Intelligent regulation and control method for microstructure in double-phase titanium alloy hot spinning forming process

Abstract

The invention provides an intelligent regulation and control method for a microstructure in a double-phase titanium alloy hot spinning forming process. The method comprises the following steps: designing an orthogonal experiment, researching internal relations among an initial phase, a phase volume fraction, process parameters and the microstructure in a constant strain rate deformation process and a variable strain rate deformation process, and constructing a microstructure prediction model; establishing a unified physical mechanism constitutive model related to the forming process, and determining the incidence relation among the rheological behavior, the deformation parameter and the microstructure; a microstructure prediction model in the double-phase titanium alloy hot spinning forming process is further constructed in an integrated mode, and the mapping relation among spinning parameters, the initial phase and the phase volume fraction and the space-time evolution rules of phases in the component is determined; and in combination with a machine learning algorithm, outputting the optimal process parameter range according to the deviation between the microstructure predicted value of the component and the target quantity in the spinning forming process. The problem that the microstructure is difficult to regulate and control in the hot spinning forming process of the double-phase titanium alloy can be solved.

Description

Technical Field

[0001] This invention relates to the field of spinning technology, and in particular to a method for intelligent control of the microstructure of titanium alloy spinning. Background Technology

[0002] With the increasing demands for component performance in the aerospace industry, titanium alloy thin-walled cylindrical parts, characterized by their lightweight and high strength, have gradually become an important material in the aerospace sector. Spin forming, as a continuous localized plastic forming process, is widely used in the manufacture of titanium alloy thin-walled cylindrical parts due to its advantages of high process flexibility, simple mold structure, low forming load, and high production efficiency. However, the microstructure and mechanical properties of titanium alloys after forming are affected by various process parameters, such as deformation temperature and strain rate, which is detrimental to the high-quality manufacturing of titanium alloy parts. Therefore, establishing a predictive model between process parameters, microstructure, and mechanical properties is of great significance for improving product quality and reducing production costs.

[0003] Addressing the optimization of spinning process parameters, titanium alloys, with their excellent specific strength and corrosion resistance, occupy a core position in high-end equipment fields such as aerospace. Hot spinning is a key manufacturing technology for their complex rotating components. The service performance of components is directly determined by their microstructure, and the synergistic effect of multiple process parameters during hot spinning significantly alters the microstructure evolution path. Therefore, the precise matching of process parameters with microstructure and properties becomes the core technology.

[0004] The optimization of hot spinning process parameters for titanium alloys still faces many unresolved issues. First, the coupling relationship between process parameters and microstructure is unclear. Existing research often focuses on the impact of a single parameter on macroscopic properties, neglecting the complex changes in microscopic mechanisms such as dynamic recrystallization and grain boundary evolution under the interaction of multiple parameters like temperature, strain, and strain rate, making it difficult to establish quantitative correlation models. Second, parameter optimization is often arbitrary, relying heavily on experience or trial-and-error methods to determine the process window, leading to poor microstructure uniformity and problems such as uneven grain size and disordered orientation distribution. Third, performance control precision is insufficient. Existing methods cannot reverse-match process parameters to target performance requirements, often resulting in an imbalance between strength and toughness in components, making it difficult to meet the stringent service requirements of high-end equipment.

[0005] This patent provides an effective solution to the aforementioned problems. By revealing the evolution mechanism of microstructure under the coupling of multiple process parameters, it establishes a precise correlation model between process parameters, microstructure, and performance, breaking through the limitations of traditional research. Based on this model, targeted optimization of process parameters can be achieved, avoiding the blindness of empirical control and significantly improving the uniformity of microstructure. Simultaneously, this method supports reverse design of process schemes based on target performance, achieving precise matching between microstructure and service performance, greatly reducing trial-and-error costs, providing reliable technical support for the high-performance manufacturing of titanium alloy hot-spun components, and promoting the development of titanium alloy forming technology towards precision and efficiency. Summary of the Invention

[0006] The purpose of this invention is to address the shortcomings of the aforementioned background technology by providing a solution that solves the problem of the lack of intelligent control over the microstructure in the existing titanium alloy spinning process, thereby achieving intelligent control over the microstructure during the titanium alloy spinning process.

[0007] To achieve the above objectives, this invention provides a method for intelligent control of the microstructure during the hot spinning process of titanium alloys, comprising the following steps:

[0008] S1. Design an orthogonal experiment to study the influence of deformation parameters on microstructure in compression test, obtain the variation law of α phase and β phase content, and thus obtain the evolution law of α phase and β phase during spinning process;

[0009] S2, based on the results of orthogonal experiments, establishes a rheological stress constitutive model through thermal stress equation and non-thermal stress equation, thereby establishing a microstructure evolution model that considers mechanisms such as α / β phase transition, dynamic recrystallization, and dynamic spheroidization;

[0010] S3. Establish a digital simulation model for titanium alloy spinning forming, and couple the constitutive model established in S2 with this model to realize intelligent control of microstructure evolution by adjusting process parameters during spinning.

[0011] S4, based on experimental data from digital simulation, uses optimization algorithms to achieve real-time control of process parameters, ultimately outputting the optimal combination of process parameters to achieve intelligent control of microstructure.

[0012] Furthermore, the hot spinning deformation parameters include deformation temperature, deformation rate, etc., and the process parameters do not change during the hot spinning forming process;

[0013] The microstructure evolution models include the coarsening / dissolution kinetics model of the primary equiaxed α phase, the lamellar α spheroidization kinetics model, the β matrix subgrain size evolution model, the β matrix grain boundary orientation difference angular distribution model, the β matrix continuous dynamic recrystallization model, and the β matrix average dislocation density evolution model. These models are used to predict the microstructure evolution under certain deformation parameters.

[0014] Furthermore, based on the thermal stress equation and the non-thermal stress equation, the rheological stress constitutive equation is established in S2, as follows:

[0015]

[0016]

[0017]

[0018] in: , as well as Here, R is the material constant, R is the ideal gas constant, and the strain rate is... , where f is the attack rate and n is the rotation speed.

[0019] As can be seen from the above equations, the magnitude of rheological stress is affected by microstructure parameters such as dislocation density, grain size, and α phase content. These microstructure parameters also influence each other and have complex evolution mechanisms. They are highly sensitive to deformation parameters, such as α / β phase transformation, dynamic recrystallization, and dynamic spheroidization. Therefore, it is necessary to establish a microstructure evolution model that considers the above mechanisms to predict the microstructure during deformation and thus predict the rheological stress during deformation.

[0020] Furthermore, an isothermal phase transformation kinetic model is established. The phase transformation of TC18 titanium alloy is a solid-state phase transformation, which can usually be described by the JMAK equation. In this invention, the stress-induced dynamic phase transformation during deformation is ignored, and only the static phase transformation is considered. The content of secondary α phase precipitation and primary α phase is calculated using the following formula:

[0021]

[0022] In the formula, t represents the volume fraction of the secondary α-structure after time t; m is a constant describing the nucleation and growth mechanism of TC18 titanium alloy; t is the holding time, including the heating and holding process before deformation (uniformly 300s) and the deformation process (related to strain rate). ref For reference time, and to better fit the experimental results, 0.001s is used here. -1 The total time for the strain rate to deform to a true strain of 0.92 was 1220 s. The temperature at which the secondary α phase completely integrates into the matrix is ​​1093 K for this alloy; k is a temperature-dependent constant representing the reaction phase transformation rate, calculated using equation (4-7):

[0023]

[0024] Where A is the prefactor (intrinsic material parameter), Q is the phase transformation / recrystallization activation energy determined by the titanium alloy composition and needs to be calibrated through thermal simulation experiments, R is the gas constant, and T is the spinning forming temperature. For equivalent change, denoted as the equivalent strain rate, and m and n as strain / strain rate sensitivity coefficients.

[0025]

[0026] Furthermore, the genetic algorithm (GA) built into MATLAB software was used to analyze the material parameters and spinning deformation parameters (including k0, Q, T, ...). - By optimizing the values ​​of f, n, etc., we can obtain functional expressions for the contents of primary and secondary α phases, thereby calculating the contents of β phase and obtaining predicted values ​​of the relationship between β phase contents and temperature at different temperatures. .

[0027] Furthermore, after obtaining the β phase content at different temperatures, considering that the α phase precipitates in a layered form on the β phase matrix, we also need to consider the spheroidization of the layered α phase, the evolution of β phase subgrains, the orientation of β phase grain boundaries, the continuous dynamic recrystallization of the β phase matrix, and the evolution of the β matrix dislocation density. Models are established for each of these considerations, and the content of the spheroidized α phase can be obtained accordingly.

[0028]

[0029] The evolution rate formula for the spheroidization fraction of the lamellar α phase is:

[0030]

[0031] Where S is the spheroidization ratio of the lamellar α phase; is the hot spin compression strain rate; c0, c1 are material constants; M bm denoted as grain boundary mobility; P is the driving force for grain boundary migration.

[0032] The fraction of dislocation cells in LAGBs:

[0033]

[0034] Grain boundary orientation difference angular distribution of the β matrix:

[0035]

[0036] in Calculated using the following formula:

[0037]

[0038] Where λ is the scale parameter, and is related to the average orientation difference angle. Related, and The orientation difference is determined by both "deformation-induced dislocation accumulation" and "recrystallization-induced high-angle grain boundary orientation difference," and the formula is derived based on thermal activation theory.

[0039]

[0040] in Let be the hot spin strain rate, a be the strain sensitivity coefficient, and b be the strain rate sensitivity coefficient. The average orientation difference of the high-angle grain boundaries formed by DRX.

[0041] β Total grain boundary density variation in the matrix:

[0042]

[0043] Changes in the overall average dislocation density of the β matrix:

[0044]

[0045] Where f HAGB It is the core functional parameter of orientation difference distribution (directly affecting the strength, toughness, and corrosion resistance of the β matrix), and is related to the recrystallization fraction X. DRX Strong correlation, derived using the Avrami recrystallization kinetic equation:

[0046]

[0047] Where f HAGB，0 To achieve a high grain boundary ratio, the annealed β-titanium alloy before spinning contains approximately 10%~20% k. DRX Let p be the recrystallization kinetic constant, q be the strain influence index, and Q be the strain rate influence index. DRX It is a dynamic recrystallization active energy.

[0048] Furthermore, by combining the sub-models established above, a prediction model for deformation parameters, microstructure, and rheological stress of the solid solution-treated TC18 titanium alloy is finally established.

[0049] Furthermore, ABAQUS is a powerful finite element analysis software widely used in numerical simulation analysis across various engineering fields. This patent utilizes ABAQUS software to build a simulation model of TC18 titanium alloy hot spinning forming. Since hot spinning forming is a dynamic problem, the explicit solver ABAQUS / Explicit is used for calculation and solution. Finite element modeling of the hot spinning forming process is performed in S3, with the following conditions input: motion mode and boundary conditions, contact conditions and friction model, heat source, heat transfer and thermal boundary conditions, and material parameter model; Motion mode: Typically, the mandrel is fixedly mounted on the machine tool spindle, and the sheet metal is clamped in the head of the mandrel by the tailstock, both rotating at the same speed as the spindle; Contact conditions: The lower surface of the sheet metal and the outer surface of the mandrel are set to face-to-face contact, and the upper surface of the sheet metal and the outer surface of the spinning wheel are set to face-to-face contact; Thermal boundary conditions: Setting a reasonable heat source, heat transfer, and thermal boundary conditions makes the simulation more convincing.

[0050] Furthermore, the microstructure evolution model established in S2 is input into ABAQUS and coupled with the finite element model established in S3, thereby realizing intelligent control of microstructure evolution by adjusting process parameters during spinning.

[0051] Furthermore, the core logic of intelligent microstructure control in S4 through the introduction of machine learning is as follows: based on spinning simulation data, a mapping model of "process parameters-microstructure-performance" is constructed, and the optimal process is output by combining real-time control and optimization. The core process can be summarized into four stages:

[0052] Data layer: Extract spinning process parameters (temperature, rotation speed, etc.) from simulation data as input, and microstructure (α / β phase content, etc.) and performance indicators (hardness, etc.) as output. After preprocessing and enhancement, the dataset is divided into training / validation / test sets to construct a high-quality dataset.

[0053] Model layer: A dual "prediction-optimization" model is constructed. A prediction model is built using a fusion model of XGBoost and CNN-LSTM to accurately map the nonlinear relationship between process technology and tissue performance. Based on the prediction model, optimization models such as genetic algorithms are combined to locate the optimal range of process parameters.

[0054] Control layer: Establish a real-time control mechanism to link the optimization model with the spinning production system, collect production data in real time and feed it back to the model, dynamically correct process parameters, and ensure timely and accurate control.

[0055] Validation layer: The model accuracy is verified through test sets and physical experiments. If the performance of the tissue does not meet the target, the dataset and model parameters are iteratively optimized until the optimal combination of process parameters that meets the requirements is output, so as to achieve intelligent control.

[0056] The above-described solution of the present invention has the following beneficial effects:

[0057] The intelligent control method for microstructure during the hot spinning forming process of titanium alloy provided by this invention can achieve intelligent control of microstructure during the hot spinning forming process of TC18 titanium alloy by adjusting process parameters. It is applicable to the optimization of process parameters when processing titanium alloy cylindrical parts in complex situations, and finds the best combination of process parameters to optimize the overall quality of titanium alloy cylindrical rotating parts.

[0058] The deep reinforcement learning framework used in this invention is unsupervised learning, which does not require a large amount of predefined label data. The spinning forming process parameter control model can obtain a large amount of data by interacting with the environment, reducing the cost of pre-acquiring data.

[0059] This invention employs a constitutive model to accurately describe the mechanical response of materials, supporting finite element simulation; and establishes a microstructure model to predict the microstructure of titanium alloys, greatly improving the efficiency of titanium alloy spinning process optimization and cost reduction.

[0060] The machine learning employed in this invention can efficiently process simulation and experimental data of titanium alloy hot spinning, accurately constructing a mapping model of "process parameters-microstructure-performance". It can rapidly iteratively optimize parameters such as temperature and rotation speed, reducing the cost of trial and error and extending the R&D cycle. Simultaneously, it can respond to production fluctuations in real time, dynamically correcting parameters to ensure consistent product microstructure and performance, driving the process from experience-driven to data-driven, and improving optimization efficiency and forming quality.

[0061] Other beneficial effects of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0062] Figure 1 This is a flowchart of the method steps of the present invention;

[0063] Figure 2 A flowchart for establishing a constitutive model for a unified physical mechanism. Detailed Implementation

[0064] The following specific examples illustrate the implementation of this disclosure. Those skilled in the art can easily understand other advantages and effects of this disclosure from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of this disclosure, and not all of them. This disclosure can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of this disclosure. It should be noted that, in the absence of conflict, the following embodiments and features in the embodiments can be combined with each other. Based on the embodiments in this disclosure, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this disclosure.

[0065] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using structures and / or functionalities other than one or more of the aspects set forth herein.

[0066] It should also be noted that the illustrations provided in the following embodiments are merely schematic representations of the basic concept of this disclosure. The illustrations only show components relevant to this disclosure and are not drawn according to the actual number, shape, and size of components in implementation. In actual implementation, the type, quantity, and proportion of each component can be arbitrarily changed, and the component layout may be more complex. Furthermore, specific details are provided in the following description to facilitate a thorough understanding of the examples. However, those skilled in the art will understand that the described aspects can be practiced without these specific details.

[0067] like Figure 1 As shown, an embodiment of the present invention provides a method for intelligent control of microstructure during the hot spinning process of titanium alloys, comprising the following steps:

[0068] S1. Design an orthogonal experiment to study the influence of deformation parameters on microstructure in compression tests, obtain the variation law of α phase and β phase content, and thus obtain the evolution law of α phase and β phase during spinning.

[0069] The hot spinning deformation parameters include deformation temperature and deformation rate, and these process parameters remain unchanged during the hot spinning process. In this embodiment, the macroscopic rheological behavior of the two-phase region of the solid solution TC18 titanium alloy was studied through orthogonal experiments. The influence of deformation parameters on the material's rheological behavior was analyzed, and a prediction model and hot working diagram based on long and short memory networks were established. At the same time, the hot deformation process was preliminarily optimized from the perspective of the hot working diagram.

[0070] In the orthogonal experiment, specimens were cut from the blank, and a thermocouple was installed by drilling a hole in the center of the specimen to measure and control the temperature. A Gleeble-3500 testing machine was used for the experiment. A tantalum sheet was placed between the specimen and the indenter, and lubricant was applied to reduce friction. Before compression, the specimen was first heated to the set experimental temperature and held at that temperature. After deformation, the specimen was immediately water-cooled to preserve the deformed microstructure. After the experiment, the microstructure at the center of the specimen was observed.

[0071] S2, based on the results of orthogonal experiments, establishes a rheological stress constitutive model through thermal stress equations and non-thermal stress equations, thereby establishing a microstructure evolution model that considers mechanisms such as α / β phase transition, dynamic recrystallization, and dynamic spheroidization.

[0072] Because this involves the establishment of numerous microstructure evolution models, this patent describes the process of establishing the lamellar α spheroidization kinetic model: during deformation, the secondary α phase precipitates in the form of lamellae on the β matrix. For lamellar α phases, dynamic spheroidization is the most important form of microstructure evolution. Dynamic spheroidization mainly occurs through grain boundary migration, a process that consumes dislocation density, leading to a reduction in flow stress. Therefore, the modeling of the dynamic spheroidization process can refer to the dynamic recrystallization process. Research has found that the spheroidization rate of the lamellar α phase is related to the strain rate, deformation temperature, and thickness of the lamellar α phase. Therefore, a formula for the evolution rate of the spheroidization fraction of the lamellar α phase is given:

[0073]

[0074] In the formula: S is the spheroidization ratio of the lamellar α phase; ε is the strain rate; c0 and c1 are material constants; M bm denoted as grain boundary mobility; P is the driving force for grain boundary migration.

[0075]

[0076] In the formula, δ is the grain boundary thickness; D ob It is the boundary self-diffusion coefficient; Q diff For the boundary diffusion activation energy, k b It is the Boltzmann constant.

[0077]

[0078] The reduction in dislocation density caused by lamellar α-spheroidization can be expressed as:

[0079]

[0080] After considering spheroidization, the content of lamellar α phase and spheroidized α phase during deformation can be calculated using the following formula:

[0081]

[0082] S3. Establish a digital simulation model for titanium alloy spinning forming. Couple the constitutive model established in S2 with this model to achieve intelligent control of microstructure evolution by adjusting process parameters during spinning.

[0083] A simulation model of TC18 titanium alloy hot spinning forming was built using ABAQUS software. Since hot spinning forming is a dynamic problem, the explicit solver ABAQUS / Explicit was used for calculation and solution. Finite element modeling of the hot spinning forming process was performed in S3, with the following conditions input: motion mode and boundary conditions, contact conditions and friction model, heat source, heat transfer and thermal boundary conditions, and material parameter model. Motion mode: Typically, the mandrel is fixedly mounted on the machine tool spindle, and the sheet metal is clamped in the head of the mandrel by the tailstock, both rotating at the same speed as the spindle. Contact conditions: The lower surface of the sheet metal and the outer surface of the mandrel are set to face-to-face contact, and the upper surface of the sheet metal and the outer surface of the spinning wheel are set to face-to-face contact. Thermal boundary conditions: Setting a reasonable heat source, heat transfer, and thermal boundary conditions makes the simulation more convincing.

[0084] Key technologies for finite element simulation of hot spinning forming of titanium alloys:

[0085] Mesh Adaptive Control: The hot spin forming of tapered parts with internal ribs involves large plastic deformation. In simulations, the sheet metal mesh often suffers significant distortion, which can severely disrupt model calculations and affect the accuracy of the results. Therefore, mesh adaptive technology is used to address this issue. ABAQUS software incorporates two mesh adaptive techniques: Arbitrary Lagrange-Eulerian Adaptive Mesh (ALE) and Adaptive Mesh Re-division (AM). For solving dynamic explicit problems using ABAQUS / Explicit in this paper, ALE adaptive mesh is more suitable.

[0086] Mass scaling factor: For large plastic deformation problems, using ALE adaptive meshes can slow down the computation. A mass scaling factor is typically used to improve the model's computational efficiency. Generally, the mass scaling factor is only suitable for quasi-static problems, and setting the factor too high can easily distort the model's calculation results and cause errors. In this paper, the mass scaling factor for the finite element thermal spin simulation model of the internally ribbed conical component is set to 600. The compliance with the quasi-static principle is verified by ensuring that the ratio of kinetic energy to internal energy in the calculated results is sufficiently small.

[0087] Restart Settings: Restart settings are an important function in ABAQUS software. They are used to resume calculations from the previous step and continue calculations when simulations are unexpectedly interrupted. This is especially important in large-scale engineering simulation models, where restart settings maximize resource utilization. The thermal rotation simulation model of the internally ribbed tapered component created in this paper is computationally complex and slow; therefore, restart analysis is necessary to prevent unexpected situations.

[0088] Hourglass Control: Hourglass mode refers to a non-physical zero-energy deformation mode, where deformation occurs in the model without stress or strain. The presence of hourglass mode reduces the computational accuracy of the model, rendering the calculation results invalid; therefore, it needs to be controlled to avoid this phenomenon. Mesh refinement and setting hourglass control options are two commonly used methods in ABAQUS software. The process of refining the mesh for sheet metal has been described in detail and will not be repeated here. This paper limits the expansion of hourglass mode by introducing a small amount of artificial "hourglass stiffness," and the hourglass control option is set to enhanced in the simulation. After the simulation is completed, the result is verified by ensuring that the ratio of pseudo-strain energy to internal energy is sufficiently small.

[0089] S4, based on experimental data from digital simulation, uses optimization algorithms to achieve real-time control of process parameters, ultimately outputting the optimal combination of process parameters to achieve intelligent control of microstructure.

[0090] Based on the S3-based digital simulation model for titanium alloy spinning, the system extracts core parameters and result data from the simulation process, constructs an "input-output" variable matrix, and ensures that the variables cover the key influencing factors and core microstructure indicators of the entire spinning process.

[0091] Input variables: Select process parameters that play a decisive role in the microstructure evolution during the spinning process, including basic process parameters, constraint parameters, and initial state parameters (initial microstructure uniformity of billet, initial α phase / β phase content ratio), totaling 10 key parameters in 3 categories, forming the input feature vector. .

[0092] Output variables: With intelligent control as the goal, quantifiable core indicators are selected, including microstructure indicators (volume fraction of primary α phase, volume fraction of secondary α phase, volume fraction of β phase, β phase grain size, volume fraction of dynamic recrystallization, spheroidization rate, etc.) and mechanical property indicators, totaling 2 categories and 9 indicators, forming the output target vector. .

[0093] By calling the simulation software's API interface using a Python script, simulation data under multiple combinations of different process parameters is automatically extracted in batches, generating raw dataset CSV files. After generating the raw dataset, the first step is to verify data quality. The Pandas library in Python is used to statistically analyze the missing rate, numerical range, and distribution characteristics of each column of data. After confirming that there are no obvious formatting errors, the data proceeds to the preprocessing stage.

[0094] First, address data defects: use multiple interpolation to fill in sample data with a missing rate of <10%, and directly remove samples with a missing rate exceeding the standard; combine the 3σ criterion and box plots to identify and delete abnormal data such as abnormal grain size and sudden changes in performance indicators, ensuring that the data conforms to the actual spinning law.

[0095] Then, standardization and feature optimization are performed: Z-score standardization is used to eliminate the influence of dimensions for the input process parameters, and Min-Max normalization is used to normalize the output indicators such as α / β phase content to the [0,1] interval; redundant parameters with correlation >0.85, such as core mold rotation speed and spool linear velocity, are removed by using Pearson correlation coefficient matrix and random forest feature importance score, and 7 core input features are retained.

[0096] Finally, data augmentation and partitioning were completed: the SMOTE algorithm was used to supplement the scarce process range samples such as high temperature and high speed; stratified sampling was used to create training, validation and test sets to ensure that the data distribution of each set is consistent and to provide high-quality data support for subsequent model training.

[0097] In summary, the intelligent control and optimization method for the microstructure during the hot spinning process of titanium alloys provided in this embodiment can dynamically adjust the process parameters during spinning based on factors such as the ideal deformation temperature, strain rate, and strain of the part. It is suitable for optimizing the process parameters of spinning parts with complex generatrices and non-axisymmetric features, finding the best process path for customizing and optimizing the overall quality of parts, and achieving high accuracy and reliability.

[0098] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0099] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for intelligent control of microstructure during hot spinning of dual-phase titanium alloys, characterized in that, Includes the following steps: S1, through orthogonal experimental design, the thermal deformation parameters are determined, and the initial values ​​during deformation at constant strain rate and variable strain rate are obtained. Harmony Phase volume fraction, process parameters, and microstructure evolution laws are used to construct a microstructure prediction model, which includes... Phase transition model Phase dynamic recrystallization kinetic model, Phase dynamic spheroidization model; S2, combined with the microstructure prediction model constructed in S1, establish a unified physical mechanism constitutive model related to the forming process, and determine deformation parameters and initial... Harmony Phase volume fraction, Mutually, The spatiotemporal characteristics and intrinsic relationship between phases and rheological stress; S3, an integrated model for predicting the microstructure during the hot spinning process of dual-phase titanium alloys was constructed to determine the spinning process parameters and initial... Harmony Phase volume fraction, forming process components Harmony The mapping relationship between the spatiotemporal evolution laws of phases; S4, combining the models established in S2 and S3 to construct a machine learning framework, based on the initial structure of the component during the spinning process. Harmony The deviation between the predicted value and the target value of the equal microstructure is used to optimize the spinning process parameters by combining machine learning algorithms, and finally output the optimal process parameter range.

2. The intelligent control method for microstructure during hot spinning of dual-phase titanium alloys according to claim 1, characterized in that, The parameters of thermal deformation include deformation temperature, constant strain rate, stepped strain rate, and deformation amount; Hot spinning parameters include thinning rate, spindle speed, spinning wheel feed rate, and spinning temperature; microstructure includes... Mutually, Mutually, Mutually, Phase, dislocation, subgrain.

3. The intelligent control method for microstructure during hot spinning of dual-phase titanium alloys according to claim 1, characterized in that, Built in S1 The phase transition model is ,in, For the current moment Matrix content, It is the initial isoaxial shape after time t. Volume fraction of phase content It is secondary after time t Phase transition volume fraction of the tissue They are primary isometric Volume fraction of phase content and secondary The proportion of phase transition volume fraction in the tissue; The phase dynamic recrystallization kinetic model is as follows: ,in The thickness is the grain boundary. It is the boundary self-diffusion coefficient; For boundary diffusion activation energy, It is the Boltzmann constant; It is an optimizable rate constant; The phase dynamic spheroidization model is ,in, For the current moment of sphericity Phase volume fraction, S is the number of layers Phase spheroidization ratio, For strain rate, For material constants, These are free dislocations inside the grain.

4. The intelligent control method for microstructure during the hot spinning forming process of titanium alloys according to claim 1, characterized in that, The constitutive equation for rheological stress is expressed as follows: , in: , as well as Here, R is the material constant, R is the ideal gas constant, and the strain rate is... , where f is the attack rate and n is the rotation speed.

5. The intelligent control method for microstructure during the hot spinning forming process of titanium alloys according to claim 1, characterized in that, Establish an expression for the β-phase content: in, t represents the volume fraction of the secondary α-structure after time t; m is a constant describing the nucleation and growth mechanism of TC18 titanium alloy; t is the holding time, including the heating and holding process before deformation (uniformly 300s) and the deformation process (related to strain rate). ref For reference, and to better fit the experimental results, the total time from deformation at a strain rate of 0.001 s⁻¹ to the true strain of 0.92 is taken as 1220 s. The temperature at which the secondary α phase completely integrates into the matrix is ​​1093 K for this alloy; k is a temperature-dependent constant representing the reaction phase transformation rate, calculated using the following formula: Where A is the prefactor (intrinsic material parameter), Q is the phase transformation / recrystallization activation energy determined by the titanium alloy composition and requires verification and calibration, R is the gas constant, and T is the spinning forming temperature. For equivalent change, denoted as the equivalent strain rate, and m and n as strain / strain rate sensitivity coefficients.

6. The intelligent control method for microstructure during the hot spinning forming process of titanium alloys according to claim 1, characterized in that, By establishing a model equation for the evolution of microstructure, the content of spheroidized α phase can be obtained: The fraction of dislocation cells in LAGBs: Grain boundary orientation difference distribution of the β matrix: ; Total grain boundary density variation in the β matrix: Changes in the overall average dislocation density of the β matrix: .

7. The intelligent control method for microstructure during hot spinning of dual-phase titanium alloys according to claim 1, characterized in that, The unified physical mechanism constitutive model constructed in S2 is as follows: in , , , , ... , , , , , , , , .

Images