Intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation

Abstract

The invention belongs to the technical field of thermal deformation, and relates to an intelligent prediction method for thermal deformation of a high-strength titanium alloy based on finite element simulation. Establishing a dynamic recrystallization fraction prediction model, and calculating a dynamic recrystallization fraction prediction value; comparing the predicted value of the dynamic recrystallization fraction with a dynamic recrystallization threshold value, triggering a metallographic complementary test and updating the prediction model; constructing a mechanical property prediction model, and calculating the numerical values of the tensile strength and the ductility; comparing the tensile strength with a tensile strength qualified threshold value, comparing the elongation with an elongation qualified threshold value, and judging whether the thermal deformation performance of the titanium alloy is qualified or not; applying a multi-objective optimization algorithm to find a temperature and strain rate combination enabling the tensile strength to be the highest; and dynamically correcting model parameters. According to the method, the problems that an existing prediction method lacks an intelligent prediction and optimization algorithm, the situation that performance is unqualified cannot be automatically found, and technological parameter optimization cannot be triggered can be solved.

Description

Technical Field

[0001] The invention relates to an intelligent prediction method for thermal deformation of a high-strength titanium alloy based on finite element simulation, and belongs to the technical field of thermal deformation. Background Art

[0002] Traditional experimental methods require a lot of time and resources and are costly. Due to the limitations of experimental conditions, it is often difficult to fully and accurately simulate the thermal deformation behavior of titanium alloys during actual processing. Secondly, the thermal deformation properties of titanium alloys are affected by many factors, such as chemical composition, grain size, deformation temperature, strain rate, etc. The interaction between these factors makes the thermal deformation properties of titanium alloys complex and changeable. Traditional experimental methods often find it difficult to fully consider these factors, resulting in low accuracy of prediction results.

[0003] With the rapid development of materials science and finite element simulation technology, people have begun to explore the use of advanced computational methods and simulation techniques to predict the thermal deformation properties of titanium alloys. As a powerful numerical simulation tool, finite element simulation technology can simulate the stress, strain and temperature field distribution of titanium alloys under complex deformation conditions, providing a new way to predict the thermal deformation properties of titanium alloys. The existing prediction methods based on finite element simulation still have many shortcomings.

[0004] Based on the above problems, existing prediction methods lack intelligent prediction and optimization algorithms, and are unable to automatically detect performance failures and trigger process parameter optimization. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation, which can solve the problem that the existing prediction methods lack intelligent prediction and optimization algorithms, cannot automatically detect unqualified performance and trigger process parameter optimization.

[0006] The technical solution adopted by the present invention is as follows: An intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation comprises the following steps: S1. Collect data related to the chemical composition, grain size, deformation temperature, strain rate and stress value of titanium alloy; S2. Calculate the peak strain and critical strain according to the collected data, and establish a dynamic recrystallization fraction prediction model to calculate the dynamic recrystallization fraction prediction value; S3, setting a dynamic recrystallization threshold, comparing the predicted value of the dynamic recrystallization fraction with the dynamic recrystallization threshold, triggering a metallographic supplementary test and updating the prediction model; S4. Construct a mechanical property prediction model to calculate the values ​​of tensile strength and elongation; S5. Set a qualified threshold value for tensile strength and a qualified threshold value for elongation, compare the size of the tensile strength with the qualified threshold value for tensile strength, and compare the size of the elongation with the qualified threshold value for elongation, and judge whether the thermal deformation performance of the titanium alloy is qualified. If it is unqualified, optimize the process parameters; S6. Based on the premise that the elongation value is qualified, a multi-objective optimization algorithm is applied to find the temperature and strain rate combination that maximizes the tensile strength; S7. The optimized process parameters are verified by finite element simulation, and closed-loop feedback is performed based on the simulation results and actual production data to dynamically correct the model parameters.

[0007] A further solution of the present invention, step S1, comprises the following steps: Use direct reading spectrometer to collect composition data of titanium alloy, including the content of elements related to Ti, Al, V, and O; Using the Gleeble thermal simulation test machine, combined with the preset temperature T and strain rate, the true stress-strain curve is obtained, and the data acquisition frequency is ≥100Hz; Use scanning electron microscope to carry out metallographic analysis on the quenched samples, determine the dynamic recrystallization fraction DRX and grain size, and the number of sampling areas is ≥5 fields of view; The local strain rate and temperature field distribution data of the deformation process are collected synchronously by laser velocimeter and infrared thermometer.

[0008] A further solution of the present invention, step S2, comprises the following steps: According to the true stress-strain curve obtained in step S1, the peak strain and the critical strain are calculated; Among them, the critical strain is the minimum strain that triggers the dynamic recrystallization fraction DRX, and the peak strain is the strain corresponding to the maximum stress point; According to the dynamic recrystallization fraction DRX obtained in step S1, the parameters K and n are measured in the laboratory.

[0009] A further solution of the present invention, step S2, further comprises the following steps: A dynamic recrystallization fraction prediction model is established to satisfy the following formula:

[0010] in, represents the dynamic recrystallization fraction, represents the kinetic coefficient, n represents the characteristic parameter reflecting the phase transition mechanism, represents the peak strain, represents the critical strain, represents the actual strain.

[0011] A further solution of the present invention, step S3, comprises the following steps: Preset dynamic recrystallization threshold , compared with the predicted and dynamic recrystallization threshold The critical strain is corrected based on the real-time strain feedback. if , maintain the current critical strain , proceed to the next step; if , triggering metallographic supplementary testing, using electron backscatter diffraction to analyze the supplementary test sample and recalibrate , update the dynamic recrystallization fraction prediction model and recalculate Predicted value.

[0012] A further solution of the present invention, step S4, comprises the following steps: tensile strength Dynamic recrystallization fraction It is related to the grain size and satisfies the following formula:

[0013] in, It represents the lattice friction stress, which indicates the inherent deformation resistance of the material when there is no grain boundary influence and is related to the purity of the material; Represents the Hall-Petch coefficient, which requires experimental calibration; The average size of the grains after recrystallization; Represents the original grain size, the average grain size before deformation.

[0014] A further solution of the present invention, step S4, further comprises the following steps: Elongation , satisfying the following formula,

[0015] in, Indicates the maximum theoretical elongation of the material, which depends on the intrinsic plasticity of the material; The coefficient that represents the influence of dynamic recrystallization on plasticity needs to be determined through data fitting.

[0016] A further solution of the present invention, step S5, comprises the following steps: Combine historical data and expert estimates to set the tensile strength threshold and elongation acceptance threshold , compare the tensile strength The qualified threshold of tensile strength Size, compare elongation and elongation acceptance threshold The size of is used to judge; If predicted and , judging that the thermal deformation performance of the high-strength titanium alloy is qualified; If predicted or , triggering the optimization of process parameters.

[0017] A further solution of the present invention, step S6, comprises the following steps: Apply a multi-objective optimization algorithm to find the temperature that gives the highest tensile strength and strain rate; The elongation must ; Parameter range: , strain rate ; Within the parameter range, 4-6 key parameter combinations are selected to cover high, medium and low change temperatures and fast, medium and slow strain rates, and the optimal solution candidate is selected.

[0018] A further solution of the present invention, step S7, comprises the following steps: Sensors are deployed inside the thermal processing equipment to collect temperature, strain rate, and pressure data in real time; Real-time data is fed into the finite element model within a specified time period to update the simulated stress-strain curve and dynamic recrystallization prediction; When the actual temperature measured by the thermal simulation test machine exceeds the range allowed by the process window, the parameter correction instruction is triggered to meet the following formula:

[0019] in, is the target temperature, To actually measure the temperature, is the original strain value before correction, The corrected strain value is used. At least 3 batches of titanium alloy are produced using the optimized parameters. The titanium alloy is randomly sampled to test the tensile strength and elongation to determine whether the thermal deformation performance of the titanium alloy is qualified. If the qualified rate is , the locking parameter is the process standard; if the qualified rate , return to step S2 to recalibrate the dynamic recrystallization fraction prediction model.

[0020] In summary, the present invention includes the following beneficial technical effects: 1. By comprehensively collecting multi-dimensional data such as the chemical composition and grain size of titanium alloys, and using advanced finite element simulation technology and dynamic recrystallization fraction prediction models, the hot deformation performance of high-strength titanium alloys is intelligently predicted. Compared with traditional experimental methods, this method can more comprehensively consider various factors affecting the hot deformation performance of titanium alloys, thus significantly improving the accuracy and reliability of prediction. Accurate prediction helps reduce material waste and production cost losses caused by unqualified performance; 2. Based on the prediction of the hot deformation performance of titanium alloys, this method can timely detect unqualified performance and automatically trigger the optimization of process parameters. By finely adjusting the process parameters, it can ensure that titanium alloy products obtain excellent mechanical properties and processing performance during the hot deformation process. This not only improves the overall quality of the products but also enhances the stability of performance, making titanium alloy products better meet the stringent requirements in practical applications; 3. It can significantly reduce manual trial and error and the number of experiments, thus reducing the time and cost in the product R & D and production processes. By quickly and accurately predicting the hot deformation performance of titanium alloys, enterprises can more quickly determine appropriate process parameters and accelerate the time to market of new products. In addition, the intelligent prediction method helps enterprises achieve lean production, improve production efficiency and resource utilization rate, further reduce costs and enhance market competitiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a schematic flow chart of an intelligent prediction method for hot deformation of high-strength titanium alloys based on finite element simulation.

[0022] Figure 2 is a schematic framework diagram of an intelligent prediction system for hot deformation of high-strength titanium alloys based on finite element simulation. DETAILED DESCRIPTION OF THE INVENTION

[0023] The present invention will be described in detail below with reference to the accompanying drawings.

[0024] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0025] The following will be a preferred and detailed description of the present invention with reference to the appended Figure 1-2 drawings.

[0026] Referring to the appended Figure 1 drawings, the present invention provides an intelligent prediction method for hot deformation of high-strength titanium alloys based on finite element simulation, including the following steps: S1. Collect data related to the chemical composition, grain size, deformation temperature, strain rate, and stress value of titanium alloy; establish a multi-dimensional input foundation to provide a complete input parameter system for subsequent models to ensure that the model covers all major influencing factors.

[0027] S2. Calculate the peak strain and critical strain based on the collected data, establish a dynamic recrystallization fraction prediction model to calculate the predicted value of the dynamic recrystallization fraction; connect the process parameters with the microstructure evolution, quantify the microstructure changes, and provide an intermediate bridge for the subsequent mechanical properties prediction.

[0028] S3. Set the dynamic recrystallization threshold, compare the predicted value of the dynamic recrystallization fraction with the dynamic recrystallization threshold, trigger the metallographic supplementary test and update the prediction model; intelligently calibrate the model reliability to avoid failure of the model due to long-term simulation error accumulation, and realize data-driven adaptability.

[0029] S4. Construct a mechanical property prediction model to calculate the values ​​of tensile strength and elongation; map from microstructure to macroscopic performance, directly link the microstructure evolution to engineering performance indicators, and provide quantifiable goals for process optimization.

[0030] S5. Set the qualified thresholds for tensile strength and elongation, compare the size of tensile strength with the qualified threshold for tensile strength, compare the size of elongation with the qualified threshold for elongation, determine whether the thermal deformation performance of titanium alloy is qualified, and optimize the process parameters if it is unqualified; realize real-time quality monitoring and process self-adjustment, replacing the traditional inefficient process that relies on manual trial and error.

[0031] S6. Based on the premise that the elongation value is qualified, a multi-objective optimization algorithm is applied to find the combination of temperature and strain rate that gives the highest tensile strength; solve the common multi-objective conflict problems in engineering and ensure the global optimality of process parameters under complex constraints.

[0032] S7. The optimized process parameters are verified through finite element simulation, and closed-loop feedback is performed based on the simulation results and actual production data to dynamically correct the model parameters; forming a closed loop of "simulation → optimization → verification → correction" to ensure the long-term applicability of the model as production conditions change.

[0033] In this solution, the dynamic recrystallization model is connected in series with the mechanical properties model, breaking through the limitations of traditional single-scale simulation and being closer to the actual material behavior. Through real-time comparison of the dynamic recrystallization score with the threshold, the directional metallographic experiment is triggered to achieve "on-demand update" of model parameters instead of fixed-cycle adjustment to avoid manual intervention lag. Prioritize the qualified elongation, and then optimize the tensile strength with a single objective to avoid multi-objective conflicts, which is more in line with the actual needs of the project. Finite element provides high-fidelity physical field data, and intelligent algorithms are responsible for decision-making optimization. The two work together to form a prediction system. Through hybrid modeling and intelligent control, high-precision prediction of the thermal deformation properties of titanium alloys and autonomous optimization of process parameters are achieved, which solves the pain points of high cost, low efficiency and insufficient precision of traditional methods, and provides a popularizable intelligent solution for complex material processing.

[0034] In one embodiment of the present invention, step S1 includes the following steps: A direct reading spectrometer is used to collect composition data of titanium alloys, including the content of elements related to Ti, Al, V, and O. The direct reading spectrometer can non-destructively detect solid samples within seconds without the need for complex chemical pretreatment, and directly output high-precision element content data, providing high-fidelity basic parameters for subsequent thermal deformation performance modeling, thus avoiding the accumulation of model errors caused by composition deviations.

[0035] For example, a direct reading spectrometer is used to collect the composition of TC4 titanium alloy, and the excitation energy of the direct reading spectrometer is 4.8 , argon flow rate is 12 , the composition of TC4 titanium alloy was measured, the Al content was 5.75%, the V content was 3.92%, the O content was 0.15%, and the rest was Ti.

[0036] Using the Gleeble thermal simulation tester, combined with the preset temperature T and strain rate, the true stress-strain curve is obtained, and the data acquisition frequency is ≥100Hz; the advantage of Gleeble is that it can accurately control the temperature and strain rate, ensuring high-fidelity dynamic response capture and process traceability. The ultra-high sampling frequency can fully record the stress fluctuation details corresponding to the evolution of the microstructure, providing high-resolution time series data for dynamic recrystallization models and mechanical property predictions.

[0037] For example, based on the Gleeble thermal simulation test machine, the sample size of TC4 titanium alloy is TC4 titanium alloy is preheated to 950 , the temperature is maintained at Keep warm within the range of 180 seconds; Combined with the preset temperature, 800 , 900 , 1000 ,1050 , and the preset strain rate, 0.001 , 0.01 , 0.1 , 1 , 10 , according to 4 groups of temperature 5 groups of strain rates For 20 groups of tests, the stress values ​​collected at constant strain rate intervals were recorded to obtain the true stress-strain curves at different temperatures T and strain rates.

[0038] Use scanning electron microscope to carry out metallographic analysis on quenched samples, determine dynamic recrystallization fraction DRX and grain size, and the number of sampling areas is ≥5 fields of view; provide statistically reliable microstructure verification data for dynamic recrystallization model and solve the accidental error problem of traditional single field of view analysis.

[0039] For example, within 1 second after the thermal deformation is completed, the alloy is put into ice-salt water for rapid cooling, an automatic grinder is used to prepare a metallographic sample, the TC4 titanium alloy is corroded by Kroll reagent for 15 seconds, and professional software is used to determine the dynamic recrystallization fraction DRX and the statistical average grain size.

[0040] For example, a laser Kepler velocimeter can be used with a sampling rate of 10 , measurement point density 1 point , collect real-time temperature data; use infrared thermal imager and Gleeble thermal simulation test machine to trigger synchronously to collect local strain rate data. Through laser velocimeter and infrared thermometer, synchronously collect local strain rate and temperature field distribution data during deformation process; synchronous monitoring of laser velocimeter + infrared thermometer realizes real-time matching of local strain rate and temperature field gradient during deformation process, provides boundary condition calibration data for finite element simulation, significantly improves the accuracy of coupling between stress-strain curve and real thermal field, and avoids model distortion caused by uniformity assumption.

[0041] In one embodiment of the present invention, step S2 comprises the following steps: According to the true stress-strain curve obtained in step S1, the peak strain and the critical strain are calculated; Among them, the critical strain is the minimum strain that triggers the dynamic recrystallization fraction DRX, and the peak strain is the strain corresponding to the maximum stress point; The peak strain is directly extracted from the high-frequency stress-strain curve of the Gleeble test, and the critical strain is determined, which provides the core input parameters for the dynamic recrystallization model and solves the subjective error problem of traditional visual interpretation.

[0042] For example, the peak strain calculated and critical strain .

[0043] According to the dynamic recrystallization fraction DRX obtained in step S1, the parameters K and n are measured in the laboratory; A dynamic recrystallization fraction prediction model is established to satisfy the following formula:

[0044] in, represents the dynamic recrystallization fraction, represents the kinetic coefficient, n represents the characteristic parameter reflecting the phase transition mechanism, represents the peak strain, represents the critical strain, represents the actual strain; Make the prediction model have physical mechanism support to avoid the risk of overfitting driven by pure data.

[0045] For example, TC4 titanium alloy was used as a sample and finite element simulation was performed using DEFORM software to obtain and ; The experimental fitting parameters of TC4 titanium alloy are obtained and ; Actual strain , brought into the dynamic recrystallization fraction prediction model; Learn that .

[0046] In one embodiment of the present invention, step S3 includes the following steps: Combine historical data and expert estimates to set the dynamic recrystallization threshold , compared with the predicted and dynamic recrystallization threshold The critical strain is corrected based on the real-time strain feedback. if , maintain the current critical strain , proceed to the next step; if , triggering metallographic supplementary testing, using electron backscatter diffraction method to analyze the supplementary test sample and recalibrate , update the dynamic recrystallization fraction prediction model and recalculate Prediction value. Automatically identify abnormal working conditions through threshold criteria, trigger retest only for unqualified samples, and reduce invalid experiments by more than 80%.

[0047] For example, the dynamic recrystallization threshold is set ; if , maintain the current critical strain ;if , triggering metallographic supplementary testing, using electron backscatter diffraction to analyze the supplementary test sample and recalibrate , among which .

[0048] In one embodiment of the present invention, step S4 comprises the following steps: tensile strength Dynamic recrystallization fraction It is related to the grain size and satisfies the following formula:

[0049] in, It represents the lattice friction stress, which indicates the inherent deformation resistance of the material without the influence of grain boundaries and is related to the purity of the material; Represents the Hall-Petch coefficient, which requires experimental calibration; The average size of the grains after recrystallization; represents the original grain size, that is, the average grain size without deformation; Elongation , satisfying the following formula,

[0050] in, Indicates the maximum theoretical elongation of the material, which depends on the intrinsic plasticity of the material; It represents the influence coefficient of dynamic recrystallization on plasticity, which needs to be determined by data fitting; For example, taking TC4 titanium alloy as a sample, assuming A classic experience value. , , ,when When the tensile strength is The formula is:

[0051] For TC4 titanium alloy as the sample, assuming , , substituting the elongation The formula is:

[0052] In one embodiment of the present invention, step S5 comprises the following steps: Combine historical data and expert estimates to set the tensile strength threshold and elongation acceptance threshold , compare the tensile strength The qualified threshold of tensile strength Size, compare elongation and elongation acceptance threshold The size of If predicted and , judging that the thermal deformation performance of the high-strength titanium alloy is qualified; If predicted or , triggering the optimization of process parameters.

[0053] As orthogonal evaluation dimensions, tensile strength and elongation cover the core requirements of material toughness and are more in line with engineering reality than a single indicator. Through hard threshold control and flexible process adjustment, the flexibility of the process window is maximized while ensuring the bottom line of quality.

[0054] For example, assuming that the tensile strength qualification threshold , elongation qualified threshold ; If predicted and , judging that the thermal deformation performance of the high-strength titanium alloy is qualified; If predicted or , triggering the optimization of process parameters.

[0055] In one embodiment of the present invention, step S6 comprises the following steps: Apply a multi-objective optimization algorithm to find the temperature that gives the highest tensile strength and strain rate; The elongation must ; Parameter range: , strain rate ; Within the parameter range, 4-6 key parameter combinations are selected to cover high, medium and low temperature changes and fast, medium and slow strain rates, and the optimal solution candidate is selected.

[0056] The elongation is used as a hard constraint to avoid the risk of brittle fracture caused by simply pursuing strength. The intelligent algorithm replaces the traditional grid search, greatly improving the calculation efficiency. The parameter combination avoids local optimality and ensures global search capability.

[0057] For example, the elongation must , Group 1 ) corresponding to , 10.5%; Group 2 ) corresponding to , 14.5%; Group 3 ( ) corresponding to , 13%; Group 4 ( ) corresponding to , 11.8%; So the optimal solution candidate: the second group ( ) corresponding to , 14.5%.

[0058] In one embodiment of the present invention, step S7 includes the following steps: Sensors are deployed inside the thermal processing equipment to collect temperature, strain rate, and pressure data in real time; Real-time data is fed into the finite element model within a specified time period to update the simulated stress-strain curve and dynamic recrystallization prediction; When the actual temperature measured by the thermal simulation test machine exceeds the range allowed by the process window, the parameter correction instruction is triggered to meet the following formula:

[0059] in, is the target temperature, To actually measure the temperature, is the original strain value before correction, is the corrected strain value. The process parameters are dynamically corrected through sensor data to avoid batch scrapping caused by traditional "post-test". The strain rate correction formula is based on physical relationships to ensure that the adjusted process still complies with the thermal deformation law of the material.

[0060] For example, a batch of thermal simulation test machines fails and the actual temperature changes from Down to (Below the lower limit of the process window ); Dynamic recrystallization fraction prediction model predicted The value of The value of is reduced to 852 MPa; Triggering strain rate compensation, we know:

[0061] Use optimized parameters to produce at least 3 batches of titanium alloys. Randomly sample the titanium alloys to test the tensile strength and elongation to determine whether the thermal deformation performance of the titanium alloy is qualified. If the qualified rate is , the locking parameter is the process standard; if the qualified rate , go back to step S2 to recalibrate the dynamic recrystallization fraction prediction model. The strict standard of 95% qualification rate for 3 batches ensures the process robustness.

[0062] See the appendix Figure 2 , the present invention also proposes an intelligent prediction system for hot deformation of high-strength titanium alloy based on finite element simulation, including: titanium alloy data acquisition module, dynamic recrystallization fraction prediction model establishment module, dynamic recrystallization threshold setting module, mechanical property prediction model construction module, qualified threshold setting module for tensile strength and elongation, process optimization module, optimization verification module; The titanium alloy data acquisition module is used to collect data related to the chemical composition, grain size, deformation temperature, strain rate, and stress value of the titanium alloy; The dynamic recrystallization fraction prediction model establishment module is used to calculate the peak strain and critical strain based on the collected data, establish a dynamic recrystallization fraction prediction model, and calculate the predicted value of the dynamic recrystallization fraction; The dynamic recrystallization threshold setting module is used to set the dynamic recrystallization threshold, compare the predicted value of the dynamic recrystallization fraction with the dynamic recrystallization threshold, trigger a metallographic supplementary test, and update the prediction model; The mechanical property prediction model construction module is used to construct a mechanical property prediction model and calculate the values of tensile strength and elongation; The qualified threshold setting module for tensile strength and elongation is used to set the qualified threshold for tensile strength and the qualified threshold for elongation, compare the tensile strength with the qualified threshold for tensile strength, compare the elongation with the qualified threshold for elongation, judge whether the hot deformation performance of the titanium alloy is qualified, and optimize the process parameters if unqualified; The process optimization module, on the premise that the elongation value is qualified, finds the combination of temperature and strain rate that maximizes the tensile strength; The optimization verification module verifies the optimized process parameters through finite element simulation, and performs closed-loop feedback based on the simulation results and actual production data to dynamically correct the model parameters.

[0063] Each of the above-mentioned modules can be implemented in whole or in part by software, hardware, and their combination, supporting being embedded in the processor of the computer device in hardware form or being independent of it, and also supporting being stored in the memory of the computer device in software form for the processor to call and execute the operations corresponding to each of the above modules.

[0064] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. The present invention extends to any new features or any new combinations disclosed in this specification, and any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention shall be included in the protection scope of the present invention. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the detailed technical features not disclosed in this embodiment, such as the specific structure, are all prior art, and those skilled in the art can obtain them from the prior art.

Claims

1. An intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation, characterized in that: The following steps are involved: S1. Collect data related to the chemical composition, grain size, deformation temperature, strain rate and stress value of titanium alloy; S2. Calculate the peak strain and critical strain according to the collected data, and establish a dynamic recrystallization fraction prediction model to calculate the dynamic recrystallization fraction prediction value; S3, setting a dynamic recrystallization threshold, comparing the predicted value of the dynamic recrystallization fraction with the dynamic recrystallization threshold, triggering a metallographic supplementary test and updating the prediction model; S4. Construct a mechanical property prediction model to calculate the values ​​of tensile strength and elongation; S5. Set a qualified threshold value for tensile strength and a qualified threshold value for elongation, compare the size of the tensile strength with the qualified threshold value for tensile strength, and compare the size of the elongation with the qualified threshold value for elongation, and judge whether the thermal deformation performance of the titanium alloy is qualified. If it is unqualified, optimize the process parameters; S6. Based on the premise that the elongation value is qualified, a multi-objective optimization algorithm is applied to find the temperature and strain rate combination that maximizes the tensile strength; S7. The optimized process parameters are verified by finite element simulation, and closed-loop feedback is performed based on the simulation results and actual production data to dynamically correct the model parameters.

2. According to claim 1, a high-strength titanium alloy thermal deformation intelligent prediction method based on finite element simulation is characterized in that: Step S1 includes the following steps: S11. Collect the composition data of titanium alloy using direct reading spectrometer, including the content of elements related to Ti, Al, V, and O; S12. Using the Gleeble thermal simulation test machine, combined with the preset temperature T and strain rate, obtain the true stress-strain curve, and the data acquisition frequency is ≥100Hz; S13. Use a scanning electron microscope to perform metallographic analysis on the quenched sample to determine the dynamic recrystallization fraction DRX and grain size. The number of sampling areas is ≥ 5 fields of view. S14. The local strain rate and temperature field distribution data of the deformation process are synchronously collected through a laser velocimeter and an infrared thermometer.

3. The intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation according to claim 2 is characterized in that: Step S2 includes the following steps: S21, calculating the peak strain and the critical strain according to the true stress-strain curve obtained in step S1; Among them, the critical strain is the minimum strain that triggers the dynamic recrystallization fraction DRX, and the peak strain is the strain corresponding to the maximum stress point; S22. According to the dynamic recrystallization fraction DRX obtained in step S1, parameters K and n are measured in the laboratory.

4. The intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation according to claim 3 is characterized in that: Step S2 further includes the following steps: S23, establish a dynamic recrystallization fraction prediction model, satisfying the following formula, in, represents the dynamic recrystallization fraction, represents the kinetic coefficient, n represents the characteristic parameter reflecting the phase transition mechanism, represents the peak strain, represents the critical strain, represents the actual strain.

5. The intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation according to claim 4 is characterized in that: Step S3 includes the following steps: S31. Presetting the dynamic recrystallization threshold , compared with the predicted and dynamic recrystallization threshold The critical strain is corrected based on the real-time strain feedback. if , maintain the current critical strain , proceed to the next step; if , triggering metallographic supplementary testing, using electron backscatter diffraction to analyze the supplementary test sample and recalibrate , update the dynamic recrystallization fraction prediction model and recalculate Predicted value.

6. The intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation according to claim 5 is characterized in that: Step S4 includes the following steps: tensile strength Dynamic recrystallization fraction It is related to the grain size and satisfies the following formula: in, It represents the lattice friction stress, which indicates the inherent deformation resistance of the material without the influence of grain boundaries and is related to the purity of the material; Represents the Hall-Petch coefficient, which requires experimental calibration; The average size of the grains after recrystallization; Represents the original grain size, the average grain size before deformation.

7. The intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation according to claim 6 is characterized in that: Step S4 further includes the following steps: Elongation , satisfying the following formula, in, Indicates the maximum theoretical elongation of the material, which depends on the intrinsic plasticity of the material; The coefficient that represents the influence of dynamic recrystallization on plasticity needs to be determined through data fitting.

8. The intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation according to claim 7 is characterized in that: Step S5 includes the following steps: Setting the tensile strength threshold and elongation acceptance threshold , compare the tensile strength The qualified threshold of tensile strength Size, compare elongation and elongation acceptance threshold The size of If predicted and , judging that the thermal deformation performance of the high-strength titanium alloy is qualified; If predicted or , triggering the optimization of process parameters.

9. The intelligent prediction method for thermal deformation of high-strength titanium alloy based on finite element simulation according to claim 8 is characterized in that: Step S6 includes the following steps: Apply a multi-objective optimization algorithm to find the temperature that gives the highest tensile strength and strain rate; Elongation ; Parameter range: , strain rate ; Within the parameter range, 4-6 key parameter combinations are selected to cover high, medium and low change temperatures and fast, medium and slow strain rates, and the optimal solution candidate is selected.

10. The intelligent prediction method for hot deformation of high-strength titanium alloy based on finite element simulation according to claim 9 is characterized in that: Step S7 includes the following steps: Sensors are deployed inside the thermal processing equipment to collect temperature, strain rate, and pressure data in real time; Real-time data is fed into the finite element model within a specified time period to update the simulated stress-strain curve and dynamic recrystallization prediction; When the actual temperature measured by the thermal simulation test machine exceeds the range allowed by the process window, the parameter correction instruction is triggered to meet the following formula: in, is the target temperature, To actually measure the temperature, is the original strain value before correction, is the corrected strain value; Use optimized parameters to produce at least 3 batches of titanium alloys, randomly sample the titanium alloys to test the tensile strength and elongation, and determine whether the thermal deformation performance of the titanium alloy is qualified. If the qualified rate , the locking parameter is the process standard; if the qualified rate , return to step S2 to recalibrate the dynamic recrystallization fraction prediction model.

Images