reference to Figure 1. Figure 1 is a diagram showing the schematic configuration of the polymer material remaining life evaluation device according to an embodiment.

[0016] The remaining life evaluation device 1 shown in Figure 1 is a device for diagnosing the remaining life of polymer materials and includes a target product data acquisition unit 11, a target product degradation unit 12, a degraded product data acquisition unit 13A, a database unit 13B, a degradation prediction model creation unit 14, an actual used degraded product data acquisition unit 15, an update unit 17, a remaining life estimation unit 16, a display unit 18, and the like.

[0017] Of the remaining life evaluation device 1, the target product data acquisition unit 11, the deteriorated product data acquisition unit 13A, and the actual use deteriorated product data acquisition unit 15 are composed of measuring instruments, while the target product deterioration unit 12 is composed of various experimental and measuring instruments.

[0018] The database unit 13B, the degradation prediction model creation unit 14, the update unit 17, the remaining lifespan estimation unit 16, and the display unit 18 can be implemented by a PC (Personal Computer) equipped with hardware such as a computing device like a CPU, a main memory device like semiconductor memory and an auxiliary storage device like a hard disk, an input device like a keyboard or USB port, and an output device like a monitor. The operation of each device and various calculation processes described later are executed based on various programs.

[0019] The program may be stored in internal memory, external storage media, or a data server (not shown in the diagram), and read and executed by the CPU. Control processing may be combined into a single program, divided into multiple programs, or a combination of these. Furthermore, some or all of the program may be implemented using dedicated hardware or modularized. Additionally, various programs may be installed from a program distribution server, internal storage media, or external storage media.

[0020] Details of each of these parts will be described later.

[0021] Next, the overall process of the remaining life evaluation method, which includes a degradation prediction model for polymer materials according to an embodiment of the present invention, will be explained with reference to Figure 2. Figure 2 is a process diagram of the remaining life evaluation method for polymer materials according to an embodiment.

[0022] As shown in Figure 2, the overall process of the polymer material remaining life diagnosis method of the present invention consists of seven steps: target product data acquisition step S1, target product degradation step S2, degraded product data acquisition and database creation step S3, degradation prediction model creation step S4, actual used degraded product data acquisition step S5, remaining life estimation and display step S6, and update step S7. Each step will be described below.

[0023] [Target product data acquisition process S1] The target product data acquisition process S1 is preferably performed by the target product data acquisition unit 11, and is a process of performing chemical structure analysis and mechanical property measurement on the target polymer material before use.

[0024] The target product data acquisition unit 11 preferably uses the same measurement principle and measuring device as the deteriorated product data acquisition unit 13A and the actual used deteriorated product data acquisition unit 15, which will be described later.

[0025] For chemical structure analysis, at least FT-IR measurements should be performed. Other analytical methods that can track changes caused by the degradation of the target polymer material may also be used, and these can be added as explanatory variables when constructing degradation prediction models as needed.

[0026] In particular, a handheld FT-IR device is desirable for evaluating remaining life in environments where polymer materials are used.

[0027] Mechanical properties are determined based on the characteristics required for the application of the material. For example, this might include hardness measured by a durometer, compression set for O-rings, or tensile strength for piping.

[0028] The chemical structure analysis method and mechanical property measurement method used in this invention can be appropriately modified depending on the target polymer material and its application, but the chemical structure analysis method must always include FT-IR measurement.

[0029] The polymer materials in question are preferably materials for nuclear power plants, such as polymer materials for O-rings, gaskets, and piping used in nuclear power plants, but are not particularly limited.

[0030] More specifically, examples include diene-based elastomers such as acrylonitrile butadiene rubber, ethylene propylene diene rubber, polypropylene, polyethylene, vinyl chloride, chloroprene rubber, and butyl rubber; fluororubber; fluororesins such as tetrafluoroethylene, perfluoroalkoxyalkane, and polychlorotrifluoroethylene; silicone rubber; and styrene butadiene rubber.

[0031] [Target product deterioration process S2] The target product degradation process S2 is preferably carried out by the target product degradation unit 12, and is a process in which the target polymer material, for which pre-use data was obtained in the target product data acquisition unit 11 / target product data acquisition process S1, is subjected to a load of degradation factors expected at the application site in a nuclear power plant.

[0032] Degradation factors include heat, radiation, and moisture. Degradation will be carried out using these factors individually, in combination, or in any other way that is anticipated. Accelerated degradation will be performed if necessary.

[0033] [Degraded product data acquisition and database creation process S3] The degradation product data acquisition and database creation process S3 is preferably carried out by the degradation product data acquisition unit 13A and the database unit 13B, and is a process of performing the same analysis and measurements on the target polymer material that was degraded in the target product degradation unit 12 / target product degradation process S2 as the chemical structure analysis and mechanical property measurement performed on the target polymer material before degradation, and a process of creating a database of the degradation conditions, chemical structure analysis results and mechanical property measurement results as a set.

[0034] As described above, the deteriorated product data acquisition unit 13A is preferably an apparatus and method with the same measurement principle as the target product data acquisition unit 11 and the actual used deteriorated product data acquisition unit 15 described later, and more preferably the same measuring apparatus. Even if the measuring apparatus is not the same, it is possible to utilize the data from each apparatus by multiplying it by a correction coefficient. In this respect, it is also possible to utilize publicly available data such as public databases and research papers.

[0035] [Degradation prediction model creation process S4] The deterioration prediction model creation step S4 is preferably carried out by the deterioration prediction model creation unit 14, and is a step in which a correlation model is created between the chemical structure analysis results by multiple regression analysis and the mechanical property measurement results using the deteriorated product data acquisition unit 13A and the database unit 13B / deteriorated product data acquisition and database creation step S3.

[0036] Step S4, the process of creating a degradation prediction model, is the process of creating a degradation prediction model that correlates the FT-IR spectrum of a polymer material with its mechanical properties. Furthermore, the degradation prediction model created here is obtained by correlating the mechanical properties of the polymer material of an actual product with the mechanical properties of the polymer material by performing a multiple regression analysis using the intensity or area of ​​multiple peaks in the FT-IR spectrum that change due to the degradation of the polymer material but do not influence each other. In addition, any type of regression model, including linear regression and nonlinear regression, can be applied when performing the multiple regression analysis.

[0037] In this invention, "products used in actual use" includes both polymer materials in a degraded state that are the target of model construction by acquiring data in the laboratory beforehand for polymer materials that are actually used, and polymer materials that have become degraded through actual use in the field.

[0038] When using FT-IR for chemical structure analysis, multiple peaks that change due to degradation by various factors on the FT-IR spectrum but do not influence each other are selected and used as feature variables (explanatory variables).

[0039] In particular, multiple peaks should include peaks representing the main components, such as the main component and additives, that make up the polymer material components. Furthermore, it is desirable to use the peak intensity ratio or peak area ratio of three or more FT-IR spectra as explanatory variables for multiple regression analysis. Additionally, multiple peaks should be analyzed to reflect changes in the polymer material of the product in actual use. - It is desirable to do so.

[0040] [Process S5: Acquisition of data on products that have deteriorated under actual use] The process S5 for acquiring data on degraded products in actual use is preferably carried out by the data acquisition unit 15 for degraded products in actual use, and is a process for performing chemical structure analysis, including FT-IR spectroscopy, and mechanical property measurements of polymer materials of products in actual use at nuclear power plants and the like.

[0041] The analysis and measurement performed in this actual use degradation product data acquisition process S5 shall be the same analysis and measurement methods that were used on the polymer material before use and the polymer material after degradation treatment in the target product data acquisition process S1 and the degradation product data acquisition and database creation process S3.

[0042] Furthermore, it is desirable to simultaneously acquire data on the usage environment of the target polymer material (temperature, humidity, and radiation (gamma ray) levels). This will ultimately make it possible to predict the degree of degradation from the usage environment data, enabling more accurate predictions of the degree of degradation. Moreover, if the degree of degradation can be predicted from the usage environment data, it may become possible to eliminate the need for on-site measurements.

[0043] Furthermore, if the mechanical properties of the target polymer material can be measured, it is desirable to use the degradation prediction model to update the degradation prediction model by reflecting the results of chemical structure analysis and mechanical property measurement of the actual product in the degradation prediction model recorded in the database unit 13B by the update unit 17 / update process S7 described later.

[0044] [Update process S7 for actual use and deterioration data] The update process S7 is preferably performed by the update unit 17, and is a process in which the degradation prediction model recorded in the database unit 13B is updated to a model that is more suitable for the application environment by adding a dataset consisting of chemical structure analysis results and mechanical property measurement results of the polymer material after actual use, obtained in the actual use degradation product data acquisition unit 15 / actual use degradation product data acquisition process S5, to the constructed database unit 13B.

[0045] [Remaining life estimation / display process S6] The remaining life estimation and display process S6 is preferably carried out by the remaining life estimation unit 16 and the display unit 18, and is a process in which the degree of deterioration of mechanical properties is predicted using the deterioration prediction model created in the deterioration prediction model creation unit 14 / deterioration prediction model creation process S4, based on the results of chemical structure analysis after use obtained in the actual use deterioration product data acquisition unit 15 / actual use deterioration product data acquisition process S5.

[0046] When FT-IR is used for chemical structure analysis, the input factors are multiple peaks that change with degradation, selected in the degradation prediction model creation process, and the output factors are either the degree of degradation or the remaining lifespan of the target polymer material.

[0047] The remaining lifespan is calculated by first determining the values ​​of the mechanical characteristics, which are designated as degradation indicators, at the end of their lifespan, and then comparing those values ​​with the current degree of degradation.

[0048] Furthermore, using FT-IR spectra, it is possible to interpret changes in chemical structure due to degradation from the changes in each peak, and to understand the contribution of each explanatory variable in the degradation prediction model, thereby allowing for the inference of the degradation mechanism. Therefore, by comparing the spectral data stored in the database unit 13B in advance, not just the differences as input factors in the prediction model, it is possible to understand the differences in the degradation mechanisms of products in actual use.

[0049] If the degradation mechanism estimated from an actual used product differs from the degradation mechanism estimated from the spectral data stored in the database unit 13B, it is possible to correct the degradation and remaining life predictions based on the degradation mechanism of the actual used product.

[0050] Next, the effects of this embodiment will be described.

[0051] The method for diagnosing the remaining life of a polymer material in this embodiment, as described above, comprises the steps of creating a degradation prediction model that correlates the FT-IR spectrum and mechanical properties of the polymer material, and acquiring the FT-IR spectrum and mechanical properties of a polymer material in actual use. The degradation prediction model is obtained by correlating it with the mechanical properties of the polymer material in actual use through multiple regression analysis using the intensity or area of ​​multiple peaks in the FT-IR spectrum that change due to the degradation of the polymer material and do not influence each other.

[0052] This makes it possible to accurately predict the degree of degradation of a target polymer material, even in complex materials where multiple peaks change due to degradation in the FT-IR spectrum.

[0053] Furthermore, by incorporating the FT-IR spectra and mechanical properties of the polymer materials of the actual products used into the degradation prediction model, it is expected that the model will be updated as needed according to the usage environment, thereby improving the prediction accuracy.

[0054] Furthermore, multiple peaks include peaks of polymer materials that are the main constituent components of the polymer material parts being studied, the explanatory variables for multiple regression analysis use the peak intensity ratio or peak area ratio of three or more FT-IR spectra, and the multiple peaks change due to the degradation of the polymer material in the actual product. - By doing so, it becomes possible to evaluate the degree of degradation with higher accuracy.

[0055] Furthermore, by using linear regression instead of multiple regression analysis, it becomes possible to easily evaluate the degree of deterioration from the standardized partial regression coefficients of each explanatory variable, and to understand the contribution of each explanatory variable to the deterioration.

[0056] Furthermore, by using a nonlinear regression method, such as random forest regression, for multiple regression analysis, evaluation can be performed with even higher accuracy.

[0057] The present invention will be described in detail below with reference to examples and comparative examples, but it is not limited in any way to the following examples.

[0058] [Example 1] The method for creating the degradation prediction model in the remaining life prediction method of Example 1 is described below.

[0059] The degradation prediction model and remaining life evaluation for Example 1 were performed using the following procedure.

[0060] In the data acquisition process for the target product, O-rings used in nuclear power plants were selected as the target polymer material. The O-rings were constructed primarily from acrylonitrile butadiene rubber (NBR), a commonly used rubber material, with additives such as plasticizers added to provide flexibility.

[0061] For chemical structure analysis, FT-IR was performed. For mechanical properties, compression set and hardness using a durometer were measured. These mechanical properties were used as indicators to understand the degree of degradation.

[0062] In the degradation process for the target product, the degradation factors were either heat, radiation, or both. When radiation and heat were used as degradation factors, sequential degradation was performed, involving degradation by radiation in the atmosphere followed by heating in nitrogen or hot water. The heat was applied under either nitrogen or hot water conditions.

[0063] Table 1 shows the specific degradation conditions.

[0064] [Table 1]

[0065] In the degradation product data acquisition and database creation process S3, FT-IR measurements, compression set, and hardness change rate were measured, similar to the target product data acquisition process S1. The specific FT-IR data and database creation are described below.

[0066] For FT-IR measurement, a Fourier transform infrared analyzer Nicolet 6700 (manufactured by Thermo Fisher Scientific) was used. The measurement conditions were a resolution of 4 cm -1 , an integration number of 32 times, and a measurement wave number range of 650 - 4000 cm -1 , and the measurement was carried out by ATR measurement.

[0067] Figure 3 shows the FT-IR spectra of acrylonitrile nitrile butadiene rubber (NBR), which is the target O-ring material. Figure 3 shows the spectra before and after deterioration respectively. The deteriorated products shown here are in the case where heat is the deterioration factor, and the deterioration conditions are 120 °C for 140 hours under a nitrogen environment.

[0068] As shown in Figure 3, before deterioration, as the main peaks resulting from the chemical structure of NBR, the butadiene group (965 cm -1 ), the nitrile group (2235 cm -1 ), and the CH group (2917 cm -1 ) were confirmed. Also, the peak (1259 cm ) was confirmed. Also, the peak (1259 cm -1 ) resulting from the phthalic acid-based material contained as a plasticizer, and the peak (1730 cm -1 ) resulting from additives and oxidation were confirmed.

[0069] On the other hand, after deterioration, in addition to the peaks confirmed before deterioration, the peaks (3361 cm -1 ) resulting from hydroxyl groups and amino groups were confirmed.

[0070] From the spectra in Figure 3, the peaks derived from the butadiene group, plasticizer, carbonyl group, and nitrile group, which change due to heating, do not drag each other, and have a large impact on mechanical properties, were extracted. In addition, as the numerical values used for creating the deterioration prediction model, the normalized peak intensity ratios of each peak, divided by the peak intensity of the CH group (2917 cm -1 ), were used.

[0071] Furthermore, when radiation (gamma rays), or both radiation (gamma rays) and heat, were used as degradation factors, and when the environment changed from nitrogen to hydrothermal fluid, similar peak changes tended to occur in the FT-IR spectrum. Therefore, we decided to use the same peaks to analyze samples degraded under other degradation conditions.

[0072] The compression set was calculated based on the dimensional change of the sample after heating it in a compressed state at a predetermined temperature and time, according to the following formula. The test method is specified in JIS K 6262. Compression set (%) = (Thickness of sample before heating - Thickness of spacer) / (Thickness of sample before heating - Thickness 30 minutes after pressure release)

[0073] Hardness was measured using a durometer, and the rate of change in hardness was calculated according to the following formula. Hardness change rate (%) = (Hardness after degradation) / (Hardness before heating) × 100

[0074] As described above, the peak intensity ratio contributing to O-ring degradation, extracted from the FT-IR spectrum, was used as the explanatory variable, and the compression set and hardness change rate were used as the dependent variables. These, along with degradation conditions and degradation environmental factors, were then compiled into a database as a dataset.

[0075] In the degradation prediction model creation process S4, a degradation prediction model was created by correlating the dependent variable and independent variables using multiple regression analysis based on the aforementioned database. In Example 1, this was carried out using multiple regression analysis with a linear regression model.

[0076] Figures 4 and 5 show degradation prediction models created using a linear multiple regression analysis of NBR degraded with various degradation factors in Example 1. The explanatory variables were the changes in peak intensity ratios (butadiene group, nitrile group, carbonyl group, plasticizer) in the FT-IR spectra, and the dependent variables were compression set and the durometer hardness change rate. Figure 4 shows the compression set, and Figure 5 shows the durometer hardness change rate. In both figures, the vertical axis represents the measured value, and the horizontal axis represents the predicted value from the prediction model.

[0077] Figures 4 and 5 show that the method described above made it possible to predict the compression set and hardness change rate from multiple peaks caused by the degradation of the FT-IR spectrum. The coefficient of determination (R) between the measured value and the predicted value at that time was calculated. 2 The values ​​were 0.79 and 0.71, indicating that it was a degradation prediction model with a high correlation between the measured and predicted values.

[0078] In the S5 process for acquiring data on degraded products in actual use, the FT-IR spectrum of a sample of an O-ring that had actually been used was measured, and the compression set and hardness change rate were predicted using the degradation prediction model created in the S4 process.

[0079] In actual use, it is assumed that mechanical properties cannot be measured. However, for accuracy verification, samples recovered after 46,080 hours of actual use under accelerated conditions equivalent to 30°C were evaluated.

[0080] Predicting the degradation of the O-ring after actual use from FT-IR spectra, the compression set was 43.6% and the hardness change rate was 3.4%. The relative errors with the measured values ​​were 37.0% for compression set and 5.0% for hardness change rate.

[0081] In the remaining life estimation and display process S6, the remaining life was calculated using the compression set calculated in the actual used product degradation data acquisition process S5, the lifespan at 80% compression set, and the actual usage time, according to the following formula. As a result, the estimated lifespan was 84,550 hours and the estimated remaining lifespan was 38,470 hours. Lifespan (hours) = 80% × Actual usage time (hours) / Predicted compression set (%) Remaining life (hours) = Life (hours) - Actual usage time (hours)

[0082] [Example 2] In the degradation prediction model creation process S4 described in Example 1, the environmental factors of whether the environment was under nitrogen or hydrothermal conditions were added to the dataset, and a degradation prediction model was created using multiple regression analysis with a linear model.

[0083] Figure 6 shows the created degradation prediction model. As shown in Figure 6, the coefficient of determination (R) in the degradation prediction model for compression set is 2 The coefficient of determination was 0.81, indicating that introducing environmental factors as explanatory variables improved the coefficient of determination.

[0084] [Example 3] In the degradation prediction model creation process S4 described in Example 2, a prediction model was created using multiple regression analysis with a random forest, which is a nonlinear regression model.

[0085] Figures 7 and 8 show the degradation prediction models created for compression set and durometer hardness change rate in Example 3, respectively. Figure 7 shows compression set, and Figure 8 shows durometer hardness change rate. In both figures, the vertical axis represents the measured value, and the horizontal axis represents the predicted value from the prediction model. The coefficient of determination (R) is calculated. 2 The coefficient of determination was 0.95 for compression set and 0.91 for hardness change rate, indicating that applying the nonlinear model improved the coefficient of determination compared to the linear model shown in Example 1.

[0086] [Example 4] In the degradation prediction model creation process S4 described in Example 1, the degradation prediction model was updated by adding data from actual used products, representing 30% of the data in the constructed database. Then, using the degradation prediction models before and after the update, degradation prediction of the mechanical properties of actual used products not included in the database was performed.

[0087] As a result, the predicted compression set values ​​obtained from both models showed a relative error of 37% before the update (results from Example 1), while after the update, predictions were possible with a relative error of 27.8%. This demonstrated that by adding data from actual used products to the database and updating the degradation prediction model, the accuracy of predicting the mechanical properties of actual used products can be improved.

[0088] [Example 5] In the remaining life evaluation method described in Example 1, the target material was changed from NBR to ethylene propylene diene rubber (EPDM).

[0089] In the degradation data acquisition and database creation process S3, the peak that changes due to degradation is identified from the spectral changes before and after degradation, specifically the CH group originating from the CH3 group (1370 cm⁻¹). -1 ), CC group (1600cm -1 ), C=O group (1730cm -1 ), CH group derived from CH2 group (2920 cm -1 ), peaks derived from plasticizers (1260 cm²) -1 ) were extracted. The values ​​used in creating the degradation prediction model were CH groups (2920 cm²) derived from CH2 units. -1 The normalized values ​​of each peak were used, obtained by dividing them by the peak intensity of ).

[0090] As a result, when a degradation prediction model was created in the same manner as in Example 1, the coefficient of determination (R 2 The result was 0.70, indicating that by appropriately extracting the peaks that change due to degradation even when the material type is changed, it was possible to predict the mechanical properties of the degraded product using the FT-IR spectrum.

[0091] [Example 6] In the degradation prediction model creation process S4 described in Example 1, a dataset was created using the peak intensity ratio as the peak area, and a degradation prediction model was created using multiple regression analysis.

[0092] As a result, the coefficient of determination (R 2 The value was 0.71, indicating that mechanical properties can be predicted using FT-IR spectra, similar to the case using peak intensity ratios.

[0093] [Comparative Example 1] Simple regression analysis was performed on each peak of the FT-IR spectrum extracted in Example 1, comparing it to mechanical properties (compression set, hardness change rate). The butadiene group showed the strongest correlation with the mechanical properties among the single peaks. The results of the simple regression analysis using only the peaks derived from the butadiene group are shown. A degradation prediction model was created in the same manner as in Example 1, except that only one peak derived from the butadiene group was used.

[0094] Figures 9 and 10 show the degradation prediction models created for compression set and durometer hardness change rate in Comparative Example 1, respectively. The vertical axis represents the measured values, and the horizontal axis represents the predicted values ​​from the prediction model.

[0095] As a result, the coefficient of determination (R 2 The values ​​were 0.47 for compression set and 0.35 for hardness change rate, which are significantly smaller than those of the linear model shown in Example 1. Although there is no absolute standard, considering that a coefficient of determination of 0.5 or higher is generally considered to indicate a strong correlation in statistics, it is thought that it is difficult to construct a sufficient degradation prediction model using regression with a single peak.

[0096] <Other> It should be noted that the present invention is not limited to the embodiments described above, and various modifications and applications are possible. The embodiments described above are explained in detail for the purpose of clearly illustrating the present invention, and are not necessarily limited to those having all the configurations described. [Explanation of Symbols]

[0097] 1…Remaining life evaluation device 11…Target product data acquisition unit 12…Deteriorated parts of the product 13A...Degraded product data acquisition unit 13B...Database Department 14…Degradation Prediction Model Creation Department 15...Acquisition unit for data acquisition of degraded products under actual use 17…Update Department 16...Remaining life estimation section 18...Display section

Claims

1. A method for diagnosing the remaining life of a polymer material including a main component and additives, A first step of obtaining the FT-IR spectrum and mechanical properties of the polymer material, A second step involves applying degradation factors expected in the application site within a nuclear power plant to the aforementioned polymer material, and obtaining the FT-IR spectrum and mechanical properties of the polymer material in a degraded state. The process includes creating a degradation prediction model by correlating the FT-IR spectrum and mechanical properties of the polymer material, The degradation prediction model is obtained by using multiple regression analysis based on the intensity or area of ​​multiple peaks in the FT-IR spectrum, which are caused by the chemical structures of the main component and additives that change due to degradation and do not affect each other, and by correlating these peaks with the mechanical properties of the polymer material. This analysis is performed using the FT-IR spectrum results of the polymer material obtained in the first step and the FT-IR spectrum results of the polymer material in a degraded state obtained in the second step. Methods for diagnosing remaining lifespan.

2. A method for diagnosing the remaining life of polymer materials, A first step of obtaining the FT-IR spectrum and mechanical properties of the polymer material, A second step involves applying a load of degradation factors expected in the application site within a nuclear power plant to the polymer material, and obtaining the FT-IR spectrum and mechanical properties of the polymer material in a degraded state. The process includes creating a degradation prediction model by correlating the FT-IR spectrum and mechanical properties of the polymer material, The degradation prediction model is obtained by correlating the FT-IR spectral results obtained in the first step and the FT-IR spectral results obtained in the second step with the mechanical properties of the polymer material through a nonlinear regression analysis using the intensity or area of ​​multiple peaks in the FT-IR spectrum that change due to the degradation of the polymer material and do not influence each other. Methods for diagnosing remaining lifespan.

3. In the remaining life diagnosis method according to claim 1 or 2, The process includes obtaining the FT-IR spectrum and mechanical properties of a polymer material used in actual operations at a nuclear power plant. Methods for diagnosing remaining lifespan.

4. In the remaining life diagnosis method described in claim 3, The degradation prediction model incorporates the results of the FT-IR spectrum and mechanical properties of the polymer material used in the nuclear power plant. Methods for diagnosing remaining lifespan.

5. In the remaining life diagnosis method according to claim 1 or 2, The explanatory variables for the aforementioned multiple regression analysis are the peak intensity ratios or peak area ratios of three or more of the aforementioned FT-IR spectra. Methods for diagnosing remaining lifespan.

6. In the remaining life diagnosis method according to claim 1, The above multiple regression analysis is replaced with linear regression. Methods for diagnosing remaining lifespan.

7. In the remaining life diagnosis method according to claim 1, The above multiple regression analysis is defined as nonlinear regression. Methods for diagnosing remaining lifespan.

8. In the remaining life diagnosis method according to claim 2 or 7, The aforementioned nonlinear regression is defined as random forest regression. Methods for diagnosing remaining lifespan.

9. In the remaining life diagnosis method according to claim 1 or 2, The polymer material is acrylonitrile butadiene rubber or ethylene propylene diene rubber. Methods for diagnosing remaining lifespan.

10. In the remaining life diagnosis method according to claim 1 or 2, The aforementioned polymer material is intended to be used as a material for nuclear power plants. Methods for diagnosing remaining lifespan.

11. In the remaining life diagnosis method according to claim 1 or 2, The system includes a remaining life estimation step that calculates the degree of degradation or remaining life of the polymer material based on the degradation prediction model. Methods for diagnosing remaining lifespan.

12. A device for diagnosing the remaining life of a polymer material, including a main component and additives, A target data acquisition unit for acquiring the FT-IR spectrum and mechanical properties of the polymer material, The polymer material is subjected to a load of degradation factors expected in its application site within a nuclear power plant, and the FT-IR spectrum and mechanical properties of the degraded polymer material are obtained from the degraded part of the target product. The system includes a degradation prediction model creation unit that creates a degradation prediction model by correlating the FT-IR spectrum and mechanical properties of the polymer material, The degradation prediction model is obtained by using multiple regression analysis based on the FT-IR spectrum results from the target data acquisition unit and the FT-IR spectrum from the degraded part of the target product. This analysis uses the intensity or area of ​​multiple peaks in the FT-IR spectrum resulting from the chemical structures of the main component and additives, which change due to degradation and do not influence each other, and correlates these peaks with the mechanical properties of the polymer material of the product in actual use. Remaining life diagnostic device.

13. A device for diagnosing the remaining lifespan of polymer materials, A target data acquisition unit for acquiring the FT-IR spectrum and mechanical properties of the polymer material, The polymer material is subjected to a load of degradation factors expected in its application site within a nuclear power plant, and the FT-IR spectrum and mechanical properties of the degraded polymer material are obtained from the degraded part of the target product. The system includes a degradation prediction model creation unit that creates a degradation prediction model by correlating the FT-IR spectrum and mechanical properties of the polymer material, The degradation prediction model is obtained by correlating the FT-IR spectral results from the target data acquisition unit and the FT-IR spectral results from the degraded part of the target product with the mechanical properties of the polymer material through a nonlinear regression analysis using the intensity or area of ​​multiple peaks in the FT-IR spectrum that change due to degradation and do not influence each other. Remaining life diagnostic device.

14. In the remaining life diagnostic device according to claim 13, The aforementioned nonlinear regression is defined as random forest regression. Remaining life diagnostic device.

Images