Vanadium-added steel cylinder section bearing performance prediction method considering non-uniform creep deterioration

Abstract

The invention discloses a vanadium-added steel cylinder section bearing performance prediction method considering non-uniform creep deterioration, and relates to the technical field of data processing, and the method comprises the following implementation steps: constructing a creep performance data set of materials in different areas of a vanadium-added steel cylinder section; performing data enhancement on the creep performance data set by using a generative adversarial neural network; constructing a creep damage evaluation index by using a time fraction method; constructing a residual strength data set after creep deterioration of materials in different areas of the vanadium-added steel cylinder section; establishing a residual intensity prediction model by using a data driving method; establishing a vanadium-added steel cylinder section finite element model; calculating creep damage distribution by using a time fraction method; calculating the residual strength distribution of the material after the vanadium-added steel cylinder section is subjected to creep deterioration; vanadium-added steel cylinder section bearing performance prediction considering non-uniform creep deterioration is realized. The method is relatively high in prediction precision, relatively high in reliability and relatively low in implementation cost, and can provide scientific guidance for applicability evaluation of the hydrogenation reactor in overdue service.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method for predicting the bearing performance of a vanadium-added steel cylinder section taking into account non-uniform creep degradation. Background Art

[0002] As key equipment for quality improvement and upgrading of the petrochemical and coal chemical industries, hydrogenation reactors are widely used in clean energy projects such as oil product improvement, heavy oil cracking, coal-to-synthetic natural gas, and direct coal liquefaction. The vanadium-added steel cylinder section is the core bearing component of the forged and welded hydrogenation reactor. Under actual service conditions, it is subjected to long-term thermal loads, and creep damage is its main failure mode. Moreover, the material properties of the vanadium-added steel cylinder section will inevitably undergo creep deterioration as the service time increases, which will also have an adverse effect on its structural bearing capacity. With the rapid development of science and technology, applicability evaluation technology is an important way to extend the service life of hydrogenation reactors, and accurately predicting the remaining bearing performance of the vanadium-added steel cylinder section after creep degradation is the core of the hydrogenation reactor applicability evaluation technology.

[0003] Currently, the existing methods for predicting the bearing performance of vanadium-added steel cylinder sections have the following two limitations:

[0004] In current engineering practice, vanadium-enriched steel cylinder segments and weld hangers are often placed inside hydrogenation reactors to assess the degradation effects of harsh environments on the material in different regions of the segment. However, unlike the actual service conditions of load-bearing components, these internal hangers are not subject to operating stresses and cannot reflect the effects of creep damage on vanadium-enriched steel, thus limiting the prediction of the load-bearing performance of vanadium-enriched steel cylinder segments.

[0005] 2. Forged-welded vanadium-added steel cylinder sections typically have a wall thickness exceeding 150 mm and an operating pressure exceeding 20 MPa. Under internal pressure, the stress distribution along the wall thickness is uneven, resulting in different creep behaviors at different wall thickness locations. Furthermore, forged-welded vanadium-added steel cylinder sections have distinct creep properties in the base metal, weld, and heat-affected zones. This results in low accuracy in existing predictions of the load-bearing performance of vanadium-added steel cylinder sections, making them unsuitable for evaluating the suitability of hydrogenation reactors undergoing extended service life. This impacts the accuracy of predictions of the load-bearing performance of vanadium-added steel cylinder sections.

[0006] Therefore, a new method for predicting the bearing performance of vanadium-added steel cylinder sections considering non-uniform creep degradation is proposed to solve the above problems. Summary of the Invention

[0007] The main purpose of the present invention is to provide a method for predicting the load-bearing performance of vanadium-added steel cylinder sections taking into account the uneven creep degradation, so as to solve the problem that the existing methods do not take into account the uneven creep behavior of vanadium-added steel cylinder sections, resulting in inaccurate prediction results and cannot be used for the applicability evaluation of extended service hydrogenation reactors.

[0008] To achieve the above-mentioned object, the technical solution adopted by the present invention is: a method for predicting the load-bearing performance of vanadium-added steel cylinder sections taking into account non-uniform creep degradation, the method comprising the following implementation steps: Step 1: Construct a creep performance dataset of materials in different regions of the vanadium-added steel cylinder; Step 2: Establish a creep performance prediction model based on the creep performance data set; Step 3: Use the time fraction method to construct the creep damage evaluation index; Step 4: Construct the residual strength data set based on the creep damage evaluation index; Step 5: Establish a residual strength prediction model based on the residual strength dataset; Step 6: Establish a finite element model of the vanadium-added steel cylinder section and calculate the stress distribution of the vanadium-added steel cylinder section under service conditions; Step 7: Obtain creep damage distribution based on the creep performance prediction model; Step 8: Input the creep damage distribution and stress distribution into the residual strength prediction model to obtain the residual strength distribution of the material after creep degradation of the vanadium-added steel cylinder; Step 9: Incorporate the material residual strength distribution into the plastic limit analysis to predict the load-bearing performance of the vanadium-added steel cylinder considering uneven creep degradation.

[0009] The vanadium-added steel refers to vanadium-improved chromium-molybdenum steel, including 1Cr-1Mo-V steel, 2.25Cr-1Mo-0.25V steel and 3Cr-1Mo-0.25V steel;

[0010] The specific method for constructing the creep performance data set of materials in different regions of the vanadium-added steel cylinder section is as follows:

[0011] According to the forging welding process of vanadium-doped steel cylinder, the corresponding forging welding simulation test blocks were made, and materials M1, M2 and M3 were cut from the weld zone, heat-affected zone and parent material zone respectively;

[0012] The working temperature of the vanadium-added steel cylinder section is T s Used as the test temperature, and n test stresses s1, s2, ..., s are selected within a reasonable range. n , where n≥5;

[0013] Creep rupture tests were performed on materials M1, M2, and M3 at the same test temperature and different test stresses, and creep performance data sets D1, D2, and D3 of materials M1, M2, and M3 under different creep conditions were obtained.

[0014] The specific method of step 2 is as follows:

[0015] Generative adversarial neural networks are used to perform data enhancement on the creep performance datasets D1, D2, and D3 measured in step 1, respectively, to obtain the creep performance enhanced datasets of materials M1, M2, and M3. 、 and ;

[0016] Based on the obtained enhanced data set, creep property prediction models of materials M1, M2 and M3 were established using empirical equations.

[0017] The step 3 uses the time fraction method to construct a creep damage evaluation index.

[0018] Step 4 is to test at the temperature and test stresses s1, s2, ..., s n Under different conditions, creep degradation treatments were performed on materials M1, M2 and M3 to varying degrees, and the test temperature Tensile tests were performed on all degraded materials under different creep conditions to obtain the residual strength data corresponding to different degrees of deterioration of materials M1, M2 and M3, and the data set was recorded as 、 and ;

[0019] Step 5 is based on the measured residual intensity data set 、 and , using the data-driven method, a residual strength prediction model for materials M1, M2 and M3 after deterioration to different degrees under different creep conditions is established.

[0020] The step 6 is to establish a finite element model of the vanadium-added steel cylinder section including the weld zone, heat-affected zone and parent metal zone according to the geometric parameters, manufacturing process and service conditions of the vanadium-added steel cylinder section, and to set the working pressure The finite element model is input as boundary conditions, and materials M1, M2 and M3 are The mechanical response data under the conditions of finite element model are input into the corresponding areas as material parameters, and then the stress distribution of vanadium-added steel cylinder section under service conditions is calculated.

[0021] The step 7 is to calculate the stress distribution of the vanadium-added steel cylinder obtained in step 6 under service conditions. , input the creep performance prediction model established in step 2, calculate the creep rupture life distribution of the vanadium-added steel cylinder section under service conditions, and use the constructed creep damage evaluation index to calculate the creep damage distribution of the vanadium-added steel cylinder section after long-term service.

[0022] Step 8 is to convert the creep damage distribution obtained in step 7 into and the stress distribution obtained in step 6 , input the residual strength prediction model established in step 5 to calculate the service life of the vanadium steel cylinder section The residual strength distribution of the material after a certain period of time.

[0023] Step 9 is to distribute the residual strength of the vanadium-added steel cylinder section material obtained in step 8. and The improved double-sided model based on Melan's theorem is input to obtain the ultimate load of the vanadium-added steel cylinder section;

[0024] In step 5, based on the remaining intensity data set , using data-driven methods, the residual strength of the material Mk after creep degradation is predicted, and the bearing performance prediction of vanadium-added steel cylinder sections considering uneven creep degradation is realized.

[0025] The present invention has the following beneficial effects:

[0026] 1. This invention adopts a solution that combines data-driven, numerical simulation, and plastic limit analysis, avoiding complex factors such as physical mechanisms and load history. It provides a reliable method for accurately predicting the in-service load-bearing performance of vanadium-added steel cylinder sections that have been in service for a long time under creep conditions. It also forms a suitability evaluation technology that systematically considers creep degradation for hydrogenation reactors that have been in service for an extended period of time, making the prediction of the load-bearing performance of vanadium-added steel cylinder sections more accurate.

[0027] 2. This invention incorporates the complex non-uniform creep degradation caused by material welding and structural load into the load-bearing performance prediction of vanadium-added steel cylinder sections, which can provide technical support for the applicability evaluation of hydrogenation reactors in extended service life.

[0028] 3. The present invention uses a generative adversarial neural network model to supplement the measured creep data set with samples, which not only reduces the demand for creep test samples of the creep empirical model, but also helps to improve the accuracy of its creep performance prediction. Therefore, the present invention has certain cost advantages. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is an overall flow chart of a method for predicting the load-bearing performance of vanadium-added steel cylinder sections taking into account non-uniform creep degradation according to the present invention; Figure 2 The actual vanadium-added steel cylinder section provided by Lanzhou Lanshi Heavy Equipment Co., Ltd. involved in the embodiment of the present invention; Figure 3 Enhanced dataset for creep properties of material M1; Figure 4 Enhanced dataset for creep properties of material M2; Figure 5 Enhanced dataset for creep properties of material M3; Figure 6This is the prediction result of the residual yield strength of material M1 after creep degradation; Figure 7 is the prediction result of the residual tensile strength of material M1 after creep degradation; Figure 8 This is the prediction result of the residual yield strength of material M2 after creep degradation; Figure 9 is the prediction result of the residual tensile strength of material M2 after creep degradation; Figure 10 This is the prediction result of the residual yield strength of material M3 after creep degradation; Figure 11 This is the prediction result of the residual tensile strength of material M3 after creep degradation; Figure 12 A finite element model of a cylinder section including a weld zone, a heat-affected zone, and a base material zone constructed according to an embodiment of the present invention; Figure 13 This is the equivalent stress distribution of the vanadium-added steel cylinder section involved in the embodiment of the present invention at a working pressure of 35 MPa and a working temperature of 550°C; Figure 14 This is the evolution law of the load-bearing performance of the vanadium-added steel cylinder section involved in the embodiment of the present invention during service at a working pressure of 35 MPa and a working temperature of 550°C. DETAILED DESCRIPTION

[0030] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0031] Please refer to Figures 1 to 11 A method for predicting the load-bearing performance of vanadium-added steel cylinder sections considering non-uniform creep degradation is shown, comprising the following implementation steps:

[0032] Step 1: Construct a creep performance dataset of materials in different regions of the vanadium-added steel cylinder. The creep performance in step 1 includes creep time-creep strain curve, creep rupture life, creep elongation after rupture, and creep section reduction rate.

[0033] Step 2: Generative adversarial neural networks are used to enhance the creep performance dataset. Based on the enhanced dataset, a creep performance prediction model for materials in different regions of the vanadium-added steel cylinder is established.

[0034] Step 3: Use the time fraction method to construct the creep damage evaluation index;

[0035] Step 4: Using creep degradation and tensile tests, construct a residual strength dataset of materials in different regions of the vanadium-added steel cylinder after creep degradation;

[0036] Step 5: Using a data-driven approach, a residual strength prediction model for different regions of the vanadium-added steel cylinder section after creep degradation is established;

[0037] Step 6: Establish a finite element model of the vanadium-added steel cylinder section including the weld zone, heat-affected zone, and base metal zone, and calculate its stress distribution under service conditions;

[0038] Step 7: Based on the stress distribution of the vanadium-added steel cylinder and the creep performance prediction model of the materials in different regions, the creep rupture life distribution of the vanadium-added steel cylinder is obtained, and its creep damage distribution is calculated using the time fractional method;

[0039] Step 8: Input the creep damage distribution and stress distribution of the vanadium-added steel cylinder segment into the residual strength prediction model of the material in the corresponding area of ​​the vanadium-added steel cylinder segment to calculate the residual strength distribution of the material after creep degradation of the vanadium-added steel cylinder segment;

[0040] Step 9: Incorporate the residual strength distribution of the material after creep degradation of the vanadium-added steel cylinder into the plastic limit analysis to realize the prediction of the load-bearing performance of the vanadium-added steel cylinder considering uneven creep degradation.

[0041] Vanadium-added steel refers to vanadium-modified chromium-molybdenum steel, including 1Cr-1Mo-V steel, 2.25Cr-1Mo-0.25V steel, and 3Cr-1Mo-0.25V steel. In this embodiment, 2.25Cr-1Mo-0.25V steel is selected as the model material. The raw material for making the forging welding simulation test block in step 1 should be taken from the base material of the actual vanadium-added steel cylinder section forging;

[0042] The welding parameters for the simulated forging welding test block in step 1 should be consistent with the welding process of the actual vanadium-doped steel cylinder, including the groove type and its size parameters, preheating temperature, solder, welding speed, welding current and post-weld heat treatment parameters;

[0043] The specific method for constructing the creep performance dataset of materials in different regions of the vanadium-added steel cylinder section is as follows:

[0044] According to the actual vanadium-added steel cylinder section (this embodiment takes the actual vanadium-added steel cylinder section with a wall thickness of 152 mm, an inner diameter of 2400 mm, and a length of 1700 mm provided by Lanzhou Lanshi Heavy Equipment Co., Ltd. as an example, Figure 2 A forging welding process (shown) was used to produce corresponding forging welding simulation test blocks (the welding parameters for producing the forging welding simulation test blocks should be consistent with the welding process of the actual vanadium-added steel cylinder section, including the groove type and its size parameters, preheating temperature, solder, welding speed, welding current, welding voltage and post-weld heat treatment parameters. The main welding process parameters used in this embodiment were provided by Lanzhou Lanshi Heavy Equipment Co., Ltd., as shown in Table 1), and materials M1, M2 and M3 were cut from the weld area, heat-affected zone and base material area respectively;

[0045] Table 1 - Main parameters of welding process used in this embodiment

[0046] The working temperature of the vanadium-added steel cylinder section is T s Used as the test temperature (T s =550℃), and select five test stresses s1, s2, ..., s5 within a reasonable range (i.e., above the working stress of the vanadium-added steel cylinder and below 90% of the minimum yield strength of materials M1, M2, and M3). The reasonable range of test stress in step 1 refers to the stress range between the actual service stress level of the vanadium-added steel cylinder material and the yield strength of the vanadium-added steel, as specifically listed in Table 2;

[0047] Table 2 - Test stress levels selected for this example

[0048] Creep rupture tests were performed on materials M1, M2, and M3 at the same test temperature and different test stresses, and the creep performance data sets D1, D2, and D3 of materials M1, M2, and M3 under different creep conditions were obtained, as shown in Table 3.

[0049] Table 3 - Creep performance data set obtained in this example (only creep rupture life data is shown here)

[0050] In step 2, the empirical equations include the Larson-Miller parameter method, the Manson-Haferd parameter method, the θ projection method, and the Ω method. The specific method of step 2 is as follows:

[0051] Generate adversarial neural network for material M k Creep performance dataset Implement data augmentation, which includes the following sub-steps:

[0052] Step 21: Randomly extract 80% of the data from the creep performance datasets D1, D2, and D3 as a training set for generative adversarial neural network fitting, and the remaining 20% ​​of the data as a test set for fitting result evaluation. The data division results of this embodiment are shown in Table 4;

[0053] Table 4 - Division results of creep rupture life data training set and test set in this example

[0054] Step 22: Based on the creep performance data set Based on the sample size of , a fully connected neural network with appropriate capacity is constructed as the generator and discriminator of the generative adversarial neural network, and the model parameters are initialized;

[0055] Step 23: Sample from the random distribution, input the generator, and obtain the material M k A batch of virtual data on creep performance;

[0056] Step 24: Based on the training set obtained in step 21 and the virtual data obtained in step 23, use the Wasserstein distance and gradient descent algorithm to optimize the model parameters of the discriminator to improve its ability to distinguish between true and false data;

[0057] Step 25: Use the discriminator obtained in step 24 to optimize the model parameters of the generator so that its output data distribution gradually approaches the data distribution of the training set in step 21;

[0058] Step 26: Repeat steps 23 to 25 until the generator's ability to fit the data distribution of the training set in step 21 is maximized;

[0059] Step 27: Use the training set obtained in step 21 to evaluate the fitting accuracy of the generator obtained in step 26, and use the test set obtained in step 21 to evaluate the generalization performance of the generator obtained in step 26. If both the fitting accuracy and generalization performance are acceptable, a usable data augmentation model is obtained; otherwise, adjust the hyperparameters of the fully connected neural network in step 22, and repeat steps 22 to 27.

[0060] Step 28: Analyze the distribution characteristics of the creep performance data on the creep stress coordinate axis to determine the creep stress ranges with missing data samples (in this embodiment, creep performance data sets D1, D2, and D3 lack creep performance data for vanadium-added steel in the normal operating stress range of the barrel segment, specifically, the range of 100 to 200 MPa);

[0061] Step 29: Using the data enhancement model obtained in step 27, supplement the creep stress interval determined in step 28 with samples to obtain a creep performance enhancement dataset. ), perform data enhancement on the creep performance datasets D1, D2 and D3 measured in step 1, and obtain the creep performance enhanced datasets of materials M1, M2 and M3 、 and , specifically Figure 3 、 Figure 4 and Figure 5 As shown;

[0062] Based on the obtained enhanced data set, the Larson-Miller empirical equation is used to establish creep property prediction models for materials M1, M2, and M3, respectively, as follows:

[0063]

[0064]

[0065] ;

[0066] Where, represents the creep stress, Indicates material M k At temperature and stress The Larson-Miller parameter under Calculated, where M k represents any one of the materials M1, M2 and M3 (i.e., its subscript k is 1, 2, 3), C LM is the material constant (set to 20 in this embodiment), Indicates material M k Creep rupture life;

[0067] Step 3 uses the time fraction method to construct a creep damage evaluation index, as follows:

[0068] ;

[0069] Where, represents the damage factor that describes the degree of creep degradation of the material, Indicates the creep service time of the material, Indicates the creep rupture life of the material.

[0070] Step 4 is as follows:

[0071] At the test temperature and test stresses s1, s2, ..., s n Under different conditions, creep degradation treatments were performed on materials M1, M2 and M3 to varying degrees, and the test temperature All degraded materials were subjected to tensile tests under k Conduct creep degradation and tensile testing, which includes the following sub-steps:

[0072] Step 41: Based on the creep performance data set measured in step 1 , determine the material M k Creep rupture life and the end time of the steady-state creep stage (See Table 5 for details);

[0073] Table 5 - End time of steady-state creep stage of each material under different creep conditions obtained in this embodiment

[0074] Step 42: Based on the creep damage evaluation index constructed in step 3, calculate the material Mk Damage factor corresponding to the end of the steady-state creep stage , see Table 6 for details;

[0075] Table 6 - Damage factors corresponding to the end of steady state of each material under different creep conditions obtained in this example

[0076] Step 43: In [0, ] m damage factors are selected at equal intervals within the range 、 、…、 ,in ≥5, no. The damage factor is recorded as In this embodiment, the damage factors corresponding to the end of the steady-state creep stage for materials M1, M2, and M3 under different creep conditions are all between 0.60 and 0.70, which are relatively close to each other. Therefore, the same five damage factors are selected for materials M1, M2, and M3 (see Table 7 for details). It must be pointed out that when the damage factors corresponding to the end of the steady-state creep stage for materials M1, M2, and M3 under different creep conditions are significantly different from each other, m damage factors can be selected for different materials and different creep conditions.

[0077] Table 7 - Five damage factors selected in this example

[0078] Step 44: From Material M k Sampling processing The same creep specimens, The creep specimen is recorded as ;

[0079] Step 45: At the test temperature and test stress s1, the processed in step 44 The creep specimens were subjected to creep interruption tests of different lengths. The creep time is given by and The product of ;

[0080] Step 46: All creep degradation specimens obtained in step 45 are further processed into high temperature tensile specimens and tested at the test temperature. All samples were subjected to tensile tests to measure their yield strength and tensile strength;

[0081] Step 47: Adjust the test stress in step 45 to s2, s3, ..., s n , repeat steps 44 to 46 respectively to obtain material Mk Residual strength data sets corresponding to different degrees of degradation under different creep conditions ), obtain the residual strength data corresponding to different degrees of deterioration of materials M1, M2 and M3 under different creep conditions, and record the data set as 、 and , see Table 8, Table 9 and Table 10 for details;

[0082] Table 8 - Residual strength data set of material M1 obtained in this example :

[0083] Table 9 - Residual strength data set of material M2 obtained in this example :

[0084] Table 10 - Residual strength data set of material M3 obtained in this example :

[0085] Step 5 is based on the measured residual intensity dataset 、 and , using a data-driven approach (a single hidden layer fully connected neural network model is used in this embodiment), a residual strength prediction model for materials M1, M2, and M3 after different degrees of degradation under different creep conditions is established, as follows:

[0086] ;

[0087] Where, Indicates the material Mk at temperature The residual yield strength under Indicates material M k At temperature The residual tensile strength under Indicates the use of data-driven methods to analyze material M k The fitting function established is based on the residual intensity data set , using data-driven methods to predict material M k The residual strength after creep degradation (the prediction results of materials M1, M2 and M3 obtained in this embodiment are as follows Figure 6 、 Figure 7 and Figure 8 、 Figure 9 、 Figure 10 and Figure 11 The implementation specifically includes the following sub-steps:

[0088] Step 51: From the Remaining Intensity Dataset 80% of the data are randomly selected as the training set for data-driven model fitting, and the remaining 20% ​​of the data are used as the test set for fitting result evaluation;

[0089] Step 52: Based on the training set obtained in step 51, use grid search and cross-validation to optimize the model hyperparameters, and then train the model to maximize its performance on material M. k The ability to predict residual strength after creep degradation;

[0090] Step 53: Use the training set obtained in step 51 to evaluate the fitting accuracy of the data-driven model trained in step 53, and use the test set obtained in step 51 to evaluate the generalization performance of the data-driven model trained in step 53. If both the fitting accuracy and the generalization performance are acceptable, then the available material M is obtained. k Residual strength prediction model after creep degradation (the performance of the creep degradation residual strength prediction model for materials M1, M2, and M3 constructed in this embodiment is shown in Table 11. The root mean square error on the training set and test set is low, the R2 determination coefficient is greater than 0.95, and the two are relatively close, so its fitting accuracy and generalization performance are considered acceptable); otherwise, the data-driven model is replaced, and steps 51 to 53 are repeated.

[0091] Table 11 - Performance of the creep degradation residual strength prediction model for materials M1, M2, and M3 constructed in this embodiment

[0092] Step 6 is as follows:

[0093] According to the geometric parameters of the vanadium-added steel cylinder section (in this embodiment, the geometric parameters of the vanadium-added steel cylinder section are wall thickness 152 mm, inner diameter 2400 mm and length 1700 mm), the manufacturing process (in this embodiment, the vanadium-added steel cylinder section is forged and welded, and the welding adopts V-shaped groove full penetration), and the service conditions (in this embodiment, the service conditions of the vanadium-added steel cylinder section are working pressure 40 MPa and working temperature 550°C), a finite element model of the vanadium-added steel cylinder section including the weld zone, heat-affected zone and base metal zone is established (see Figure 12 ), the working pressure = 40 MPa is input into the finite element model as the boundary condition, and the materials M1, M2 and M3 are = The mechanical response data at 550℃ are input as material parameters into the corresponding areas of the finite element model, and then the stress distribution of the vanadium steel cylinder under service conditions is calculated (such as Figure 13As shown, it must be pointed out that the distribution cloud diagram can only show part of the calculation results of the surface of the vanadium-added steel cylinder section, and it is difficult to clearly explain the data mapping and transfer relationship in the subsequent calculation steps. Therefore, in this embodiment, the subsequent calculation process is described using mathematical expressions, as follows:

[0094] ;

[0095] Where, Indicates that the vanadium steel cylinder section is at working temperature and work pressure The stress distribution under Indicates the spatial position of any point in the vanadium steel cylinder in the cylindrical coordinate system, A finite element model that takes into account the inhomogeneous mechanical behavior of the weld zone, heat-affected zone, and base material zone.

[0096] Step 7: The stress distribution of the vanadium-added steel cylinder obtained in step 6 under service conditions , input the creep performance prediction model established in step 2, and calculate the creep rupture life distribution of the vanadium-added steel cylinder under service conditions, specifically:

[0097] ;

[0098] Where, Indicates that the vanadium steel cylinder section is at working temperature and work pressure Creep rupture life distribution under ;

[0099] The constructed creep damage evaluation index is used to calculate the creep damage distribution of vanadium-added steel cylinder sections after long-term service, specifically:

[0100] ;

[0101] Where, Indicates the creep service time of vanadium-added steel cylinder section, Indicates that the vanadium steel cylinder section is at working temperature and work pressure Serve Creep damage distribution after time.

[0102] Step 8 is to convert the creep damage distribution obtained in step 7 into and the stress distribution obtained in step 6 , input the corresponding area material residual strength prediction model established in step 5, calculate the service life of vanadium steel cylinder section The residual strength distribution of the material after a certain period of time is as follows:

[0103] ;

[0104] Where, Indicates that the vanadium steel cylinder section is at working temperature and work pressure Serve The residual yield strength distribution of the material after a certain period of time, Indicates that the vanadium steel cylinder section is at working temperature and work pressure Serve The residual tensile strength distribution of the material after a certain period of time.

[0105] The specific steps of step 9 are as follows:

[0106] The residual strength distribution of the vanadium-added steel cylinder section material obtained in step 8 and Enter the improved two-sided model based on Melan's theorem, as follows:

[0107] ;

[0108] Where α represents the load multiplier, represents the yield function, Indicates the Gauss point number of the finite element discrete model of the vanadium-added steel cylinder. represents the total number of Gaussian points, Indicates the corner point number of the load space involved in the vanadium-added steel cylinder section. represents the total number of load corner points, Indicates the load corner The elastic stress field generated in The stress vector at each Gaussian point is The finite element discrete model of the vanadium steel cylinder is shown in Figure 1. The self-equilibrium residual stress at each Gaussian point is The finite element discrete model of the vanadium steel cylinder is shown in Figure 1. The back stress at each Gaussian point describes the kinematic strengthening effect. The finite element discrete model of the vanadium steel cylinder is shown in Figure 1. The cylindrical coordinate position of the Gaussian point, The finite element discrete model of the vanadium steel cylinder is shown in Figure 1. Gauss point at the working temperature and work pressure Serve The residual yield strength after time, The finite element discrete model of the vanadium steel cylinder is shown in Figure 1. Gauss point at the working temperature and work pressure Serve The residual tensile strength after a certain period of time, The finite element discrete model of the vanadium steel cylinder is shown in Figure 1. Stress-to-nodal load conversion factors at Gaussian points related to meshing and boundary conditions;

[0109] Based on the dimensionality reduction iterative framework, the effective set algorithm is used to solve the improved double-sided model and obtain the vanadium steel cylinder section at the working temperature. and work pressure Serve The ultimate load after a long time is used to predict the bearing performance of the vanadium steel cylinder section with uneven creep degradation; the calculation results of this embodiment are as follows Figure 14 As shown, it should be noted that in this embodiment, only the vanadium-added steel cylinder section with complete structure is considered, and when it is partially damaged, it is considered that it loses its bearing capacity.

[0110] The geometric parameters and manufacturing process of the vanadium-added steel cylinder section involved in this embodiment are consistent with engineering practice. However, it should be pointed out that in order to accelerate the deterioration of the vanadium-added steel cylinder section and reduce the cost of creep testing, the working temperature of 550°C and the working pressure of 35 MPa used in this embodiment are higher than engineering practice. That is, the service conditions used in this embodiment are for illustration only.

[0111] It must be pointed out that the formulas in steps 6, 7 and 8 can clearly show how the working stress distribution, creep life distribution, creep damage distribution and residual strength distribution of the vanadium-added steel cylinder section in this embodiment are mapped and calculated one by one. The formula in step 9 can clearly express how to incorporate the residual strength calculation results into the calculation framework of plastic limit analysis. These formulas can assist technical personnel in this field to understand and apply this application, and can facilitate their programming to implement the technical content of this application.

[0112] In the present invention, a method for predicting the bearing performance of vanadium-added steel cylinder sections considering uneven creep degradation is provided. First, a creep performance dataset of materials in different regions of the vanadium-added steel cylinder section is constructed; a generative adversarial neural network is used to perform data enhancement on the creep performance dataset, and based on the enhanced dataset, a creep performance prediction model for materials in different regions of the vanadium-added steel cylinder section is established. By adopting a solution combining data-driven, numerical simulation, and plastic limit analysis, complex factors such as physical mechanism and load history are avoided, providing a reliable method for accurately predicting the in-service bearing performance of vanadium-added steel cylinder sections that have been serving in creep conditions for a long time. A set of applicability evaluation technologies that systematically consider creep degradation are formed for hydrogenation reactors that have been in service for an extended period of time, so that the prediction accuracy of the bearing performance of vanadium-added steel cylinder sections is higher. A creep damage evaluation index is constructed using a time fractional method; a residual strength dataset of materials in different regions of the vanadium-added steel cylinder section after creep degradation is constructed using creep degradation and tensile tests; a residual strength prediction model for materials in different regions of the vanadium-added steel cylinder section after creep degradation is established using a data-driven method; a system including weld zone, heat-affected zone and parent material is established. The invention provides a finite element model of a vanadium-added steel cylinder section in a certain area, and calculates its stress distribution under service conditions; based on the stress distribution of the vanadium-added steel cylinder section and the creep performance prediction model of the material in different areas, the creep rupture life distribution of the vanadium-added steel cylinder section is obtained, and the creep damage distribution is calculated using the time fractional method; the creep damage distribution and stress distribution of the vanadium-added steel cylinder section are input into the residual strength prediction model of the material in the corresponding area of ​​the vanadium-added steel cylinder section, and the residual strength distribution of the material after creep degradation of the vanadium-added steel cylinder section is calculated. By incorporating the complex non-uniform creep degradation caused by material welding and structural load into the load-bearing performance prediction of the vanadium-added steel cylinder section, technical support can be provided for the applicability evaluation of the extended service hydrogenation reactor; the residual strength distribution of the material after creep degradation of the vanadium-added steel cylinder section is incorporated into the plastic limit analysis, and the load-bearing performance prediction of the vanadium-added steel cylinder section considering the non-uniform creep degradation is realized. By adopting the generative adversarial neural network model to supplement the sample of the measured creep data set, it not only reduces the demand for creep test samples of the creep empirical model, but also helps to improve the accuracy of its creep performance prediction. Therefore, the invention has certain cost advantages.

[0113] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for predicting the load-bearing performance of vanadium-added steel cylinder sections considering non-uniform creep degradation, characterized in that: The method comprises the following implementation steps: Step 1: Construct a creep performance dataset of materials in different regions of the vanadium-added steel cylinder; Step 2: Establish a creep performance prediction model based on the creep performance data set; Step 3: Use the time fraction method to construct the creep damage evaluation index; Step 4: Construct the residual strength data set based on the creep damage evaluation index; Step 5: Establish a residual strength prediction model based on the residual strength dataset; Step 6: Establish a finite element model of the vanadium-added steel cylinder section and calculate the stress distribution of the vanadium-added steel cylinder section under service conditions; Step 7: Obtain creep damage distribution based on the creep performance prediction model; Step 8: Input the creep damage distribution and stress distribution into the residual strength prediction model to obtain the residual strength distribution of the material after creep degradation of the vanadium-added steel cylinder; Step 9: Incorporate the material residual strength distribution into the plastic limit analysis to predict the load-bearing performance of the vanadium-added steel cylinder considering uneven creep degradation.

2. The method according to claim 1, wherein: The specific method for constructing the creep performance data set of materials in different regions of the vanadium-added steel cylinder section is as follows: According to the forging welding process of vanadium-doped steel cylinder, the corresponding forging welding simulation test blocks were made, and materials M1, M2 and M3 were cut from the weld zone, heat-affected zone and parent material zone respectively; The working temperature of the vanadium-added steel cylinder section is T s Used as the test temperature, and n test stresses s1, s2, ..., s are selected within a reasonable range. n , where n≥5; At the same test temperature and different test stresses, creep rupture tests are carried out on materials M1, M2 and M3, respectively, and creep performance data sets D1, D2 and D3 of materials M1, M2 and M3 under different creep conditions are obtained.

3. The method according to claim 1, wherein: The remaining intensity data set constructed in step 4 is specifically as follows: At the test temperature and test stresses s1, s2, ..., s n Under different conditions, materials M1, M2 and M3 were subjected to creep degradation treatments of different degrees, and the creep degradation was tested at the test temperature. Tensile tests were performed on all degraded materials under different creep conditions to obtain the residual strength data sets corresponding to different degrees of deterioration of materials M1, M2 and M3. 、 and ; Step 5 is based on the measured residual intensity data set 、 and , using the data-driven method, a residual strength prediction model for materials M1, M2 and M3 after deterioration to different degrees under different creep conditions is established.

4. The method according to claim 1, wherein: The step 6 is specifically as follows: According to the geometric parameters, manufacturing process and service conditions of the vanadium-added steel cylinder, a finite element model of the vanadium-added steel cylinder including the weld zone, heat-affected zone and base metal zone was established. The finite element model is input as boundary conditions, and materials M1, M2 and M3 are The mechanical response data under the condition of finite element model are input as material parameters to the corresponding areas of the finite element model, and then the vanadium steel cylinder section is calculated at the working temperature. and work pressure Stress distribution under ,in Indicates the spatial position inside the vanadium-added steel cylinder section in the cylindrical coordinate system.

5. The method according to claim 4, characterized in that: The specific step of step 7 is to calculate the stress distribution of the vanadium-added steel cylinder obtained in step 6 under service conditions. , input the creep performance prediction model established in step 2, calculate the creep rupture life distribution of the vanadium-added steel cylinder section under service conditions, and use the constructed creep damage evaluation index to calculate the creep damage distribution of the vanadium-added steel cylinder section after long-term service.

6. The method according to claim 1, wherein: Step 8 will be creep damage distribution and the stress distribution obtained in step 6 , input the residual strength prediction model established in step 5 to calculate the service life of the vanadium steel cylinder section The residual strength distribution of the material after a certain period of time.

7. The method according to claim 1, wherein: The specific steps of step 9 are as follows: The residual strength distribution of the material obtained in step 8 and The improved double-sided model based on Melan's theorem is input to obtain the ultimate load of the vanadium-added steel cylinder section.

8. The method according to claim 3, wherein: In step 4, the material M k Conduct creep degradation and tensile testing, including the following steps: Step 41: Based on the creep performance data set , determine the material M k Creep rupture life and the end time of the steady-state creep stage ; Step 42: Calculate material M based on creep damage evaluation index k Damage factor corresponding to the end of the steady-state creep stage ; Step 43: In [0, ] m damage factors are selected at equal intervals within the range 、 、…、 ,in ≥5, no. The damage factor is recorded as ; Step 44: From Material M k Sampling processing The same creep specimens, The creep specimen is recorded as ; Step 45: At the test temperature and test stress s1, the processed in step 44 The creep specimens were subjected to creep interruption tests of different lengths. The creep time is given by and The product of ; Step 46: All creep degradation specimens obtained in step 45 are further processed into high temperature tensile specimens and tested at the test temperature. All samples were subjected to tensile tests to measure their yield strength and tensile strength; Step 47: Adjust the test stress in step 45 to s2, s3, ..., s n , repeat steps 44 to 46 respectively to obtain material M k Residual strength data sets corresponding to different degrees of degradation under different creep conditions .

9. The method according to claim 3, wherein: In step 5, based on the measured residual intensity data set , using data-driven methods to predict material M k The residual strength after creep degradation specifically includes the following steps: Step 51: From the Remaining Intensity Dataset 80% of the data are randomly selected as the training set for data-driven model fitting, and the remaining 20% ​​of the data are used as the test set for fitting result evaluation; Step 52: Based on the training set obtained in step 51, use grid search and cross-validation to optimize the model hyperparameters, and then train the model to maximize its performance on material M. k Ability to predict residual strength after creep degradation; Step 53: Use the training set obtained in step 51 to evaluate the fitting accuracy of the data-driven model trained in step 53, and use the test set obtained in step 51 to evaluate the generalization performance of the data-driven model trained in step 53. If both the fitting accuracy and the generalization performance are acceptable, then the available material M is obtained. k The residual strength prediction model after creep degradation is used; otherwise, the data-driven model is replaced and steps 51 to 53 are repeated.

Images