Method for predicting weathering resistance degradation of a curing agent based on finite element analysis

Abstract

The present application relates to the technical field of finite element analysis, and particularly relates to a method for predicting the weather resistance degradation of a curing agent based on finite element analysis. The method comprises the following steps: performing finite element analysis on the reaction molecular structure of the curing agent based on the material property data of the curing agent and the environmental data of the application scenario of the curing agent to obtain finite element curing agent reaction molecular structure data; performing analysis on the bond breaking behavior characteristics of the curing agent under the influence of the environment according to the finite element curing agent reaction molecular structure data to generate environmental influence curing agent bond breaking behavior characteristic data; establishing an environmental stress tensor-bond breaking evaluation model through the environmental influence curing agent bond breaking behavior characteristic data; and performing mapping optimization of the bond breaking and weather resistance degradation relationship of the curing agent on the environmental stress tensor-bond breaking evaluation model to obtain an environmental stress tensor-weather resistance degradation evaluation model. The present application realizes efficient prediction of the weather resistance degradation of the curing agent through finite element analysis technology.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of finite element analysis, and particularly relates to a prediction method for weather resistance degradation of a curing agent based on finite element analysis. BACKGROUND

[0002] With the development of computer simulation technology, the finite element analysis method is gradually introduced into the field of material performance research, and the mechanical model of the material is constructed to simulate its response under different loads. As the core additive for the molding and performance guarantee of high polymer materials, the stability of the weather resistance of the curing agent directly determines the service life and safety reliability of the terminal product in complex environments. Among many types of curing agents, such as methyl tetrahydrophthalic anhydride, due to its low toxicity, high reactivity and excellent curing effect, it is widely used in key fields such as electronic device packaging, composite material preparation and building coating. These application scenarios often come with harsh environmental tests, such as continuous temperature fluctuations and humidity cycles in electronic packaging, long-term exposure to ultraviolet radiation and rain erosion in outdoor coatings, and environmental stress coupling under mechanical load in composite materials. These factors will continuously act on the molecular structure of the curing agent, causing micro changes such as chemical bond rupture and crosslinking network destruction, and then leading to the degradation of the macro performance of the material, causing equipment failure or safety hazards. However, the existing analysis technology for weather resistance degradation of the curing agent mainly relies on natural aging test or accelerated aging test, and the weather resistance life of the material is inferred by performance detection under long-term exposure or intensified environmental conditions. The natural aging test period can be as long as several years or even decades, which is difficult to meet the needs of rapid research and development and application of new materials. Moreover, there is a lack of in-depth exploration of the correlation mechanism between the evolution of the micro molecular structure of the curing agent and environmental factors, especially the quantification of the molecular bond breaking behavior under the interaction of temperature, humidity, ultraviolet radiation and mechanical stress, which cannot fundamentally reveal the internal rules of weather resistance degradation. SUMMARY

[0003] Therefore, the present application provides a prediction method for weather resistance degradation of a curing agent based on finite element analysis to solve at least one of the above technical problems.

[0004] To achieve the above-mentioned purpose, a prediction method for weather resistance degradation of a curing agent based on finite element analysis comprises the following steps:

[0005] Step S1: Obtain curing agent material property data and curing agent application scenario environment data; perform material performance clustering processing on the differences in curing agent composition based on the curing agent material property data, and generate clustered curing agent material performance data;

[0006] Step S2: Perform finite element analysis processing on the reaction molecular structure of the curing agent based on the clustered curing agent material performance data and the curing agent application scenario environment data to obtain finite element curing agent reaction molecular structure data;

[0007] Step S3: Perform curing agent unit micro-bond breaking behavior analysis according to the finite element curing agent reaction molecular structure data to generate curing agent unit micro-bond breaking behavior data; perform environmental impact curing agent bond breaking behavior feature analysis based on the curing agent unit micro-bond breaking behavior data to generate environmental impact curing agent bond breaking behavior feature data;

[0008] Step S4: Establish a mapping relationship between the environmental stress tensor and the bond breaking evaluation of each cluster curing agent through the environmental impact curing agent bond breaking behavior feature data to obtain an environmental stress tensor-bond breaking evaluation model;

[0009] Step S5: Perform bond breaking and weathering degradation relationship mapping optimization of the environmental stress tensor-bond breaking evaluation model to obtain an environmental stress tensor-weathering degradation evaluation model;

[0010] Step S6: Perform unit environmental stress tensor feature analysis of the trend time-varying based on the finite element curing agent reaction molecular structure data according to the curing agent application scenario environment data to generate trend time-varying unit environmental stress tensor data; transmit the trend time-varying unit environmental stress tensor data to the environmental stress tensor-weathering degradation evaluation model to perform curing agent weathering performance degradation trend prediction to generate curing agent weathering performance degradation trend data.

[0011] Further, step S1 includes the following steps:

[0012] Step S11: Obtain curing agent material property data and curing agent application scenario environment data;

[0013] Step S12: Perform material composition analysis according to the curing agent material property data to obtain curing agent material composition data;

[0014] Step S13: Perform curing agent composition difference phase difference tensor analysis according to the curing agent material composition data to obtain curing agent composition difference phase difference tensor data;

[0015] Step S14: Perform curing agent material performance clustering processing according to the curing agent material composition data and the curing agent composition difference phase difference tensor data to generate clustering curing agent material performance data.

[0016] Further, step S2 includes the following steps:

[0017] Step S21: Perform curing agent reaction isomeric environment analysis according to the curing agent application scenario environment data to generate curing agent reaction isomeric environment data;

[0018] Step S22: Perform environmental field intensity gradient feature analysis on the curing agent reaction isomeric environment data to generate environmental field intensity gradient feature data;

[0019] Step S23: Design the environmental dynamic unit grid density according to the environmental field intensity gradient characteristic data to obtain environmental dynamic unit grid density data;

[0020] Step S24: Design the finite element non-structure grid nodes of the curing agent reaction through the environmental dynamic unit grid density data to obtain the finite element grid node data of the curing agent, and perform finite element analysis processing of the curing agent reaction by taking the clustering curing agent material performance data and the curing agent reaction isomerization environment data as the grid attribute nodes of the finite element grid node data of the curing agent to obtain the finite element curing agent reaction data;

[0021] Step S25: Perform molecular structure analysis of the finite element curing agent reaction data to obtain finite element curing agent reaction molecular structure data.

[0022] Further, step S3 includes the following steps:

[0023] Step S31: Perform curing agent unit microscopic bond breaking behavior analysis according to the finite element curing agent reaction molecular structure data to generate curing agent unit microscopic bond breaking behavior data;

[0024] Step S32: Perform curing agent unit isomerization environment and microscopic bond breaking behavior correlation processing on the curing agent unit microscopic bond breaking behavior data based on the curing agent reaction isomerization environment data to generate curing agent unit environment-bond breaking behavior data;

[0025] Step S33: Extract the curing agent reaction environmental impact factor of the curing agent unit environment-bond breaking behavior data; perform single environmental impact factor curing agent microscopic bond breaking behavior analysis on the curing agent unit environment-bond breaking behavior data according to the curing agent reaction environmental impact factor to generate single environmental impact factor-bond breaking behavior data;

[0026] Step S34: Perform environmental impact factor interaction effect curing agent microscopic bond breaking behavior analysis through the curing agent unit environment-bond breaking behavior data and the single environmental impact factor-bond breaking behavior data to generate interaction environmental impact factor-bond breaking behavior data;

[0027] Step S35: Perform environmental impact curing agent bond breaking behavior characteristic analysis based on the single environmental impact factor-bond breaking behavior data and the interaction environmental impact factor-bond breaking behavior data to generate environmental impact curing agent bond breaking behavior characteristic data.

[0028] Further, the single environmental impact factor-bond breaking behavior data of step S33 includes temperature conduction impact factor-bond breaking behavior data, humidity diffusion-bond breaking behavior data, ultraviolet absorption characteristic-bond breaking behavior data, and mechanical stress characteristic-bond breaking behavior data.

[0029] Further, step S4 comprises the following steps:

[0030] Step S41: Perform environmental influence curing agent bond breaking probability distribution analysis according to the environmental influence curing agent bond breaking behavior characteristic data, and generate environmental influence curing agent bond breaking probability distribution data;

[0031] Step S42: Perform environmental stress tensor bond breaking probability distribution data analysis according to the environmental influence curing agent bond breaking probability distribution data, and generate environmental stress tensor-bond breaking probability distribution data;

[0032] Step S43: Perform material performance difference bond breaking probability distribution classification processing on the environmental stress tensor-bond breaking probability distribution data according to the clustering curing agent material performance data, and generate classified environmental stress tensor-bond breaking probability distribution data;

[0033] Step S44: Establish a mapping relationship for environmental stress tensor and bond breaking evaluation of each clustering curing agent through the classified environmental stress tensor-bond breaking probability distribution data, to obtain an environmental stress tensor-bond breaking evaluation model.

[0034] Further, step S44 comprises the following steps:

[0035] A random forest algorithm is used to establish a tree structure mapping relationship between the environmental stress tensor and the bond breaking probability, to obtain an environmental stress tensor-bond breaking tree structure model, and the environmental stress tensor-bond breaking tree structure model is set with a bond breaking evaluation threshold parameter corresponding to the environmental stress tensor through the classified environmental stress tensor-bond breaking probability distribution data, to obtain an environmental stress tensor-bond breaking evaluation model.

[0036] Further, step S5 comprises the following steps:

[0037] Step S51: Perform curing agent bond breaking behavior node analysis according to the curing agent unit micro-bond breaking behavior data, and generate curing agent bond breaking behavior node data;

[0038] Step S52: Perform curing agent unit bond breaking and weathering degradation correlation analysis based on the curing agent bond breaking behavior node data, and generate curing agent unit bond breaking-weathering degradation correlation data;

[0039] Step S53: Perform curing agent bond breaking and weathering degradation mapping feature analysis according to the curing agent unit bond breaking-weathering degradation correlation data, and generate curing agent bond breaking-weathering degradation mapping feature data;

[0040] Step S54: Transmit the curing agent bond breaking-weathering degradation mapping feature data to the environmental stress tensor-bond breaking evaluation model to perform bond breaking and weathering degradation relationship mapping optimization of the curing agent, to obtain an environmental stress tensor-weathering degradation evaluation model.

[0041] Further, step S53 comprises the following steps:

[0042] Step S531: Perform the weathering degradation influence analysis of the broken bond node according to the broken bond unit-bond breaking-weathering degradation correlation data, generate the broken bond node-weathering degradation influence data, and perform the weathering degradation influence characteristic analysis of the broken bond central node on the broken bond node-weathering degradation influence data, generate the broken bond central node-weathering degradation influence characteristic data;

[0043] Step S532: Perform the unit weighting coefficient analysis of the broken bond-weathering degradation according to the broken bond unit-bond breaking-weathering degradation correlation data, generate the broken bond-weathering degradation unit volume weighting coefficient;

[0044] Step S533: Perform the curing agent broken bond-weathering degradation mapping characteristic analysis through the broken bond-weathering degradation unit volume weighting coefficient and the broken bond central node-weathering degradation influence characteristic data, generate the curing agent broken bond-weathering degradation mapping characteristic data.

[0045] Further, step S6 comprises the following steps:

[0046] Step S61: Perform the unit environment boundary condition analysis of the trend time-varying change of the finite element curing agent reaction molecular structure data based on the curing agent application scene environment data, generate the trend time-varying unit environment boundary condition data;

[0047] Step S62: Perform the unit environment stress tensor feature analysis of the trend time-varying change according to the trend time-varying unit environment boundary condition data, generate the trend time-varying unit environment stress tensor data;

[0048] Step S63: Transmit the trend time-varying unit environment stress tensor data to the environment stress tensor-weathering degradation evaluation model to perform the curing agent weathering performance degradation trend prediction, generate the curing agent weathering performance degradation trend data.

[0049] The application has the beneficial effects that the application lays a precise and efficient foundation for subsequent analysis through systematic processing of the curing agent material properties and application scene environment data. By obtaining material property data (such as the purity content, molecular weight distribution, etc. of methyl tetrahydrophthalic anhydride) and environmental data (such as temperature, humidity, ultraviolet radiation, external pressure, etc.) of the curing agent, the close fit of the analysis object and the actual application scene is ensured, and the prediction deviation caused by data loss is avoided. The combination of material composition analysis and phase difference tensor analysis can quantify the potential influence of the composition difference (such as impurity content, isomer ratio) of different curing agents on performance. Based on the clustering processing of composition data and phase difference tensor, curing agents with similar performance are classified, which not only reduces the calculation amount of subsequent finite element analysis, but also formulates differentiated prediction strategies for different clusters, significantly improving the analysis efficiency and pertinence, especially suitable for the diversified application needs of curing agents in electronic packaging, coatings and other scenes. By constructing a curing agent reaction molecular structure model through finite element analysis, precise mapping from the macro environment to the micro molecular level is realized, providing key support for revealing the weathering degradation mechanism. The heterogeneous environment analysis accurately captures the influence of environmental non-uniformity on molecular reaction in complex environments (such as local high temperature areas in electronic packaging and ultraviolet radiation differences in outdoor scenes), avoiding errors caused by the traditional uniform environment assumption. The combination of environmental field strength gradient feature analysis and dynamic grid density design ensures the calculation accuracy of key areas (such as areas with steep temperature gradients) while simplifying the grid in flat areas, balancing simulation accuracy and computational cost. In addition, the clustered material performance data and heterogeneous environment data are used as grid attributes, combined with potential energy surface scanning technology to analyze molecular structure, which can accurately reflect the micro characteristics such as molecular bond energy and bond length of different clustering curing agents in complex environments. Focusing on the correlation analysis of environmental factors and curing agent micro bond breaking behavior, the core mechanism of weathering performance degradation is revealed, providing key basis for subsequent model construction. Through unit micro bond breaking behavior analysis, the easy bond breaking sites in the curing agent molecule are accurately identified, and the "environment-bond breaking" correlation is established combined with heterogeneous environment data, which clarifies the bond breaking rules in different environmental regions. The combination of single environmental impact factor analysis and interaction effect analysis not only quantifies the bond breaking characteristics under the independent action of each factor, but also reveals the enhancement effect of multi-factor synergistic action (such as high temperature and high humidity accelerating bond breaking together), and the generated environmental impact bond breaking behavior characteristic data integrates the correlation mode of micro bond breaking rules and environmental factors. By establishing the mapping relationship between environmental stress tensor and bond breaking evaluation, the quantitative evaluation of the bond breaking behavior of different clustering curing agents is realized, providing a universal model basis for the weathering performance analysis of various curing agents (including methyl tetrahydrophthalic anhydride and other anhydride curing agents). Bond breaking probability distribution analysis converts the bond breaking behavior under the influence of the environment into quantifiable probability data, breaking through the limitations of qualitative description of curing agent bond breaking rules, making the bond breaking characteristics of different types of curing agents more statistically comparable.The correlation analysis of environmental stress tensor and bond breaking probability integrates the comprehensive effects of temperature, humidity, ultraviolet and other environmental factors, avoiding the one-sidedness of single factor analysis, and is suitable for curing agents and other curing agents affected by multiple factors in complex scenarios. Combined with clustering material performance for classification processing, the model can adapt to the characteristic differences of different component curing agents (such as different purity of methyl tetrahydrophthalic anhydride and composite curing agents containing different accelerators), improving the universality of the model for various curing agents. The application of random forest algorithm and the setting of threshold parameter not only enhances the fitting ability of the model to nonlinear mapping relationship, but also ensures the accuracy of bond breaking evaluation of different clustering curing agents through classification data calibration. By constructing the correlation mapping between bond breaking and weathering degradation, the microscopic bond breaking behavior is closely linked to the macroscopic performance degradation. The precise positioning of the bond breaking site that affects the performance by node analysis avoids the interference of irrelevant bond breaking information, making the correlation analysis more targeted. For example, it is clear that the bond breaking of crosslinking points of acid anhydride curing agents such as methyl tetrahydrophthalic anhydride has a significant impact on mechanical strength, or the N-H bond breaking in amine curing agents has an effect on corrosion resistance. The impact characteristic analysis of the central node of bond breaking and the calculation of unit weighting coefficient further quantify the influence weight and spatial distribution difference of different bond breaking nodes, making the mapping features more in line with the actual performance degradation law of various curing agents, such as the higher influence weight of bond breaking of methyl tetrahydrophthalic anhydride in high temperature region on overall weathering performance. The environmental stress tensor-weathering degradation evaluation model obtained by optimizing the mapping features realizes the direct prediction from environmental stress to macroscopic weathering degradation. By introducing the time-varying environmental stress analysis, the long-term weathering performance degradation trend of various curing agents is dynamically predicted. The trend time-varying unit environmental boundary condition analysis simulates the time sequence change of future environmental parameters based on actual application scenario data, making the prediction scene more close to the real use of various curing agents. The generation of time-varying environmental stress tensor converts the dynamic environmental change into a stress input that the model can recognize, capturing the cumulative effect of environmental factors over time, and avoiding the short-term prediction limitations of various curing agents caused by static environmental assumptions. Finally, the degradation trend data generated by inputting the time-varying stress tensor into the weathering degradation evaluation model can intuitively reflect the performance degradation law of various curing agents in long-term use, such as the tensile strength curve of methyl tetrahydrophthalic anhydride and the insulation performance degradation trend of epoxy curing agent, providing quantitative basis for engineering decisions such as selection of various curing agents, application scenario adaptation and maintenance cycle formulation, effectively solving the problem that traditional accelerated test is difficult to simulate long-term time sequence change.

[0050] Therefore, the curing agent weathering performance degradation prediction method based on finite element analysis can reduce the analysis object complexity through material performance clustering, improve the calculation efficiency by combining finite element analysis and dynamic grid design, greatly shorten the prediction period, and meet the needs of rapid research and application of new materials. And, the correlation between microstructure evolution and environmental factors is established, the molecular bond breaking behavior under the interaction of temperature, humidity, ultraviolet radiation, mechanical stress and other factors is accurately quantified, the internal law of weathering performance degradation is fundamentally revealed, and the accurate prediction from micro mechanism to macro performance degradation trend is realized. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 A step flowchart of the curing agent weathering performance degradation prediction method based on finite element analysis;

[0052] Figure 2 A detailed implementation step flowchart of step S4 in the curing agent weathering performance degradation prediction method based on finite element analysis; Figure 1

[0053] Figure 3 A detailed implementation step flowchart of step S5 in the curing agent weathering performance degradation prediction method based on finite element analysis; Figure 1

[0054] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0055] The technical method of the present application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0056] In addition, the accompanying drawings are only schematic illustrations of the present application, and are not necessarily drawn to scale. The same reference numerals in the drawings represent the same or similar parts, and thus repeated description thereof will be omitted. Some block diagrams shown in the drawings are functional entities, which do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0057] To achieve the above-mentioned object, please refer to Figures 1 to 3 ​​The application provides a method for predicting the weatherability degradation of a curing agent through finite element analysis. Figure 1 As shown in FIG. 1, which is a schematic diagram of the steps of the method for predicting the weatherability degradation of a curing agent through finite element analysis, the method comprises the following steps:

[0058] Step S1: Obtain curing agent material property data and curing agent application scenario environment data; perform material performance clustering processing on the curing agent component differences according to the curing agent material property data to generate clustered curing agent material performance data.

[0059] In the embodiment of the application, in the application scenario in the laboratory, a gas chromatograph-mass spectrometer is used to detect the curing agent sample (such as a methyltetrahydrophthalic anhydride sample) in the laboratory to obtain material property data including molecular weight distribution, acid value, purity, impurity type and content; at the same time, a temperature sensor, a humidity sensor, an ultraviolet radiation meter and a stress sensor arranged in the application scenario of the laboratory are used to continuously collect environment data, covering the daily temperature variation range, the relative humidity fluctuation interval, the ultraviolet radiation intensity and the mechanical stress value, which are used for finite element simulation test of the weatherability degradation of the curing agent in actual application. According to the material property data obtained by the gas chromatograph-mass spectrometer, a chemical metrology method is used to quantitatively analyze each component in the curing agent to determine the main component content, the isomer ratio and the impurity component proportion of the methyltetrahydrophthalic anhydride, and form the curing agent material component data. Based on the material component data, a tensor decomposition algorithm is used to calculate the phase difference tensor of the component differences between different samples, each element of the tensor corresponding to the difference degree of two samples in a specific component content, so as to obtain the curing agent component difference phase difference tensor data. The material component data and the phase difference tensor data are used as input variables, a K-means clustering algorithm is used, the component similarity and the phase difference tensor distance are used as clustering indexes, the samples with a component difference less than 5% and a phase difference tensor distance less than 0.1 are classified into the same class, and finally a plurality of sets of clustered curing agent material performance data are generated, each set of data containing common characteristic parameters such as the average molecular weight, the average acid value and the main impurity content of the curing agent.

[0060] Step S2: Perform finite element analysis processing on the curing agent reaction molecular structure based on the clustered curing agent material performance data and the curing agent application scenario environment data to obtain finite element curing agent reaction molecular structure data.

[0061] In the embodiment of the present application, according to the collected curing agent application scene environment data, the regions with temperature difference greater than 10℃, humidity difference greater than 20%, ultraviolet radiation intensity difference greater than 30%, and mechanical stress value difference greater than 15% are divided, which are defined as heterogeneous environment regions, the environmental parameter ranges of each region are summarized, and the curing agent reaction heterogeneous environment data is generated. Gradient calculation is performed on the temperature, humidity, ultraviolet radiation and mechanical stress parameters in the heterogeneous environment data, and the spatial interpolation method is used to obtain the change rate of each parameter at different positions, for example, the ratio of the temperature difference between adjacent two points to the distance is calculated, so as to determine the size and direction of the environmental field intensity gradient, and the environmental field intensity gradient characteristic data is formed. According to the environmental field intensity gradient characteristic data, in the region with gradient change rate greater than 5% / cm, the grid density of 0.1cm*0.1cm is adopted; in the region with gradient change rate between 2% / cm and 5% / cm, the grid density of 0.5cm*0.5cm is adopted; in the region with gradient change rate less than 2% / cm, the grid density of 1cm*1cm is adopted, so as to complete the design of the environmental dynamic unit grid density, and obtain the environmental dynamic unit grid density data. According to the above grid density data, the finite element unstructured grid of the curing agent reaction is constructed by using the Delaunay triangulation method, the spatial coordinates of each grid node are determined, and the curing agent finite element grid node data is formed; the molecular weight, acid value in the clustering curing agent material performance data and the temperature, humidity and other parameters in the heterogeneous environment data are assigned to the corresponding grid nodes as grid attributes, the balance equation of the grid nodes is established by using the displacement method in the finite element analysis, and the stress, strain and energy data of each node are solved to form the finite element curing agent reaction data. Based on the potential energy surface scanning technology, the energy calculation of the molecular structure in the finite element curing agent reaction data is performed in the molecular dynamics simulation software, the potential energy value corresponding to each conformation is recorded by changing the bond length and bond angle, the stable conformation with the lowest potential energy is found, and the spatial structure, bond length, bond angle and bond energy and other parameters of the molecule are determined, so as to obtain the finite element curing agent reaction molecular structure data.

[0062] Step S3: performing curing agent unit micro-bond breaking behavior analysis according to the finite element curing agent reaction molecular structure data to generate curing agent unit micro-bond breaking behavior data; performing curing agent bond breaking behavior characteristic analysis under the influence of the environment based on the curing agent unit micro-bond breaking behavior data to generate environmental influence curing agent bond breaking behavior characteristic data;

[0063] In the embodiment of the present application, according to the bond length, bond angle and bond energy parameters recorded in the finite element curing agent reaction molecular structure data, the breaking energy barrier of each chemical bond in the methyltetrahydrophthalic anhydride curing agent molecule is calculated by means of a molecular dynamics simulation device. When the external input energy exceeds the energy barrier, it is determined that the chemical bond breaks, and the specific position, time and corresponding bond energy change of the broken bond are recorded, thereby generating curing agent unit micro-bond breaking behavior data containing bond breaking frequency, average breaking time and other information. Then, the environment parameters of each region in the curing agent reaction isomerization environment data are matched one by one with the broken bond positions in the curing agent unit micro-bond breaking behavior data, and the bond breaking types and quantities in different environment regions are counted, such as the proportion of the number of C-O bond breakages in the high temperature region, to establish the corresponding relationship between the isomerization environment and the micro-bond breaking behavior, and generate the curing agent unit environment-bond breaking behavior data. From the data, temperature, humidity, ultraviolet radiation and mechanical stress are extracted as independent environmental impact factors, and other factors are kept unchanged, and only the value of a single factor is changed, for example, the humidity, ultraviolet radiation and mechanical stress are fixed, and only the temperature is gradually increased from room temperature to a specific high temperature, and the bond breaking behavior at each temperature is recorded, to obtain the temperature conduction impact factor-bond breaking behavior data. Using the same method, humidity diffusion-bond breaking behavior data, ultraviolet absorption characteristics-bond breaking behavior data and mechanical stress characteristics-bond breaking behavior data are obtained. Then, the curing agent unit environment-bond breaking behavior data and the single environmental impact factor-bond breaking behavior data are compared, the difference between the bond breaking rate when multiple factors act together and the sum of the bond breaking rates when each single factor acts alone is calculated, the interactive effect of the environmental impact factor is quantified, and the interactive environmental impact factor-bond breaking behavior data is generated. Finally, the above two types of data are integrated, the bond breaking probability, bond breaking rate and bond breaking position distribution law of each chemical bond under different environmental conditions are counted, and the environmental impact curing agent bond breaking behavior characteristic data is generated.

[0064] Step S4: mapping relationship establishment of environmental stress tensor and bond breaking evaluation of each cluster curing agent is carried out through the environmental impact curing agent bond breaking behavior characteristic data to obtain the environmental stress tensor-bond breaking evaluation model;

[0065] In the embodiment of the present application, according to the environmental influence on the breaking behavior characteristics data of the curing agent, the probability and statistics method is used to statistically analyze the breaking events under different environmental conditions, to calculate the probability of breaking in each environmental combination, and to form the environmental influence on the breaking probability distribution data of the curing agent, which contains the probability density function of different breaking types. The temperature, humidity, ultraviolet radiation, mechanical stress and other environmental parameters are integrated into the environmental stress tensor, each component of the tensor corresponds to the intensity of an environmental factor, and the mathematical relationship between the components of the environmental stress tensor and the breaking probability is established by multiple regression analysis to generate the environmental stress tensor-breaking probability distribution data. Combined with the clustering curing agent material performance data, the environmental stress tensor-breaking probability distribution data is grouped according to the clustering categories, so that each group of data corresponds to the breaking probability characteristics of a type of curing agent, and the classified environmental stress tensor-breaking probability distribution data is formed. Using the random forest algorithm, the classified environmental stress tensor is used as the input variable and the breaking probability is used as the output variable to construct multiple decision trees, each decision tree is trained and generated based on different sample subsets and feature subsets, the final output result is determined by the majority voting principle, and the environmental stress tensor-breaking tree structure model is formed. According to the classified environmental stress tensor-breaking probability distribution data, the breaking threshold of different clustering curing agents under a specific environmental stress tensor is set, and when the breaking probability exceeds the threshold, it is determined as significant breaking, so as to complete the threshold parameter setting of the model, and the environmental stress tensor-breaking evaluation model is obtained.

[0066] Step S5: mapping and optimizing the breaking and weathering degradation relationship of the curing agent in the environmental stress tensor-breaking evaluation model to obtain the environmental stress tensor-weathering degradation evaluation model;

[0067] In the embodiment of the present application, according to the micro bond breaking behavior data of the curing agent unit, the specific molecular position of the bond breaking is identified, the bond breaking sites which have a significant impact on the overall performance of the curing agent are marked, such as the key connection points in the crosslinked network, and the curing agent bond breaking behavior node data is generated. The performance of the curing agent samples with different bond breaking degrees is detected by the mechanical property testing equipment, and the corresponding relationship between the bond breaking rate and the tensile strength, hardness, insulation resistance and other weather resistance performance parameters is obtained, and the curing agent unit bond breaking-weathering degradation correlation data is established. Based on the curing agent unit bond breaking-weathering degradation correlation data, the sensitivity analysis method is used to calculate the influence weight of each bond breaking node on the weathering performance degradation, and the top 20% of the bond breaking nodes with the highest influence weight are selected as the bond breaking hub nodes. The influence law of the bond breaking hub nodes on the weathering performance under different bond breaking degrees is analyzed, and the bond breaking hub node-weathering degradation influence characteristic data is generated. At the same time, according to the spatial distribution of the curing agent unit, the volume proportion of the units at different positions is calculated, and the bond breaking-weathering degradation unit volume weighting coefficient is determined by combining the influence degree of the bond breaking on the overall performance. The coefficient is positively correlated with the unit volume proportion and the influence weight. The bond breaking-weathering degradation unit volume weighting coefficient and the bond breaking hub node-weathering degradation influence characteristic data are fused, and the quantitative mapping relationship between the bond breaking behavior and the weathering performance degradation is established by weighted summation, and the curing agent bond breaking-weathering degradation mapping characteristic data is generated. The data is substituted into the environmental stress tensor-bond breaking evaluation model to correct the output parameters of the model, so that the model can directly output the weathering performance degradation index, and the environmental stress tensor-weathering degradation evaluation model is obtained.

[0068] Step S6: Based on the curing agent application scene environment data, the unit environmental stress tensor characteristic analysis of the trend time-varying of the finite element curing agent reaction molecular structure data is performed, and the trend time-varying unit environmental stress tensor data is generated. The trend time-varying unit environmental stress tensor data is transmitted to the environmental stress tensor-weathering degradation evaluation model to predict the weathering performance degradation trend of the curing agent, and the curing agent weathering performance degradation trend data is generated.

[0069] In the embodiment of the present application, based on the curing agent application scene environment data, the time series analysis method is adopted to perform trend fitting on the past long-time environmental parameters such as temperature, humidity, ultraviolet radiation, and mechanical stress, to predict the future environmental parameter change rule, to determine the environmental parameter range of each time node, to generate trend time-varying unit environmental boundary condition data, and the data contains the maximum value, minimum value and average value of each environmental factor at different time. According to the trend time-varying unit environmental boundary condition data, combined with the unit space distribution in the finite element curing agent reaction molecular structure data, the stress calculation method in the finite element analysis is adopted to calculate the environmental stress tensor borne by each unit at each time node, and each component of the tensor dynamically changes with time, so as to generate trend time-varying unit environmental stress tensor data, and the data contains the stress tensor value of each unit at different time. The trend time-varying unit environmental stress tensor data is input into the environmental stress tensor-weathering degradation evaluation model, and the model calculates the weathering performance parameters of the curing agent at each time node according to the preset mapping relationship, such as the tensile strength retention rate and the insulation resistance decay rate. By continuously calculating the performance parameters at different time, the curve of the change of the weathering performance with time is obtained, and the weathering performance degradation trend data of the curing agent is generated, which can directly reflect the performance degradation rule of the curing agent in the future use process.

[0070] Further, step S1 comprises the following steps:

[0071] Step S11: Obtain curing agent material property data and curing agent application scene environment data;

[0072] In the embodiment of the present application, in the application scene of the laboratory, for the analyzed curing agent (such as methyl tetrahydrophthalic anhydride), a gas chromatograph-mass spectrometer is used for material property detection, each component in the sample is separated by a chromatographic column, and the molecular mass and structure of each component are measured by a mass spectrometer, to obtain material property data including main component content, isomer ratio, impurity type and content, and acid value; meanwhile, in the curing agent application scene of the laboratory, temperature sensors, humidity sensors, ultraviolet radiation meters and stress sensors are arranged at a specific interval, data is collected continuously for 30 days, and temperature fluctuation range, relative humidity change interval, ultraviolet radiation intensity peak value and mechanical stress average value are recorded daily to form curing agent application scene environment data. All detection instruments are calibrated, and the sensor acquisition frequency is set to once per hour to ensure the accuracy and continuity of the data.

[0073] Step S12: Perform material component analysis according to the curing agent material property data to obtain curing agent material component data;

[0074] In the embodiment of the present application, according to the obtained curing agent material characteristic data, the curing agent sample is analyzed by using an infrared spectrometer, the type and quantity of functional groups in the molecule are determined through the position and intensity of the characteristic absorption peak, and the components are quantitatively calculated in combination with the component separation results obtained by the gas chromatograph-mass spectrometer. For example, by comparing the characteristic peak area of the standard sample, the content of the main component methyl tetrahydrophthalic anhydride of the curing agent is calculated, and the proportion of the isomers (such as cis and trans structures) is determined, and the impurities such as phthalic anhydride residues and water are identified and the proportion is determined, and finally the curing agent material component data containing the component name, chemical structure and mass fraction is formed, and the quantitative result of each component is accurate to two decimal places.

[0075] Step S13: performing phase difference tensor analysis of curing agent component difference according to the curing agent material component data to obtain curing agent component difference phase difference tensor data;

[0076] In the embodiment of the present application, based on the curing agent material component data, 10 curing agent samples of different batches are selected, and the component mass fraction of each sample is used as the basis data to construct a component vector. Using tensor decomposition algorithm, the component vector of each sample is converted into a three-dimensional tensor, wherein the first dimension represents the component type, the second dimension represents the sample number, and the third dimension represents the mass fraction. The phase difference between any two sample tensors is calculated, and each element of the phase difference tensor is solved by matrix operation, and the element value reflects the difference degree of the two samples in a certain component. For example, the difference between the curing agent sample A and the curing agent sample B in the main component of methyl tetrahydrophthalic anhydride is reflected by the numerical value in the corresponding position of the tensor, and finally the curing agent component difference phase difference tensor data containing the component difference between all samples is formed.

[0077] Step S14: performing curing agent material performance clustering processing according to the curing agent material component data and the curing agent component difference phase difference tensor data to generate clustered curing agent material performance data.

[0078] In the embodiment of the present application, based on the curing agent material composition data and the composition difference phase difference tensor data (reflecting the composition difference degree between samples), the K-means clustering algorithm is used for processing. First, the number of clusters is determined, and each curing agent sample is converted into a multi-dimensional vector containing composition features and phase difference features, wherein the composition features are taken from the mass fraction of each component in the material composition data, and the phase difference features are taken from the element values in the phase difference tensor data. Calculate the Euclidean distance between any two sample vectors, the smaller the distance value, the closer the sample performance. Randomly select multiple samples as cluster centers, assign each sample to the nearest cluster center category, and then recalculate the new center of each category (i.e. the average of all sample vectors in this category). Repeat the above assignment and update process until the variation of the cluster center in the last three iterations is less than 0.001, and stop iteration. Finally, a plurality of sets of clustering data are obtained, each set of data containing the average composition mass fraction, the average phase difference tensor, the common functional group feature and the sample quantity of the samples in the category. Summarize each set of data to clarify the commonality of the same type of curing agent in terms of composition and performance difference, generate clustering curing agent material performance data, ensure that the composition similarity between the same type of samples is not less than 90%, and the phase difference tensor distance is not more than 0.05, providing classification basis for subsequent targeted modeling of finite element analysis.

[0079] Further, step S2 comprises the following steps:

[0080] Step S21: analyzing the heterogeneous environment of curing agent reaction according to the curing agent application scene environment data to generate curing agent reaction heterogeneous environment data;

[0081] In the embodiment of the present application, according to the curing agent application scene environment data, the application scene of the curing agent in the electronic packaging is divided into environment monitoring areas. The area where the temperature daily fluctuation exceeds 15℃ is marked as a temperature heterogeneous area, the area where the relative humidity fluctuation exceeds 30% is marked as a humidity heterogeneous area, the area where the ultraviolet radiation intensity difference exceeds 50% is marked as a radiation heterogeneous area, and the area where the mechanical stress value changes more than 20% is marked as a stress heterogeneous area. The spatial coordinates of each heterogeneous area are located, and the maximum value, minimum value and change period of the environmental parameters in each area are recorded, for example, the temperature range of the area near the chip and the edge area of the packaging in the temperature heterogeneous area. The spatial distribution and parameter characteristics of all heterogeneous areas are summarized to generate curing agent reaction heterogeneous environment data, which contains the boundary coordinates, dominant environmental factors and parameter fluctuation range of each heterogeneous area, ensuring that the division accuracy of each area is controlled within ±0.5 centimeters.

[0082] Step S22: analyzing the environmental field intensity gradient characteristics of the curing agent reaction heterogeneous environment data to generate environmental field intensity gradient characteristic data;

[0083] In the embodiment of the present application, for the curing agent reaction isomerism environment data, the environmental parameter monitoring points in each isomerism region are selected, and the spacing between adjacent monitoring points is set to 1 centimeter. The numerical difference method is used to calculate the environmental field strength gradient. For the temperature field, the temperature gradient value is obtained by calculating the ratio of the temperature difference between two adjacent points to the spatial distance. For the humidity field, the humidity gradient is calculated by the ratio of the relative humidity difference to the distance. For the ultraviolet radiation field and the mechanical stress field, the same method is used to calculate the radiation gradient and the stress gradient respectively. The gradient size and direction at each monitoring point are recorded, and the points with the same gradient value are connected by the contour drawing method to form the gradient field distribution map. The gradient change is marked as a high gradient area in the region (gradient value is more than 5 units / cm), and the gradient change is marked as a low gradient area in the region (gradient value is less than 2 units / cm), and the environmental field strength gradient characteristic data including gradient value, direction and region division are generated, and the calculation of all gradient values is accurate to two decimal places.

[0084] Step S23: designing the environmental dynamic unit grid density according to the environmental field strength gradient characteristic data to obtain the environmental dynamic unit grid density data;

[0085] In the embodiment of the present application, according to the environmental field strength gradient characteristic data, the grid density division standard is set: the high gradient area (gradient value is more than 5 units / cm) uses 0.2 cm*0.2 cm grid unit size to capture the subtle gradient change; the medium gradient area (gradient value is between 2-5 units / cm) uses 0.5 cm*0.5 cm grid unit size; the low gradient area (gradient value is less than 2 units / cm) uses 1 cm*1 cm grid unit size. According to the boundary coordinates of each region, transition grids are set at the junction of the high gradient area and the medium gradient area, and the medium gradient area and the low gradient area. The size of the transition grid changes gradually in a linear proportion to avoid calculation errors caused by sudden changes in grid size. The grid density is associated with the coordinate range of the corresponding region to form the environmental dynamic unit grid density data. The grid unit size in each coordinate interval is clearly defined in the data to ensure that the grid covers all isomerism regions without overlapping.

[0086] Step S24: designing the finite element unstructured grid nodes of the curing agent reaction through the environmental dynamic unit grid density data to obtain the curing agent finite element grid node data, and performing finite element analysis processing on the curing agent reaction by taking the clustering curing agent material performance data and the curing agent reaction isomerism environment data as the grid attribute nodes of the curing agent finite element grid node data to obtain the finite element curing agent reaction data;

[0087] In the embodiment of the present application, according to the dynamic unit grid density data of the environment, the Delaunay triangulation method is used for finite element non-structural grid node design. First, the boundary contour of the curing agent reaction region is determined, and the discrete points on the contour are used as the initial boundary nodes. The spacing between adjacent boundary nodes is determined according to the grid density of the region. The spacing between adjacent boundary nodes in the high gradient region is 0.1 cm, the spacing between adjacent boundary nodes in the medium gradient region is 0.25 cm, and the spacing between adjacent boundary nodes in the low gradient region is 0.5 cm. The internal nodes are generated according to the grid density requirements in the region. The internal nodes in the high gradient region are uniformly distributed at an interval of 0.2 cm x 0.2 cm, the internal nodes in the medium gradient region are distributed at an interval of 0.5 cm x 0.5 cm, and the internal nodes in the low gradient region are distributed at an interval of 1 cm x 1 cm. All nodes are connected to form triangular elements, and it is ensured that the internal angle of each element is between 30 degrees and 120 degrees to avoid the occurrence of abnormal elements. The three-dimensional coordinates of each node (accurate to 0.01 mm) are recorded to form the curing agent finite element grid node data, which includes node number and corresponding coordinate information. The clustering curing agent material performance data is assigned to the grid nodes as material properties. The average molecular weight, acid value and other parameters are assigned according to the clustering category to which the node belongs, and the functional group density is distributed to each node by interpolation method. The curing agent reaction isomeric environment data is assigned to the nodes as environmental properties. The temperature, humidity and other parameters are determined according to the parameter range of the isomeric region where the node is located, and the parameter of the node at the junction of the region is calculated by linear transition. The displacement method in finite element analysis is used for reaction analysis, and the balance equation of each node is established. The equation includes the stiffness coefficient corresponding to the material property and the load value corresponding to the environmental property. The equation set is solved by Gaussian elimination method to obtain the displacement of each node, and then the strain is calculated according to the geometric equation, the stress is obtained through the physical equation, and the energy value (including elastic potential energy and thermal energy) at the node is calculated to generate the finite element curing agent reaction data.

[0088] Step S25: performing molecular structure analysis on the finite element curing agent reaction data to obtain finite element curing agent reaction molecular structure data.

[0089] In the embodiment of the present application, the potential energy surface scanning technology is used to analyze the molecular structure for the finite element curing agent reaction data. The key chemical bonds in the curing agent molecule (such as the C-O bond and C-C bond of methyl tetrahydrophthalic anhydride) are selected, the bond length change step is set to 0.01 angstrom, the bond angle change step is set to 1 degree, and the total potential energy of the molecule is calculated at each step. The potential energy values in different conformations are solved by quantum chemical calculation method, and the three-dimensional surface of the potential energy changing with the bond length and bond angle is drawn, and the lowest point in the surface corresponds to the stable conformation of the molecule. The bond length, bond angle, dihedral angle and bond energy data in the stable conformation are recorded, and the correlation between the stress distribution in the finite element reaction data and the molecular conformation is analyzed, for example, the molecular bond length change corresponding to the high stress area. The molecular structure parameters of all grid nodes are summarized to generate finite element curing agent reaction molecular structure data, which includes the three-dimensional coordinates of the molecule, the bond parameters and the potential energy value at each node.

[0090] Further, step S3 comprises the following steps:

[0091] Step S31: analyzing the curing agent unit micro-bond breaking behavior according to the finite element curing agent reaction molecular structure data to generate curing agent unit micro-bond breaking behavior data;

[0092] In the embodiment of the present application, according to the finite element curing agent reaction molecular structure data, for example, focusing on the easy breaking bond sites such as C-O bond and C-C bond in methyl tetrahydrophthalic anhydride molecule, the molecular mechanics method is used to calculate the breaking energy barrier of each chemical bond. By analyzing the bond length, bond angle and bond energy parameters in the molecular structure data, the breaking criterion is set: when the bond length exceeds 15% of the initial bond length, or the bond energy is lower than 30% of the initial bond energy, it is determined that the bond breaking occurs. The molecular structure of each curing agent unit is dynamically monitored, and the specific position of the bond breaking (such as the first C-O bond on the molecular chain), the bond breaking time (in the time step of finite element analysis) and the bond energy change value before and after the bond breaking are recorded. The bond breaking information of all units is summarized, the bond breaking frequency (the number of bond breakings per unit time), the average breaking time and the molecular conformation parameters at the time of bond breaking of different chemical bonds are counted, and the curing agent unit micro-bond breaking behavior data is generated.

[0093] Step S32: correlating the curing agent unit micro-bond breaking behavior data with the isomerization environment and micro-bond breaking behavior of the curing agent unit based on the curing agent reaction isomerization environment data to generate curing agent unit environment-bond breaking behavior data;

[0094] In the embodiments of the present application, based on the curing agent reaction isomerism environment data, the boundary coordinates of each isomerism region are matched with the broken bond position coordinates in the obtained broken bond behavior data of the curing agent unit microcosm, to determine the isomerism environment region to which each broken bond event belongs. For each isomerism region, the broken bond type (such as C-O bond broken bond proportion) occurring therein, the broken bond number and the broken bond rate (total number of broken bonds per unit time) are counted, and the environment parameters (such as temperature range, humidity value) of the region are associated. For example, in the high-temperature sub-region of the temperature isomerism region, the corresponding relationship between the number of C-O bond broken bonds and the maximum temperature of the region is recorded; in the high-humidity sub-region of the humidity isomerism region, the association between the broken bond rate of C-C bond and the relative humidity is recorded. The associated information is classified and arranged according to the region, to generate the curing agent unit environment-broken bond behavior data, which contains the region number, the environment parameter range, the broken bond type and the corresponding statistical value, to ensure that each broken bond event can be accurately matched to a unique isomerism environment region.

[0095] Step S33: extracting the curing agent reaction environment influence factor of the curing agent unit environment-broken bond behavior data; performing single environment influence factor microcosm broken bond behavior analysis on the curing agent unit environment-broken bond behavior data according to the curing agent reaction environment influence factor, to generate single environment influence factor-broken bond behavior data;

[0096] In the embodiment of the present application, from the curing agent unit environment-bond breaking behavior data, four independent variables of temperature, humidity, ultraviolet radiation intensity and mechanical stress are separated and determined as the curing agent reaction environment influencing factors. For each factor, the single environment influencing factor micro-bond breaking behavior analysis is carried out by using the control variable method. Taking the temperature conduction influencing factor analysis as an example, the humidity is fixed as the average relative humidity in the environmental data, the ultraviolet radiation intensity is set as the average radiation value monitored, the mechanical stress is maintained at the average stress level, and only the temperature parameter is changed. The temperature adjustment range covers the lowest temperature to the highest temperature recorded in the application scenario, and each 10 DEG C is set as a gradient, and the stable time length of each gradient is the average duration of the temperature interval in the environmental data. Under each temperature gradient, the bond length change of each chemical bond in the curing agent unit is tracked through molecular dynamics simulation, and when the bond length exceeds 12% of the initial bond length, the bond breaking event is recorded, the total number of bond breaking in unit time, the bond breaking proportion of different chemical bonds and the position distribution of bond breaking are counted, and the temperature conduction influencing factor-bond breaking behavior data is formed. For the humidity diffusion influencing factor analysis, the temperature is fixed as the average temperature, the ultraviolet radiation and the mechanical stress are maintained at the average value, only the relative humidity is changed, each 20% relative humidity is set as a gradient, the stable time length of each gradient is consistent with the average duration of the corresponding humidity interval, the bond breaking behavior is monitored and counted, and the humidity diffusion-bond breaking behavior data is generated. In the ultraviolet absorption characteristic influencing factor analysis, the temperature, humidity and mechanical stress are fixed as the average value, each 20 W / m2 of ultraviolet radiation intensity is set as a gradient, and the stable time length of each gradient matches the average duration of the corresponding radiation intensity. The bond breaking data is counted to generate the ultraviolet absorption characteristic-bond breaking behavior data. In the mechanical stress characteristic influencing factor analysis, the temperature, humidity and ultraviolet radiation are fixed as the average value, each 10 MPa of mechanical stress is set as a gradient, and the stable time length of each gradient matches the average duration of the corresponding stress level. The bond breaking data is counted to generate the mechanical stress characteristic-bond breaking behavior data. In all analysis processes, the bond breaking statistics under each gradient is repeated five times, and the average value of the five results is taken as the final data under the gradient to ensure the reliability of the data.

[0097] Step S34: The curing agent micro-bond breaking behavior analysis of the interaction effect of the environmental influencing factors is carried out through the curing agent unit environment-bond breaking behavior data and the single environment influencing factor-bond breaking behavior data, and the interactive environmental influencing factor-bond breaking behavior data is generated.

[0098] In the embodiments of the present application, the temperature and humidity, temperature and ultraviolet radiation, and humidity and mechanical stress are selected as the three typical interactive combinations for analysis by combining the environmental-bond breaking behavior data of the curing agent unit and the single environmental impact factor-bond breaking behavior data. Taking the interactive effect of temperature and humidity as an example, the temperature is set at an interval of 5°C and the humidity is set at an interval of 10% RH to form a grid-shaped combination parameter, covering the extreme value range of both. At each combination parameter, the difference between the actual bond breaking rate (taken from the environmental-bond breaking behavior data) and the sum of the bond breaking rate under the action of temperature alone and the bond breaking rate under the action of humidity alone is calculated, and the difference is the interactive effect value. If the difference is positive, it indicates that there is a synergistic enhancement effect; if it is negative, it indicates that there is an antagonistic weakening effect. The interactive effect value under different combination parameters, the corresponding bond breaking type and the bond breaking position distribution are recorded, and the temperature-humidity interactive environmental impact factor-bond breaking behavior data is generated. The same logic is used to analyze the other two interactive combinations, and the corresponding interactive environmental impact factor-bond breaking behavior data is generated, and the calculation of all interactive effect values is accurate to three decimal places.

[0099] Step S35: Perform curing agent bond breaking behavior characteristic analysis of environmental impact based on the single environmental impact factor-bond breaking behavior data and the interactive environmental impact factor-bond breaking behavior data, and generate environmental impact curing agent bond breaking behavior characteristic data.

[0100] In the embodiment of the present application, the single environmental impact factor-bond breaking behavior data and the interactive environmental impact factor-bond breaking behavior data are integrated, and variance analysis and regression analysis are used to analyze the characteristics of the bond breaking behavior of the curing agent under the environmental impact. First, for the single environmental impact factor, the average change amplitude of the bond breaking probability is calculated when each factor (temperature, humidity, ultraviolet radiation, mechanical stress) changes by one standard unit within its value range. For example, when the temperature increases by 10°C, the increase percentage of the C-O bond breaking probability is calculated, and when the humidity increases by 15% RH, the change value of the C-C bond breaking probability is calculated. The influence weight of each factor on different chemical bonds is determined by calculating the absolute value of these change amplitudes, and the weight value is expressed as a percentage and the sum is 100%. For the interactive environmental impact factor, the interactive effect intensity of the combination of temperature and humidity, temperature and ultraviolet radiation, humidity and mechanical stress, etc. in different parameter intervals is analyzed. The difference between the actual bond breaking rate under the interaction and the sum of the bond breaking rates of each single factor is calculated. When the difference is positive, it is defined as a synergistic effect, and when it is negative, it is defined as an antagonistic effect. The parameter interval with an effect intensity of more than 20% (for example, when the temperature is 60-80°C and the humidity is 70-90% RH, the synergistic effect makes the bond breaking rate increase by 35%) is recorded. At the same time, the high-frequency area of bond breaking under each environmental condition is determined by spatial distribution statistics, and the top 10% units with the highest bond breaking frequency are marked as sensitive areas. The structure parameters such as bond length and bond angle in these areas are measured to determine the molecular structure characteristics of the sensitive areas. The single factor influence weight, interactive effect intensity and parameter interval, sensitive area structure characteristics, etc. are integrated and classified according to the type of chemical bond. For example, the temperature-humidity synergistic effect significant interval of C-O bond and the corresponding sensitive area bond length range, and the ultraviolet-stress synergistic effect interval of C-C bond, etc. All data are verified by three repeated calculations to ensure that the error is controlled within 5%. Finally, the environmental impact curing agent bond breaking behavior characteristic data including the environmental sensitivity, interaction rule and spatial distribution characteristics of each chemical bond breaking are formed.

[0101] Further, the single environmental impact factor-bond breaking behavior data in step S33 includes temperature conduction impact factor-bond breaking behavior data, humidity diffusion-bond breaking behavior data, ultraviolet absorption characteristic-bond breaking behavior data, and mechanical stress characteristic-bond breaking behavior data.

[0102] Further, as an embodiment of the present application, referring to Figure 2 , it is Figure 1 the detailed step flowchart of step S4 in the embodiment. In the embodiment, step S4 includes the following steps:

[0103] Step S41: performing environmental impact curing agent bond breaking probability distribution analysis according to the environmental impact curing agent bond breaking behavior characteristic data to generate environmental impact curing agent bond breaking probability distribution data;

[0104] In the embodiments of the present application, according to the environmental influence on the breaking behavior characteristics data of the curing agent, a probability statistical method is used to develop the breaking probability distribution analysis. The environmental influence factors are divided into several continuous intervals at fixed intervals, and each interval contains a specific range combination of temperature, humidity, ultraviolet radiation and mechanical stress. The breaking events in each interval are counted, the ratio of the number of breaking events to the total number of observations is calculated, and the breaking frequency of the interval is obtained. The breaking frequency data is fitted as a continuous distribution function using probability density function fitting technology, and the function parameters are adjusted by the least square method to control the deviation of the fitting curve and the actual data within 4%. For each type of chemical bond, a corresponding probability distribution function is established, and the function type, parameters and applicable environmental interval range are recorded. The distribution functions of all intervals and the associated environmental parameters are summarized to generate the environmental influence on the breaking probability distribution data of the curing agent.

[0105] Step S42: According to the environmental influence on the breaking probability distribution data of the curing agent, the breaking probability distribution data analysis of the environmental stress tensor is carried out, and the environmental stress tensor-breaking probability distribution data is generated;

[0106] In the embodiments of the present application, based on the environmental influence on the breaking probability distribution data of the curing agent, the environmental stress tensor is constructed. The environmental stress tensor is a four-dimensional vector, and each component corresponds to the standardized value of temperature, humidity, ultraviolet radiation and mechanical stress. The standardized process is obtained by calculating the ratio of the actual value of the parameter to the maximum value, so that the numerical range of each component is controlled between 0 and 1. The tensor decomposition method is used to convert the environmental parameter combination into the characteristic expression of the stress tensor, and the weight coefficient of each component in the tensor is determined. The breaking probability distribution data of each environmental interval is calculated with the corresponding environmental stress tensor, and the mathematical relationship between the stress tensor components and the breaking probability distribution parameters is established by multiple regression analysis. The calculation process retains four significant digits after the decimal point. The probability distribution rules of different chemical bonds under different stress tensor characteristics are recorded, including the distribution function type, parameters and corresponding tensor component range, and the environmental stress tensor-breaking probability distribution data is generated.

[0107] Step S43: According to the clustering of the curing agent material performance data, the breaking probability distribution classification processing of the material performance difference is carried out on the environmental stress tensor-breaking probability distribution data, and the classified environmental stress tensor-breaking probability distribution data is generated;

[0108] In the embodiment of the present application, according to the clustering curing agent material performance data, the environmental stress tensor-bond breaking probability distribution data obtained in step S42 is classified according to the clustering categories. Each clustering category corresponds to a group of curing agents with similar material characteristics. When classifying, according to the material composition similarity and performance parameter difference, samples with a material composition similarity of more than 90% and a performance parameter difference of less than 8% are classified into the same category. The statistical characteristics of the bond breaking probability distribution are recalculated for each category of data, including the probability mean, standard deviation and distribution range under the same stress tensor. The distribution characteristic differences between different clusters are compared, and the significant stress tensor interval and the corresponding probability parameters are marked. The classified data set is arranged according to the clustering number, and each data set contains the bond breaking probability distribution details of the curing agent in different stress tensors of this category, and the classified environmental stress tensor-bond breaking probability distribution data is generated.

[0109] Step S44: mapping relationship establishment of environmental stress tensor and bond breaking evaluation of each clustering curing agent through classified environmental stress tensor-bond breaking probability distribution data, to obtain an environmental stress tensor-bond breaking evaluation model.

[0110] In the embodiment of the present application, for the classified environmental stress tensor-bond breaking probability distribution data, a random forest algorithm is used to establish the mapping relationship between the environmental stress tensor and the bond breaking evaluation of each clustering curing agent. For each clustering curing agent type, the four components of the environmental stress tensor are used as input variables, and the bond breaking probability is used as output variable, and a forest model composed of 100 decision trees is constructed. The training samples of each decision tree are randomly selected from 70% of the clustering data, and 3 components are randomly selected as feature variables. By calculating the prediction results of each tree, the majority voting method is used to determine the final bond breaking probability prediction value. After the model training is completed, according to the bond breaking probability distribution data of this clustering, the bond breaking evaluation threshold is set: when the predicted bond breaking probability exceeds the upper limit value of the 90% confidence interval of this clustering, it is determined as “significant bond breaking”; between the upper limit of the 50% confidence interval and the upper limit of the 90% confidence interval, it is determined as “slight bond breaking”; lower than the upper limit of the 50% confidence interval, it is determined as “no significant bond breaking”. The threshold parameter is embedded in the model to form an environmental stress tensor-bond breaking evaluation model for this clustering. Repeat the above process for all clusters to generate multiple models, each model including forest structure parameters, threshold determination criteria and applicable clustering range.

[0111] Further, step S44 includes the following steps:

[0112] The preset random forest algorithm is used to establish the mapping relationship between the environmental stress tensor and the bond breaking probability of various tree structures, to obtain the environmental stress tensor-bond breaking tree structure model, and the threshold parameter setting of the environmental stress tensor corresponding to the bond breaking is performed on the environmental stress tensor-bond breaking probability distribution data through classification, to obtain the environmental stress tensor-bond breaking evaluation model.

[0113] In the embodiment of the present application, for the classification of environmental stress tensor-bond breaking probability distribution data, the random forest algorithm is used to construct the tree structure mapping relationship between the environmental stress tensor and the bond breaking probability. First, the algorithm parameters are determined: the input of each decision tree is the four components (standardized values of temperature, humidity, ultraviolet radiation and mechanical stress) of the environmental stress tensor, and the output is the bond breaking probability; the tree structure is constructed by using a top-down recursive splitting method, each internal node is determined according to the node impurity minimization principle from the three components selected at random, and the optimal splitting feature and splitting threshold are determined; the maximum depth of the tree is limited to 12 layers, and the minimum sample size contained by the leaf node is set to 6, to avoid overfitting. For each cluster data set, divide it into training set and validation set according to the ratio of 7:3, repeatedly construct 150 decision trees based on the training set, and use different bootstrap samples for each tree. Finally, the average value of the prediction results of all trees is taken to form the environmental stress tensor-bond breaking tree structure model. Then, the threshold parameter of bond breaking evaluation is set by using the classification of environmental stress tensor-bond breaking probability distribution data. For each curing agent type corresponding to the cluster, the distribution characteristics of all bond breaking probability data in this category are counted, and the probability values corresponding to the cumulative distribution functions reaching 90%, 70% and 50% are calculated, respectively, as the thresholds of "severe bond breaking", "slight bond breaking" and "no obvious bond breaking". These thresholds are embedded into the tree structure model, and when the bond breaking probability output by the model exceeds the "severe bond breaking" threshold, it is marked as high-risk bond breaking; when it is between the "slight bond breaking" and "severe bond breaking" thresholds, it is marked as medium-risk bond breaking; and when it is lower than the "no obvious bond breaking" threshold, it is marked as low-risk bond breaking. The accuracy of the threshold division is verified by the validation set data to ensure that the three risk labels are consistent with the actual bond breaking situation, and finally the environmental stress tensor-bond breaking evaluation model is formed. The model includes tree structure parameters, threshold standards and risk labeling rules.

[0114] Further, as an embodiment of the present application, referring to FIG. 5, it is a detailed step flow diagram of step S5 in the embodiment, and step S5 in the embodiment includes the following steps: Figure 3 Figure 1 Step S51: Perform curing agent bond breaking behavior node analysis according to the curing agent unit micro-bond breaking behavior data, and generate curing agent bond breaking behavior node data;

[0115] Step S51: Perform curing agent bond breaking behavior node analysis according to the curing agent unit micro-bond breaking behavior data, and generate curing agent bond breaking behavior node data;

[0116] ​In the embodiment of the present application, according to the micro bond breaking behavior data of the curing agent unit, node recognition technology is used to locate and analyze the bond breaking behavior. The specific molecular position of each bond breaking event is determined through molecular structure analysis, and is marked as a bond breaking node, including the bond serial number on the molecular chain, the functional group it is in, and the spatial coordinates. All bond breaking nodes are counted, and the bond breaking frequency (number of bond breakings per unit time) and the bond breaking influence range (number of adjacent bond structure changes triggered after bond breaking) of each node are calculated. According to the bond breaking frequency from high to low, the top 20% of nodes are selected as key bond breaking nodes, and their molecular structure parameters (bond length, bond angle) and environmental correlation characteristics at the time of bond breaking are recorded. The position information, structure parameters, bond breaking frequency and influence range of these nodes are summarized to generate curing agent bond breaking behavior node data.

[0117] Step S52: Perform curing agent unit bond breaking and weathering degradation correlation analysis based on the curing agent bond breaking behavior node data to generate curing agent unit bond breaking-weathering degradation correlation data.

[0118] In the embodiment of the present application, based on the curing agent bond breaking behavior node data, correlation analysis is carried out in combination with the weathering performance test results. The macroscopic performance parameters (such as tensile strength, hardness, insulation resistance) of the curing agent under different bond breaking degrees are measured by a mechanical testing device, and are divided into gradients according to the bond breaking rate (proportion of broken nodes to total nodes) of the bond breaking nodes. Three parallel tests are performed under each gradient to take the average value. The bond breaking rate of each bond breaking node and the change rate (ratio of performance decline value to initial value) of the corresponding performance parameters are calculated, and the mathematical relationship (such as linear or exponential relationship) between the two is determined through correlation analysis. For example, the tensile strength decreases by 10% for every 10% increase in the bond breaking rate of the key bond breaking node. The correlation between the bond breaking rate and the performance change rate of all nodes is classified and sorted according to the node type to generate curing agent unit bond breaking-weathering degradation correlation data, which includes node number, bond breaking rate range, corresponding performance parameters and change rate.

[0119] Step S53: Perform curing agent bond breaking and weathering degradation mapping feature analysis according to the curing agent unit bond breaking-weathering degradation correlation data to generate curing agent bond breaking-weathering degradation mapping feature data.

[0120] In the embodiment of the present application, for the curing agent unit bond breaking-weathering degradation correlation data, a weight analysis and feature extraction method is used for mapping feature analysis. The influence weight of each bond breaking node on the weathering performance degradation is calculated, and the weight value is determined based on the correlation strength of the bond breaking rate and the performance change rate, and the weight sum is 100%. The nodes with the top 30% influence weight are selected as the bond breaking hub nodes, and the contribution degree change of the hub nodes to the performance degradation in different bond breaking stages (initial, middle and late) is analyzed. At the same time, according to the spatial distribution of the curing agent unit, the volume proportion of the unit at different positions is calculated, and the unit volume weighting coefficient of bond breaking-weathering degradation is determined by combining the influence weight of the bond breaking node in the unit, and the coefficient is calculated by using the weighted average method, and the sum of the product of the volume proportion and the influence weight is the coefficient value of the unit. The contribution degree characteristics of the hub nodes and the unit volume weighting coefficient are integrated to generate the mapping feature data of the curing agent bond breaking-weathering degradation.

[0121] Step S54: transmitting the curing agent bond breaking-weathering degradation mapping feature data to the environmental stress tensor-bond breaking evaluation model to optimize the bond breaking and weathering degradation relationship mapping of the curing agent, to obtain the environmental stress tensor-weathering degradation evaluation model.

[0122] In the embodiment of the present application, the curing agent bond breaking-weathering degradation mapping feature data is combined with the environmental stress tensor-bond breaking evaluation model to optimize the relationship mapping. The influence weight, weighting coefficient and other parameters in the mapping feature data are embedded into the bond breaking evaluation model by a feature fusion method to establish a direct correlation between the bond breaking probability and the weathering performance degradation degree. The model optimization process uses an iterative adjustment method to correct the conversion coefficient from bond breaking to degradation in the model according to the actual correlation data of the bond breaking probability and the performance degradation, so that the deviation between the predicted value and the measured value of the degradation degree output by the model is controlled within 4%. The optimized model can directly receive the environmental stress tensor input and output the degradation evaluation results including the bond breaking probability and the corresponding weathering performance parameters (such as strength retention rate and insulation resistance decay rate). The model stability is tested by multiple sets of verification data to ensure that the error fluctuation of five consecutive predictions does not exceed 2%, and finally the environmental stress tensor-weathering degradation evaluation model is obtained. The model includes feature parameters, conversion coefficients and degradation degree grading standards (such as threshold values of mild, moderate and severe degradation).

[0123] Further, step S53 includes the following steps:

[0124] Step S531: performing weathering degradation influence analysis of the bond breaking node based on the curing agent unit bond breaking-weathering degradation correlation data, generating bond breaking node-weathering degradation influence data, and performing weathering degradation influence characteristic analysis of the bond breaking hub node on the bond breaking node-weathering degradation influence data, to generate bond breaking hub node-weathering degradation influence characteristic data;

[0125] In the embodiment of the present application, according to the bond breaking-resistance degradation correlation data of the curing agent unit, the sensitivity analysis method is used to carry out the resistance degradation influence analysis of the bond breaking node. The ratio of the bond breaking rate change of each bond breaking node to the resistance performance parameter change is calculated, and the greater the ratio, the more significant the influence of the node on the resistance degradation. The ratio is defined as the influence coefficient. According to the influence coefficient from large to small, the top 25% of the bond breaking nodes are extracted as the bond breaking hub nodes, and the influence coefficient and the corresponding performance parameter change rule are recorded. For each bond breaking hub node, the continuous change curve of the resistance performance under different bond breaking degrees (bond breaking rate from 0% to 100%) is analyzed, and the critical bond breaking rate (bond breaking rate value at which the performance significantly decreases) and the degradation rate (performance decrease amplitude corresponding to unit bond breaking rate) are determined. The influence coefficient, critical bond breaking rate and degradation rate of all bond breaking hub nodes are summarized to generate the bond breaking hub node-resistance degradation influence characteristic data.

[0126] Step S532: Perform unit weighting coefficient analysis of bond breaking-resistance degradation according to the bond breaking-resistance degradation correlation data of the curing agent unit, and generate the bond breaking-resistance degradation unit volume weighting coefficient;

[0127] In the embodiment of the present application, based on the bond breaking-resistance degradation correlation data of the curing agent unit, the unit weighting coefficient analysis of bond breaking-resistance degradation is carried out in combination with the spatial distribution information of the curing agent unit. The volume proportion of each curing agent unit in the overall structure is determined by the volume calculation method, and the volume proportion is the ratio of the unit volume to the total volume, accurate to four decimal places. According to the influence coefficients of all bond breaking nodes in the unit, the average influence coefficient of the unit is calculated, and the calculation method is the sum of the product of the influence coefficients of the nodes in the unit and the bond breaking frequency proportion of the node. The unit volume proportion and the average influence coefficient are multiplied to obtain the bond breaking-resistance degradation weighting coefficient of the unit, and the weighting coefficient value range is controlled between 0 and 1. Repeat the above calculation for all units to generate the bond breaking-resistance degradation unit volume weighting coefficient.

[0128] Step S533: Perform curing agent bond breaking and resistance degradation mapping feature analysis through the bond breaking-resistance degradation unit volume weighting coefficient and the bond breaking hub node-resistance degradation influence characteristic data, and generate the curing agent bond breaking-resistance degradation mapping feature data.

[0129] In the embodiment of the present application, the broken bond-weathering degradation unit volume weighting coefficient is fused and analyzed with the broken bond central node-weathering degradation influence characteristic data. The broken bond-weathering degradation influence characteristic data of the broken bond central node is used to determine the effect of the broken bond of the main chemical bond of the curing agent on the weathering degradation of the curing agent, to exclude the influence of the broken bond of the irrelevant chemical bond of the curing agent on the weathering degradation of the curing agent, and to determine the actual coefficient of each broken bond of the curing agent in the overall unit by the broken bond-weathering degradation unit volume weighting coefficient, so as to determine the actual relationship between the broken bond of each unit of the curing agent and the weathering performance degradation in the finite element analysis. The weighting parameters of all broken bond central nodes are summarized, and the overall influence coefficient, the overall critical broken bond rate and the overall degradation rate are analyzed. The change law of these overall parameters with the broken bond process is analyzed to determine the quantitative mapping relationship between the broken bond rate and the weathering performance degradation degree. The overall parameters and the mapping relationship are integrated to generate the broken bond-weathering degradation mapping characteristic data of the curing agent.

[0130] Further, step S6 comprises the following steps:

[0131] Step S61: based on the curing agent application scene environment data, the unit environment boundary condition analysis of the trend time-varying change of the finite element curing agent reaction molecular structure data is performed to generate trend time-varying unit environment boundary condition data;

[0132] In the embodiment of the present application, according to the curing agent application scene environment data and the finite element curing agent reaction molecular structure data, the time series decomposition method is used to analyze the trend time-varying change of the unit environment boundary condition. The environment data is divided into several time segments according to the annual period, and each time segment contains continuous monitoring values of temperature, humidity, ultraviolet radiation and mechanical stress. The long-term change trend (such as the annual increasing or decreasing rule) and the periodic fluctuation characteristics (such as seasonal change) of each environmental parameter are separated by a trend extraction algorithm. Combined with the unit space coordinates in the finite element curing agent reaction molecular structure data, each unit is matched to the corresponding environment monitoring area, and the environment parameter values at future time nodes (such as the end of each quarter) are predicted according to the trend of the environmental parameters of the unit area, including the maximum value, the minimum value and the average value. The predicted values are associated with the unit coordinates to generate trend time-varying unit environment boundary condition data, which contains the environmental parameter range of each unit at different time nodes.

[0133] Step S62: according to the trend time-varying unit environment boundary condition data, the trend time-varying change of the unit environment stress tensor characteristic analysis is performed to generate trend time-varying unit environment stress tensor data;

[0134] In the embodiment of the present application, based on the trend time-varying unit environmental boundary condition data, the finite element stress calculation method is used to analyze the characteristics of the trend time-varying unit environmental stress tensor. The environmental parameters (temperature, humidity, ultraviolet radiation, mechanical stress) at each time node are converted into corresponding stress components, and the conversion is based on the physical conversion relationship between the parameters and the stress (such as the linear relationship between temperature change and thermal stress). The stress components are combined into the unit environmental stress tensor, and the eigenvalues (reflecting the stress intensity) and eigenvectors (reflecting the stress direction) of each stress tensor are calculated. The stress tensor of the same unit at different time nodes is tracked, the amplitude of the eigenvalue and the deflection angle of the eigenvector are recorded, and the evolution law of the stress tensor with time is analyzed (such as the intensity increasing and the direction stable, or the intensity fluctuating and the direction periodically changing). The different environmental stress tensor eigenvalues, eigenvectors and change rates of all units at each time node are summarized to generate the trend time-varying unit environmental stress tensor data, so as to ensure that the stress calculation accuracy is consistent with the accuracy of the finite element analysis.

[0135] Step S63: The trend time-varying unit environmental stress tensor data is transmitted to the environmental stress tensor-weathering degradation evaluation model for curing agent weathering performance degradation trend prediction, and the curing agent weathering performance degradation trend data is generated.

[0136] In the embodiment of the present application, the trend time-varying unit environmental stress tensor data is input into the environmental stress tensor-weathering degradation evaluation model for curing agent weathering performance degradation trend prediction. After the model receives the different environmental stress tensor data of each unit at each time node, the corresponding bond breaking probability is calculated according to the built-in mapping relationship, and then the bond breaking probability is converted into the weathering performance parameters (such as the tensile strength retention rate and the insulation resistance decay rate) through the bond breaking-weathering degradation conversion coefficient. The performance parameters of all units at the same time node are spatially weighted and averaged (the weight is the volume proportion of the unit), and the overall weathering performance parameter value is obtained. The overall parameter values of all time nodes are continuously calculated to form a curve of the change of the weathering performance with time, and the time node at which the performance parameter is first lower than 70% of the initial value (defined as the performance significant degradation point) is marked in the curve. The curve data, the parameter values at each time node and the significant degradation point information are summarized to generate the curing agent weathering performance degradation trend data.

[0137] Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting, the scope of the present application is defined by the appended claims rather than the above description, and therefore all changes falling within the meaning and scope of the equivalent elements of the application file are intended to be included in the present application.

[0138] The foregoing is considered as illustrative only of the principles of the application. Numerous modifications and changes will readily occur to those skilled in the art, and it is intended to embrace all such modifications and changes that fall within the scope of the application. Accordingly, the application is not to be restricted in scope to the specific embodiments disclosed herein but is to be accorded the full scope that the principles and novel features request appropriately granted.

Claims

1. A method for predicting the degradation of the weathering resistance of a curing agent based on finite element analysis, characterized by, The method comprises the following steps: Step S1: obtaining curing agent material characteristic data and curing agent application scene environment data; According to the curing agent material characteristic data, the material performance clustering processing of the curing agent component difference is performed to generate the clustering curing agent material performance data; Wherein, step S1 comprises the following steps: Step S11: obtaining curing agent material characteristic data and curing agent application scene environment data; Step S12: according to the curing agent material characteristic data, the material component analysis is carried out to obtain the curing agent material component data; Step S13: according to the curing agent material component data, the phase difference tensor analysis of the curing agent component difference is carried out to obtain the curing agent component difference phase difference tensor data; Step S14: according to the curing agent material component data and the curing agent component difference phase difference tensor data, the curing agent material performance clustering processing is carried out to generate the clustering curing agent material performance data; Step S2: based on the clustering curing agent material performance data and the curing agent application scene environment data, the finite element analysis processing of the curing agent reaction molecular structure is carried out to obtain the finite element curing agent reaction molecular structure data; Wherein, step S2 comprises the following steps: Step S21: according to the curing agent application scene environment data, the heterogeneous environment analysis of the curing agent reaction is carried out to generate the curing agent reaction heterogeneous environment data; Step S22: the environment field intensity gradient characteristic analysis is carried out on the curing agent reaction heterogeneous environment data to generate the environment field intensity gradient characteristic data; Step S23: according to the environment field intensity gradient characteristic data, the environment dynamic unit grid density design is carried out to obtain the environment dynamic unit grid density data; Step S24: through the environment dynamic unit grid density data, the finite element non-structure grid node design of the curing agent reaction is carried out to obtain the curing agent finite element grid node data, and the clustering curing agent material performance data and the curing agent reaction heterogeneous environment data are taken as the grid attribute node of the curing agent finite element grid node data. The finite element analysis processing of the curing agent reaction is carried out to obtain the finite element curing agent reaction data; Step S25: the molecular structure analysis of the finite element curing agent reaction is carried out on the finite element curing agent reaction data to obtain the finite element curing agent reaction molecular structure data; Step S3: according to the finite element curing agent reaction molecular structure data, the curing agent unit microcosmic bond breaking behavior analysis is carried out to generate the curing agent unit microcosmic bond breaking behavior data; based on the curing agent unit microcosmic bond breaking behavior data, the environmental influence curing agent bond breaking behavior characteristic analysis is carried out to generate the environmental influence curing agent bond breaking behavior characteristic data; Step S4: through the environmental influence curing agent bond breaking behavior characteristic data, the mapping relationship establishment of the environmental stress tensor and the bond breaking evaluation of each clustering curing agent is carried out to obtain the environmental stress tensor-bond breaking evaluation model; Wherein, step S4 comprises the following steps: Step S41: according to the environmental influence curing agent bond breaking behavior characteristic data, the environmental influence curing agent bond breaking probability distribution analysis is carried out to generate the environmental influence curing agent bond breaking probability distribution data; Step S42: according to the environmental influence curing agent bond breaking probability distribution data, the bond breaking probability distribution data analysis of the environmental stress tensor is carried out to generate the environmental stress tensor-bond breaking probability distribution data; Step S43: Perform bond breaking probability distribution classification processing on the environmental stress tensor-bond breaking probability distribution data according to the cluster curing agent material performance data, to generate classified environmental stress tensor-bond breaking probability distribution data; Step S44: Perform mapping relationship establishment of the environmental stress tensor and bond breaking evaluation of each cluster curing agent through the classified environmental stress tensor-bond breaking probability distribution data, to obtain an environmental stress tensor-bond breaking evaluation model; Step S5: Perform bond breaking and weathering degradation relationship mapping optimization of the curing agent on the environmental stress tensor-bond breaking evaluation model, to obtain an environmental stress tensor-weathering degradation evaluation model; Step S5 includes the following steps: Step S51: Perform curing agent bond breaking behavior node analysis according to the curing agent unit micro-bond breaking behavior data, to generate curing agent bond breaking behavior node data; Step S52: Perform curing agent unit bond breaking and weathering degradation correlation analysis based on the curing agent bond breaking behavior node data, to generate curing agent unit bond breaking-weathering degradation correlation data; Step S53: Perform curing agent bond breaking and weathering degradation mapping feature analysis according to the curing agent unit bond breaking-weathering degradation correlation data, to generate curing agent bond breaking-weathering degradation mapping feature data; Step S54: Transmit the curing agent bond breaking-weathering degradation mapping feature data to the environmental stress tensor-bond breaking evaluation model for curing agent bond breaking and weathering degradation relationship mapping optimization, to obtain an environmental stress tensor-weathering degradation evaluation model Step S6: Perform unit environmental stress tensor feature analysis of the trend time-varying change based on the curing agent application scene environmental data on the finite element curing agent reaction molecular structure data, to generate trend time-varying unit environmental stress tensor data; transmit the trend time-varying unit environmental stress tensor data to the environmental stress tensor-weathering degradation evaluation model for curing agent weathering performance degradation trend prediction, to generate curing agent weathering performance degradation trend data.

2. The method of claim 1, wherein the finite element analysis-based prediction of the weathering performance degradation of the curing agent is based on a finite element analysis of a test specimen of the curing agent. Step S3 includes the following steps: Step S31: Perform curing agent unit micro-bond breaking behavior analysis according to the finite element curing agent reaction molecular structure data, to generate curing agent unit micro-bond breaking behavior data; Step S32: Perform curing agent unit isomeric environment and micro-bond breaking behavior correlation processing on the curing agent unit micro-bond breaking behavior data based on the curing agent reaction isomeric environmental data, to generate curing agent unit environment-bond breaking behavior data; Step S33: Extract the curing agent reaction environmental impact factor of the curing agent unit environment-bond breaking behavior data; perform single environmental impact factor curing agent micro-bond breaking behavior analysis on the curing agent unit environment-bond breaking behavior data according to the curing agent reaction environmental impact factor, to generate single environmental impact factor-bond breaking behavior data; Step S34: Perform environmental impact factor interaction effect curing agent micro-bond breaking behavior analysis through the curing agent unit environment-bond breaking behavior data and the single environmental impact factor-bond breaking behavior data, to generate interactive environmental impact factor-bond breaking behavior data; Step S35: based on the single environmental impact factor-bond breaking behavior data and the interactive environmental impact factor-bond breaking behavior data, the curing agent bond breaking behavior characteristic analysis of environmental impact is performed to generate the environmental impact curing agent bond breaking behavior characteristic data.

3. The method of claim 2, wherein the finite element analysis-based prediction of the weathering performance degradation of the curing agent is based on a finite element analysis of the curing agent. The single environmental impact factor-bond breaking behavior data in step S33 includes temperature conduction impact factor-bond breaking behavior data, humidity diffusion-bond breaking behavior data, ultraviolet absorption characteristic-bond breaking behavior data, and mechanical stress characteristic-bond breaking behavior data.

4. The method of claim 1, wherein the finite element analysis-based prediction of the weathering performance degradation of the curing agent is based on a finite element analysis of a test specimen of the curing agent. Step S44 includes the following steps: The preset random forest algorithm is used to establish the mapping relationship between the environmental stress tensor and the bond breaking probability of various tree structures, so as to obtain the environmental stress tensor-bond breaking tree structure model, and the environmental stress tensor-bond breaking evaluation threshold parameter setting of the environmental stress tensor-bond breaking tree structure model is performed through the classified environmental stress tensor-bond breaking probability distribution data, so as to obtain the environmental stress tensor-bond breaking evaluation model.

5. The method of claim 1, wherein the finite element analysis-based prediction of the weathering performance degradation of the curing agent is based on a finite element analysis of a test specimen of the curing agent. Step S53 includes the following steps: Step S531: according to the curing agent unit bond breaking-weathering degradation correlation data, the weathering degradation influence analysis of the bond breaking node is performed to generate the bond breaking node-weathering degradation influence data, and the weathering degradation influence characteristic analysis of the bond breaking node-weathering degradation influence data is performed to generate the bond breaking node-weathering degradation influence characteristic data; Step S532: according to the curing agent unit bond breaking-weathering degradation correlation data, the unit weighting coefficient analysis of the bond breaking-weathering degradation is performed to generate the bond breaking-weathering degradation unit volume weighting coefficient; Step S533: through the bond breaking-weathering degradation unit volume weighting coefficient and the bond breaking node-weathering degradation influence characteristic data, the curing agent bond breaking and weathering degradation mapping characteristic analysis is performed to generate the curing agent bond breaking-weathering degradation mapping characteristic data.

6. The method of claim 1, wherein the finite element analysis-based prediction of the weathering performance degradation of the curing agent is based on a finite element analysis of a test specimen of the curing agent. Step S6 includes the following steps: Step S61: based on the curing agent application scene environmental data, the trend time sequence change unit environmental boundary condition analysis of the finite element curing agent reaction molecular structure data is performed to generate the trend time-varying unit environmental boundary condition data; Step S62: according to the trend time-varying unit environmental boundary condition data, the trend time sequence change unit environmental stress tensor characteristic analysis is performed to generate the trend time-varying unit environmental stress tensor data; Step S63: the trend time-varying unit environmental stress tensor data is transmitted to the environmental stress tensor-weathering degradation evaluation model to perform the curing agent weathering performance degradation trend prediction, and the curing agent weathering performance degradation trend data is generated.

Images