A method for optimizing a cladded reinforcing bar rolling process based on finite elements

Abstract

The application discloses a kind of based on finite element's cladding reinforcement rolling process optimization method, comprising the following steps: S1, using thermal simulation experiment, obtains the stress-strain curve of base cladding material at different temperatures, obtains the material parameters of base layer material and cladding material using Jmatpro software, establishes finite element model;S2, cladding reinforcement rolling is divided into rough, medium, finish rolling, after rough rolling in the established finite element model, blank head, tail part is cut off, and the grid is re-division, then again in the established finite element model is carried out medium rolling, after medium rolling, blank head, tail part is cut off again, and the grid is re-division, finish rolling is carried out again, until rolling is finished;S3, based on finite element model, optimization is carried out to pass and rolling parameter.The application improves the accuracy of simulation result compared with traditional finite element model, shortens the process optimization cycle, and reduces the optimization cost.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of rolling process optimization, and particularly relates to a cladding steel bar rolling process optimization method based on finite elements. BACKGROUND

[0002] In recent years, with the continuous advancement of China's industrialization process, tunnels, bridges and marine engineering construction are increasing, and steel bars, as the most widely used building steel, have a huge demand. Traditional carbon steel bars are prone to corrosion and expansion due to low chromium and nickel content, and when the volume increases beyond the elastic limit of concrete, it will cause damage to the concrete structure, resulting in a significant reduction in the service life of the building. Cladding steel is a new type of composite steel bar with carbon steel as the base and stainless steel as the outer layer. The corrosion resistance of the steel bar is close to that of stainless steel bar, but the cost is only one-third of that of stainless steel bar, and it can be widely used in marine engineering, petroleum, transportation and infrastructure fields.

[0003] Steel bar rolling is a process of interactive influence of microstructure evolution and macro deformation, especially in the composite rolling of two metals, and the deformation trend is more complex. At present, there is no mature process for cladding steel bar rolling in steel plants, and relying on industrial production to explore the best rolling parameters is not only inefficient but also expensive. Finite element technology is a method of simulating real physical systems by mathematical approximation, which not only has the advantage of short time consumption, but also can effectively obtain the stress and strain conditions of the product in the processing process.

[0004] At present, there is no mature technology for finite element simulation and rolling process optimization of cladding steel bars. Patent No. CN202010429397.1 discloses a metal composite plate rolling process optimization method, which does not optimize the rolling process, but mainly uses rapid quenching to fix the interface structure of the metal composite plate, and then obtains the interface characteristics of the intermediate link in the composite plate rolling. Patent No. CN201811130653.6 discloses a low carbon steel DSIT rolling process optimization method, which mainly optimizes the rolling temperature through thermal simulation, but the thermal simulation equipment is limited and mainly used for simulating plates, which is not suitable for cladding steel bars. Patent No. CN202110732028.4 discloses a ball flat steel rolling process optimization method based on finite element simulation, which mainly re-divides the finite element grid, but does not introduce the process optimization process in detail.

[0005] In summary, the existing simulation and process optimization technology cannot be applied to cladding steel bars, and using traditional process to roll cladding steel bars is prone to problems such as inconsistent deformation of the inner and outer layers, resulting in uneven cladding thickness distribution. Therefore, it is urgent to design a simulation and process optimization method suitable for cladding steel bars. SUMMARY

[0006] The purpose of the present application is to provide a finite element-based cladding steel bar rolling process optimization method to solve the problem that the existing simulation and process optimization technology cannot be applied to cladding steel bars and the uneven thickness distribution of cladding is prone to occur when cladding steel bars are rolled by using traditional process due to inconsistent deformation of inner and outer layers.

[0007] To achieve the above purpose, the present application provides a finite element-based cladding steel bar rolling process optimization method, comprising the following steps:

[0008] S1, obtaining stress-strain curves of base cladding materials at different temperatures by using thermal simulation experiment, obtaining material parameters of the base layer and the cladding layer by using Jmatpro software, and establishing a finite element model;

[0009] S2, dividing the cladding steel bar rolling into rough rolling, medium rolling and finish rolling, cutting off the head and tail of the blank after the rough rolling in the established finite element model, re-dividing the grid, then performing the medium rolling in the established finite element model, cutting off the head and tail of the blank again after the medium rolling, re-dividing the grid, and then performing the finish rolling until the rolling is completed;

[0010] S3, optimizing the pass and rolling parameters based on the finite element model.

[0011] In a specific embodiment, the S1 specific steps comprise:

[0012] S11, processing the cladding steel bar into a thermal simulation sample with a specified shape;

[0013] S12, clamping the thermal simulation sample, and then heating it to 900℃, 950℃, 1000℃, 1050℃ and 1100℃ respectively;

[0014] S13, after heating, keeping warm, and then compressing at a strain rate of 1s -1 After compression, the stress-strain curve is derived.

[0015] In a specific embodiment, the S1 specific steps further comprise:

[0016] S14, measuring the content of each element in the cladding steel bar by using a spectrometer;

[0017] S15, then inputting the content of each element into the Jmatpro software;

[0018] S16, deriving the material parameters of the base layer and the cladding layer by using the Jmatpro software.

[0019] In a specific embodiment, the material parameters include thermal conductivity, specific heat, and Poisson's ratio.

[0020] In a specific embodiment, the S2 specific steps include:

[0021] S21, a rough rolling simulation model of the cladding reinforcing bar is established by using ABAQUS software, and calculation is submitted;

[0022] S22, after the rough rolling simulation model is calculated, the rough rolling simulation result is imported into the HYPERMESH software;

[0023] S23, the head and tail of the rough rolling simulation result are cut off by using a cutting plane, and the grid is re-divided;

[0024] S24, the rough rolling simulation result after the grid re-division is imported into ABAQUS, and a medium rolling simulation model is established based on the result;

[0025] S25, after the medium rolling simulation model is calculated, the medium rolling simulation result is imported into the HYPERMESH software;

[0026] S26, the head and tail of the medium rolling simulation result are cut off by using a cutting plane, and the grid is re-divided;

[0027] S27, the medium rolling simulation result after the grid re-division is imported into ABAQUS, and a finish rolling simulation model is established based on the result;

[0028] S28, after the finish rolling simulation model is calculated, the finished product cladding reinforcing bar is rolled out.

[0029] In a specific embodiment, the S3 specific steps include:

[0030] S31, according to the industrial site situation, the rolling process parameters of the cladding reinforcing bar are determined;

[0031] S32, according to the limited cladding reinforcing bar industrial experiment, the rolling process parameters are preliminarily screened;

[0032] S33, the rolling process parameters after the preliminary screening are twice screened by using orthogonal variance analysis, and the rolling process parameters to be optimized are determined;

[0033] S34, taking the rolling process parameters to be optimized as variables, a BP neural network algorithm is constructed to realize the prediction of the minimum value of the cladding thickness;

[0034] S35, the BP neural network prediction value is taken as an adaptive value, and the genetic algorithm is used to optimize the rolling parameters, and the optimization target is to maximize the adaptive value, that is, the greater the minimum value of the cladding thickness is, the better;

[0035] S36, the optimized result is output.

[0036] In a specific embodiment, in the S34 step:

[0037] In the constructed BP neural network algorithm, the automatic optimization determination of the number of hidden layers and nodes is completed through the embedded particle swarm algorithm.

[0038] In a specific embodiment, the specific steps of the embedded particle swarm algorithm are as follows:

[0039] S341, the number of hidden layers and nodes are used as particle swarm optimization independent variables;

[0040] S342, the range of the number of hidden layers and nodes is determined;

[0041] S343, the number of initial particles and the number of iterations are determined;

[0042] S344, the minimum value of the coating thickness is maximized as the target, and the automatic optimization is performed to obtain the optimal number of hidden layers and nodes.

[0043] In a specific embodiment, in step S344, the fitness value of each particle is the minimum value of the coating thickness predicted by the BP neural network corresponding to the particle.

[0044] In a specific embodiment, in S2, in the rough rolling stage, each node of the composite surface is defined as a friction constraint, and in the intermediate rolling stage, each node of the composite surface is defined as a binding constraint.

[0045] Compared with the prior art, the present application has the following beneficial effects:

[0046] Compared with the traditional finite element model, the present application improves the accuracy of the simulation results, shortens the process optimization period, and reduces the optimization cost.

[0047] The present application effectively solves the problems of grid distortion leading to calculation termination and the like by cutting the head and tail and re-dividing the grid.

[0048] The present application effectively connects the entire rolling process, that is, the rough rolling result is used as the intermediate rolling input, and the intermediate rolling result is used as the finish rolling input, compared with the traditional finite element model, the simulation error of the present application is reduced from 22.22% (traditional model) to 5.71%.

[0049] The present application effectively realizes the prediction of the minimum value of the coating thickness of the coated reinforcement, and obtains the optimal rolling parameters of the coated reinforcement. The cost of experimental optimization is reduced, and the optimization period is shortened. Industrial experiment verification shows that the predicted value of the coating thickness only differs by 2% from the actual value.

[0050] In addition to the purposes, features and advantages described above, the present application has other purposes, features and advantages. The present application will be further described in detail below. BRIEF DESCRIPTION OF DRAWINGS

[0051] The accompanying drawings, which form a part of this application, are included to provide a further understanding of the application, illustrate the preferred embodiments of the application and assist in

[0052] Figure 1 is a step-by-step simulation calculation flow chart of an embodiment of the application;

[0053] Figure 2 is a schematic diagram of a coated steel bar section of an embodiment of the application;

[0054] Figure 3 is a simulation model crescent groove grid division schematic diagram of an embodiment of the application;

[0055] Figure 4 is a simulation result schematic diagram of an embodiment of the application;

[0056] Figure 5 is a practical rolling result schematic diagram of an embodiment of the application;

[0057] Figure 6 is a traditional simulation model rolling schematic diagram;

[0058] Figure 7 is an optimization process schematic diagram of an embodiment of the application;

[0059] Wherein, 1, base layer; 2, coating. DETAILED DESCRIPTION

[0060] The embodiments of the application will be described in detail below, and the specific embodiments described herein are only used to explain the application and not to limit the application.

[0061] A method for optimizing a coated steel bar rolling process based on finite elements, comprising the following steps:

[0062] S1, using a thermal simulation experiment, obtaining stress-strain curves of base and coating materials at different temperatures, using Jmatpro software to obtain material parameters of the base layer 1 and the coating layer 2, and establishing a finite element model.

[0063] The S1 specific steps include:

[0064] S11, processing the coated steel bar into a specified shape of a thermal simulation sample.

[0065] S12, clamping the thermal simulation sample, and then heating to 900 DEG C, 950 DEG C, 1000 DEG C, 1050 DEG C, 1100 DEG C, respectively.

[0066] S13, after heating, keeping warm for 5 min, and then at 1s -1Strain rate compression, compression amount is 80%, after compression, the stress-strain curve is exported.

[0067] S14, the content of each element in the coated reinforcement is measured by a spectrometer.

[0068] S15, then input the content of each element into Jmatpro software.

[0069] S16, the material parameters of the base layer and the coating are exported by Jmatpro software.

[0070] The material parameters include thermal conductivity, specific heat, and Poisson's ratio.

[0071] S2, the coated reinforcement rolling is divided into rough, medium and finish rolling, after the rough rolling ends in the established finite element model, the head and tail of the blank are cut off, and the grid is re-divided, then the medium rolling is carried out in the established finite element model, after the medium rolling ends, the head and tail of the blank are cut off again, and the grid is re-divided, and then the finish rolling is carried out until the rolling ends.

[0072] The specific steps of S2 include:

[0073] S21, a simulation model of rough rolling of coated reinforcement is established by using ABAQUS software, and the operation is submitted;

[0074] S22, after the rough rolling simulation model is calculated, the rough rolling simulation result is imported into HYPERMESH software;

[0075] S23, the head and tail of the rough rolling simulation result are cut off by using cutting plane, and the grid is re-divided;

[0076] S24, the rough rolling simulation result after grid redivision is imported into ABAQUS, and a medium rolling simulation model is established based on the result;

[0077] S25, when the medium rolling simulation model is calculated, the medium rolling simulation result is imported into HYPERMESH software;

[0078] S26, the head and tail of the medium rolling simulation result are cut off by using cutting plane, and the grid is re-divided;

[0079] S27, the medium rolling simulation result after grid redivision is imported into ABAQUS, and a finish rolling simulation model is established based on the result;

[0080] S28, after the finish rolling simulation model is calculated, the finished coated reinforcement is rolled out.

[0081] In S2, in the rough rolling stage, each node of the composite surface is defined as a friction constraint, and in the medium rolling stage, each node of the composite surface is defined as a binding constraint.

[0082] S3, optimizing the pass and rolling parameters based on the finite element model.

[0083] The S3 includes the following specific steps:

[0084] S31, determining the cladding steel bar rolling process parameters according to the industrial field conditions;

[0085] S32, preliminarily screening the rolling process parameters according to the limited cladding steel bar industrial experiments;

[0086] S33, using the orthogonal variance analysis to perform secondary screening on the preliminarily screened rolling process parameters, and determining the rolling process parameters to be optimized;

[0087] S34, using the roulette method to randomly generate 300-800 combinations with the rolling process parameters to be optimized as variables;

[0088] Substituting the 300-800 combinations into the finite element model to obtain the cladding minimum values corresponding to the 300-800 combinations;

[0089] According to the 300-800 combinations and the corresponding cladding minimum values, a BP neural network algorithm is constructed to realize the prediction of the minimum cladding thickness;

[0090] In the S34, the following steps are included:

[0091] In the constructed BP neural network algorithm, the number of hidden layers and nodes is automatically optimized and determined by embedding a particle swarm algorithm.

[0092] The specific steps of the embedded particle swarm algorithm are as follows:

[0093] S341, taking the number of hidden layers and nodes as the particle swarm optimization independent variables;

[0094] S342, determining the range of the number of hidden layers and nodes;

[0095] S343, determining the number of initial particles and the number of iterations;

[0096] S344, taking the minimization of the cladding thickness as the target, and performing automatic optimization to obtain the optimal number of hidden layers and nodes.

[0097] In the step S344, the fitness value of each particle is the cladding thickness minimum value predicted by the BP neural network corresponding to the particle.

[0098] S35, using the genetic algorithm to optimize the rolling parameters with the BP neural network prediction value as the fitness value, and the optimization target is the maximization of the fitness value, i.e., the greater the cladding thickness minimum value, the better;

[0099] S36, output the optimized result.

[0100] Example 1

[0101] First, the cladding reinforcement is processed into a cylinder of φ8x10mm using a machining center, and then clamped at both ends of the thermal simulator. Then, heating is performed, followed by compression at a rate of 1s -1 -1 to 80%, and finally the stress-strain curve is obtained.

[0102] The elemental contents of C, Si, Mn, P, S, Cr, Ni, Ti, and V in the base layer and the cladding layer are measured using a spectrometer, and then the above elemental contents are input into the Jmatpro software to derive the material parameters of the base layer and the cladding layer, including the thermal conductivity, specific heat, and Poisson's ratio.

[0103] According to the above material parameters, a finite element model of the cladding reinforcement is established, and the entire rolling process is divided into rough rolling, intermediate rolling, and finish rolling, of which the rough rolling and intermediate rolling are six passes, and the finish rolling is two passes. The model establishment includes material property definition, analysis step definition, boundary condition definition, and mesh division.

[0104] (1) Material property definition.

[0105] The cladding material of each simulation model is 316L, and the base material is 20MnSiV. The thermal stress-strain curves of the above two materials have been measured by thermal simulation, and the material properties are simulated by the software Jmatpro. It should be particularly noted that the rollers are rigid bodies and do not need to define material properties.

[0106] (2) Analysis step definition.

[0107] The analysis step type of each simulation model is dynamic, temperature-displacement, display analysis step, and the analysis step length is 10s. To speed up the calculation speed and improve the simulation efficiency, the mass scaling of the rolled piece is performed, and the scaling factor is set to 10000.

[0108] (3) Boundary condition definition.

[0109] During the rolling process, steel bars are primarily subject to the combined effects of force and temperature fields. Therefore, the boundary conditions for 20MnSiV / 316L steel bars include load, temperature, and contact definitions. First, the load is defined. The initial speed of the workpiece for each simulation model is 600 mm / s. Next, the temperature is defined. According to actual measurements with a thermometer, the workpiece temperatures for roughing, intermediate, and finishing rolling are 1050°C, 1040°C, and 1020°C, respectively. The surface temperature of each roll is 120°C, and the ambient temperature is 30°C. Furthermore, the workpiece dissipates heat naturally in the air through convection and radiation, with a heat dissipation coefficient set at 0.17. Finally, the contact definition is performed. Each simulation model requires two sets of contact definitions: the first set is between the rolls and the workpiece surface, and the second set is between the cladding stainless steel tube and the base mandrel surface. The contact definition results for each simulation model are shown in Table 1. Within the friction constraint, the friction coefficient is related to the sliding velocity, compressive stress, and rolling temperature.

[0110] Table 1 Contact definition results of roughing, medium and finishing rolling models

[0111]

[0112] Through testing, it was found that after the rough rolling of the covered steel bars, the bonding strength of the composite surface reached the national calibration value. Therefore, in the rough rolling model, the composite surface was defined as a friction constraint, and in the intermediate rolling and finishing rolling models, the composite surface was defined as a binding constraint.

[0113] (4) Network division.

[0114] Meshing is the most important step in establishing a finite element model. The purpose of meshing is to discretize the continuum into small units, thereby simulating the real physical system. The mesh geometry, size, and density will affect the calculation accuracy of the finite element model to a certain extent. In ABAQUS software, the main mesh shapes are tetrahedrons, hexahedrons, and wedges. Compared with tetrahedrons and wedges, hexahedron meshes have higher quality and are easier to converge in simulation results. Therefore, hexahedron meshes are preferred when meshing. In addition, the mesh needs to be refined in areas where stress is more concentrated or curvature changes greatly, such as fillets and sharp corners, while the mesh can be appropriately diluted in areas where curvature changes more gently to improve calculation efficiency.

[0115] According to the above principles, the mesh size, shape and type of each simulation model are shown in Table 2.

[0116] Table 2 Grid shape, size and type of rough, medium and finishing rolling models

[0117]

[0118] To effectively simulate the metal flow trend in the rolling process, the roll fillet and the crescent groove are refined in mesh. The mesh size at the above-mentioned positions is 0.5 mm. The mesh division result of the crescent groove is shown in Figure 3 .

[0119] In addition, to prevent mesh distortion in the rolling process, a step-by-step simulation method is adopted, which divides the rolling process into rough rolling, intermediate rolling and finish rolling. After the rough rolling is completed, the head and tail of the mesh distortion are cut off, and the mesh is re-divided. Finally, the mesh-redivided model is imported into the ABAQUS software as the input of the intermediate rolling model. The schematic diagram of the step-by-step simulation is shown in Figure 1 , and the specific steps are as follows:

[0120] (1) According to the steps described in the foregoing, the ABAQUS software is used to establish the clad reinforcement rough rolling simulation model, and the operation is submitted;

[0121] (2) After the model calculation is completed, the rough rolling simulation result is imported into the HYPERMESH software;

[0122] (3) The head and tail of the rolled piece are cut off by using solid cutting, and the mesh is re-divided. The specific steps are as follows:

[0123] 1) Use the "mesh-surface" command to convert the rolled piece at the end of rough rolling into a surface.

[0124] 2) Use the "surface-solid" command to convert the surface into a solid.

[0125] 3) Delete the original mesh after the rough rolling is completed.

[0126] 4) Use the "solid cutting" command to cut off the head and tail with large deformation.

[0127] 5) Use the "mesh creation" command to perform hexahedral mesh division on the cut-off solid.

[0128] 6) Import the mesh-divided model into ABAQUS as the input of the intermediate rolling model, and perform simulation.

[0129] (4) When the intermediate rolling simulation model calculation is completed, the intermediate rolling simulation result is imported into the HYPERMESH software; the head and tail of the intermediate rolling simulation result are cut off by using the cutting plane, and the mesh is re-divided; the mesh-redivided intermediate rolling simulation result is imported into ABAQUS, and the finish rolling simulation model is established based on the result; after the finish rolling simulation model calculation is completed, the finished clad reinforcement is rolled out. The step-by-step simulation result is shown in Figure 4 . The error statistics of the step-by-step simulation result are shown in Table 3.

[0130] Table 3 Error statistics of step-by-step simulation result

[0131]

[0132] To evaluate the pros and cons of the step-by-step simulation, a traditional finishing model of clad reinforcement is established. Unlike the step-by-step simulation, the traditional finishing simulation directly sets the rolled piece as a cylinder with uniform distribution of cladding. The traditional finishing model is shown in Fig. 4. The simulation error of the traditional finishing model is shown in Table 4. Figure 6

[0133] Table 4 Simulation error statistics of traditional finishing model

[0134]

[0135] From Tables 3 and 4, it can be seen that the step-by-step simulation result is very close to the actual value, with a maximum simulation error of only 5.71%, while the traditional simulation result is 20%.

[0136] Process optimization steps:

[0137] First, the industrial site conditions determine the rolling parameters of the clad reinforcement, including 16 parameters such as intermediate rolling temperature, finishing temperature, heating time, heating temperature, rough rolling hole type, intermediate rolling hole type, rough rolling reduction, finishing reduction, round hole diameter (product hole before), flat oval groove bottom width, flat oval hole height, crescent groove chamfer radius, product hole roll diameter, product hole rolling speed, push tension, and initial rolling temperature.

[0138] Next, using 25 industrial experiment data, non-standard regression analysis is performed on the above parameters. The analysis is shown in Table 5.

[0139] Table 5 Non-standard regression analysis results

[0140]

[0141]

[0142] From Table 5, select the 8 parameters with larger regression coefficients as the dependent variables of orthogonal variance analysis. They are round hole diameter (product hole before), flat oval groove bottom width, flat oval hole height, crescent groove chamfer radius, product hole roll diameter, product hole rolling speed, push tension, and initial rolling temperature.

[0143] Then, orthogonal analysis is performed on the above 8 parameters. The orthogonal table type is finally selected as the nine-factor eight-level orthogonal experiment table L 64 (8 9 ), and the factor level values are shown in Table 6.

[0144] Table 6 Factor level values

[0145]

[0146] The orthogonal analysis results are shown in Table 7.​

[0147] Statistical values of each factor of Table 7

[0148]

[0149] According to the results of Table 7, it is finally determined that the round hole diameter, the flat ellipse groove bottom width, the flat ellipse hole height, the finished product hole rolling speed and the initial rolling temperature have greater influence on the minimum value of the coating, and the above five parameters are used as the input parameters of the BP neural network.

[0150] Using the roulette method, the above five parameters are randomly generated 500 parameter combinations, and the 500 combinations are substituted into the finite element model to obtain the minimum value of the coating corresponding to the 500 combinations;

[0151] According to the 500 parameter combinations and the corresponding minimum value of the coating, a BP neural network algorithm is constructed, wherein 400 parameter combinations are training samples and 100 parameter combinations are prediction samples. In the BP neural network, the number of hidden layers and nodes is optimized by the particle swarm algorithm, and the specific steps are as follows:

[0152] (1) The number of hidden layers is set to 1-10. The number of hidden layer nodes is set to 10-100.

[0153] (2) The number of initial particle swarm is set to 20, and the number of iterations is set to 5000;

[0154] (3) Automatic optimization is performed to obtain the best number of hidden layers and nodes. It should be particularly noted that the fitness value of each particle is the minimum value of the coating thickness predicted by the BP neural network corresponding to the particle.

[0155] Then, the optimized number of hidden layers 4 and the number of nodes 13 are substituted into the BP neural network algorithm.

[0156] Finally, the rolling parameters are optimized by using the genetic algorithm, and the specific steps are as follows:

[0157] The population size is set to 20, and the number of iterations is 2000;

[0158] The optimized BP neural network is used as the fitness value of the genetic algorithm;

[0159] The fitness value is maximized as the target to optimize the rolling parameters.

[0160] Finally, the optimized parameter combination is: the round hole diameter is 1.23d 钢筋 , the flat ellipse groove bottom width is 0.95d 钢筋 , the flat ellipse hole height is 0.69d 钢筋, the finished hole rolling speed 48 rad / s, the initial rolling temperature 1100 °C. At this time the minimum coating thickness reached 0.49 mm.

[0161] The above parameters are applied to the actual production, the actual production of the minimum coating thickness of 0.5 mm, compared with Figure 5 It can be seen that the minimum coating thickness is significantly improved, in addition, the actual value (0.5 mm) and the predicted value (0.49 mm) of the coating thickness can be compared, and the error of the predicted value of the coating thickness is only 2%.

[0162] The above is a further detailed description of the present application in combination with specific preferred embodiments, and cannot be considered as limiting the specific implementation of the present application to these descriptions. For ordinary skilled persons in the art to which the present application belongs, without departing from the concept of the present application, a number of simple deductions and substitutions can be made, which should be considered as belonging to the protection scope of the present application.

Claims

1. A method for optimizing the rolling process of coated steel bars based on finite element method, characterized in that: The following steps are involved: S1. Using thermal simulation experiments, obtain the stress-strain curves of the base and cladding materials at different temperatures, use Jmatpro software to obtain the material parameters of the base and cladding, and establish a finite element model; S2, the rolling of the covered steel bar is divided into rough rolling, medium rolling, and finishing rolling. After the rough rolling is completed in the established finite element model, the head and tail of the billet are cut off and the mesh is re-divided. Then, medium rolling is performed in the established finite element model. After the medium rolling is completed, the head and tail of the billet are cut off again and the mesh is re-divided. Then, finishing rolling is performed until the rolling is completed; The specific steps of S2 include: S21. Use ABAQUS software to establish a simulation model for rough rolling of coated steel bars and submit the calculation; S22. After the rough rolling simulation model is calculated, the rough rolling simulation results are imported into the HYPERMESH software; S23, using a cutting plane to cut off the head and tail of the rough rolling simulation result, and re-dividing the grid; S24, importing the rough rolling simulation results after mesh re-division into ABAQUS, and establishing a medium rolling simulation model based on the results; S25. After the calculation of the intermediate rolling simulation model is completed, the intermediate rolling simulation results are imported into the HYPERMESH software; S26, using a cutting plane to cut off the head and tail of the rolling simulation result, and re-dividing the mesh; S27, importing the intermediate rolling simulation results after mesh re-division into ABAQUS, and establishing a finishing rolling simulation model based on the results; S28, after the calculation of the finishing rolling simulation model is completed, the finished coated steel bar is rolled out; S3. Optimize the pass profile and rolling parameters based on the finite element model; The specific steps of S3 include: S31. Determine the rolling process parameters of the coated steel bars according to the industrial site conditions; S32. Based on limited industrial experiments on coated steel bars, preliminary screening of rolling process parameters was conducted; S33, performing secondary screening on the rolling process parameters after the preliminary screening using orthogonal variance analysis to determine the rolling process parameters that need to be optimized; S34, using the rolling process parameters to be optimized as variables, constructing a BP neural network algorithm to achieve prediction of the minimum coating thickness; S35, using the BP neural network prediction value as the adaptability value, and optimizing the rolling parameters using a genetic algorithm, wherein the optimization goal is to maximize the adaptability value, that is, the larger the minimum value of the coating thickness, the better; S36. Output the optimized result.

2. The method for optimizing the rolling process of coated steel bars based on finite element method according to claim 1, characterized in that: The specific steps of S1 include: S11, processing the covering steel bar into a thermal simulation specimen of a specified shape; S12, clamping the thermal simulation sample, and then heating it to 900°C, 950°C, 1000°C, 1050°C, and 1100°C respectively; S13, after heating, keep warm, then -1 After compression is completed, the stress-strain curve is derived.

3. The method for optimizing the rolling process of coated steel bars based on finite element method according to claim 2, characterized in that: The specific step S1 further includes: S14. Measure the content of each element in the covered steel bar using a spectrometer; S15. Then input the content of each element into Jmatpro software; S16. Use Jmatpro software to export the material parameters of the base layer and the covering layer.

4. The method for optimizing the rolling process of coated steel bars based on finite element method according to claim 3, characterized in that: The material parameters include thermal conductivity, specific heat, and Poisson's ratio.

5. The method for optimizing the rolling process of coated steel bars based on finite element method according to claim 1, characterized in that: In the step S34: In the constructed BP neural network algorithm, the number of hidden layers and nodes is automatically optimized through the embedded particle swarm algorithm.

6. The method for optimizing the rolling process of coated steel bars based on finite element method according to claim 5, characterized in that: The specific steps of the embedded particle swarm algorithm are as follows: S341, the number of hidden layers and the number of nodes are used as independent variables for particle swarm optimization; S342, determining the range of the number of hidden layers and the number of nodes; S343, determining the number of initialized particle swarms and the number of iterations; S344. With the goal of maximizing the minimum value of the coating thickness, automatic optimization is performed to obtain the optimal number of hidden layers and nodes.

7. The method for optimizing the rolling process of coated steel bars based on finite element method according to claim 6, characterized in that: In step S344 , the fitness value of each particle is the minimum coating thickness predicted by the BP neural network corresponding to the particle.

8. The method for optimizing the rolling process of coated steel bars based on finite element method according to claim 1, characterized in that: In the above S2, in the rough rolling stage, each node of the composite surface is defined as a friction constraint, and in the intermediate rolling stage, each node of the composite surface is defined as a binding constraint.

Images