Method for predicting mixed crystal grade in hot forging discontinuous deformation process of large forge piece

Abstract

A method for predicting the mixed crystal grade in the hot forging discontinuous deformation process of a large forge piece belongs to the technical field of material forming simulation, a grain structure evaluation port is moved forward to have the characteristics of defect prediction and process optimization, and the mixed crystal defect in the hot forging process of the large forge piece is favorably controlled. Comprising the following steps: S1, establishing a microstructure cellular automaton model in the hot forging discontinuous deformation process; s2, carrying out macro-micro multi-scale FE-CA coupling on the hot forging discontinuous deformation; wherein the step S1 sequentially comprises the following steps: establishing an evolution model of a microstructure in a dynamic recrystallization process, establishing an evolution model of cellular dislocation density in a static recovery process, determining a static recrystallization nucleation rate, establishing an evolution model of a microstructure in a sub-dynamic recrystallization process, coarsening crystal grains, and establishing a microstructure CA model; and the step S2 comprises acquisition of node physical field quantity, data processing and a simulation process.

Description

Technical Field

[0001] This invention belongs to the field of material forming simulation technology, specifically relating to a method for predicting the mixed crystal level in the discontinuous deformation process of hot forging of large forgings. Background Technology

[0002] The hot forging process for large forgings is complex, and the shapes and sizes of forgings vary greatly. The rationality of the hot forging process parameters is a key factor determining the success or failure of the process. Inappropriate process parameters often lead to macroscopic problems such as difficulty in metal flow and inability to fill the mold cavity smoothly, as well as microscopic defects such as cracks, coarse grains, and mixed grains. These macroscopic defects, inherent in the size of large forgings, are often easy to observe, but their microscopic structural defects are difficult to observe and control in a timely manner, posing challenges to crack initiation and the control of grain refinement and homogenization.

[0003] The finite element method (FEM) is widely used as a convenient visualization tool for simulating the forging and heat treatment processes of large forgings. While FEM simulation can obtain information on field quantities such as temperature, strain, and stress fields during hot working, it cannot reproduce the dynamic processes of grain nucleation, growth, and post-forging grain microstructure evolution during hot forging. Furthermore, it cannot predict the mixed-grain microstructure morphology during hot forging through macroscopic FEM simulation. Compared to macroscopic FEM, cellular automata (CA) have the advantage of quantitatively, continuously, and dynamically reproducing the microstructure morphology during recrystallization and phase transformation in revealing the intrinsic mechanisms and non-uniform distribution of microstructure evolution. However, CA cannot couple the macroscopic field quantity information generated by high-temperature deformation with the microstructure evolution.

[0004] Given that material scales typically span multiple orders of magnitude, and different numerical simulation methods correspond to only one or a few orders of magnitude, multi-scale integrated simulation techniques are imperative. In recent years, Chen et al. developed a cellular automata (CA) model combining topological deformation techniques to simulate the microstructure evolution of 30Cr2Ni4MoV rotor steel during high-temperature austenitization and dynamic recrystallization (DRX) processes. Duan et al. used a finite element method coupled with CA to study the relationship between deformation parameters and microstructure, dynamically reflecting changes in microstructure during finite element simulations. Jiang et al., by coupling a cellular automata (CA) model with a finite element (FE) model, studied the physical field and dynamic recrystallization (DRX) evolution of thin-walled magnesium (Mg) alloy wheels during reverse extrusion (BE) processes, and verified the simulation results through industrial experiments. Pourian et al. used the finite element method and CA model to discuss the microstructure of hexagonal close-packed titanium polycrystalline materials, particularly the influence of neighboring grains on the behavior of the central grain. Kwiecien et al. integrated the finite element and CA methods to simulate grain refinement behavior during composite metal forming processes. Currently, the multi-scale integrated simulation technology involved in existing reports is mostly focused on the continuous deformation analysis of small rolled materials or large forgings, and only considers dynamic recrystallization behavior. However, there are few reports on the discontinuous deformation process of hot forging of large forgings. Summary of the Invention

[0005] The main objective of this invention is to overcome the shortcomings of existing technologies and provide a method for predicting the mixed grain level in the discontinuous deformation process of hot forging of large forgings. This invention moves the grain structure evaluation point forward, giving it the characteristics of both defect prediction and process optimization. It provides an intuitive and effective research method for understanding the influence of deformation parameters on the post-forging grain structure during the hot forging process of large forgings, and helps to control mixed grain defects in the hot forging process of large forgings.

[0006] This invention is achieved through the following technical solution: a method for predicting the mixed grain level in the discontinuous deformation process of hot forging of large forgings, comprising the following steps:

[0007] S1. Using dislocation density as the basic variable, track the change of dislocation density in the cell during discontinuous hot forging deformation and establish a cellular automaton (CA) model of the microstructure of discontinuous hot forging deformation process.

[0008] S1-1. Establish an evolution model of the microstructure in the dynamic recrystallization process;

[0009] S1-1-1 Establishing a dislocation density evolution model; During hot deformation, the work hardening caused by deformation leads to the proliferation of dislocation density, while the softening caused by dynamic recovery and dynamic recrystallization leads to the annihilation of dislocation density. KOCKS and MECKING proposed the Kocks-Mecking (KM model) physical model to describe the evolution of dislocation density during hot deformation. The KM model can be used to predict the evolution of dislocation density in grains.

[0010]

[0011] In equations (1) to (3), ρ ij is the cell dislocation density at coordinate (i,j); k1 is the parameter affecting dislocation density due to work hardening; k2 is the parameter affecting dislocation density due to dynamic softening; α is a constant with a value of 0.5; b is the Burgers vector; μ is the shear modulus; is the hardening rate; σ sat It is saturated stress;

[0012] In addition, the saturation stress σ sat The relationship between the yield stress σ0 and the yield stress σ0 is:

[0013]

[0014] In equation (4), σ0 is the yield stress and ε is the strain;

[0015] S1-1-2. Establish a dynamic recrystallization model; after the dislocation density in hot-deformed metals reaches the critical dislocation density, dynamic recrystallization begins to occur.

[0016] Critical dislocation density ρ c for:

[0017]

[0018] In equation (5), the dislocation mean free thread l is;

[0019]

[0020] The grain boundary mobility M is:

[0021]

[0022] The dislocation line energy τ is:

[0023]

[0024] Grain boundary energy γ between the matrix and recrystallized grains i for:

[0025]

[0026] Grain boundary energy γ at large angle grain boundaries m for:

[0027]

[0028] In equations (5) to (10), is the Burgers vector; μ refers to the shear modulus; k is a constant with a value of 10; δ is the grain boundary thickness of a specific material; D ob Q is the self-diffusion coefficient of the grain boundary at absolute zero. b K is the activation energy for grain boundary diffusion. B θ is the Boltzmann constant; i θ represents the orientation difference between recrystallized grains and adjacent grains. m Large-angle orientation angle; v is Poisson's ratio;

[0029] When the dislocation density of the matrix reaches a critical value, nucleation occurs preferentially in high dislocation density regions such as defect points and grain boundaries with a certain nucleation rate. This invention only considers the case of grain nucleation at grain boundaries and ignores the influence of other crystal defects on nucleation.

[0030]

[0031] In equation (11), Nucleation rate; It is the strain rate; C is the material constant; m is a constant, m=1; Q act It is the nucleation activation energy; R is the gas constant, R = 8.314 J·mol⁻¹ -1 ·K -1 T is the deformation temperature, in K.

[0032] Based on the assumption of spherical grains, the experimentally obtained dynamic recrystallization volume percentage x drx for:

[0033]

[0034] In equation (12), r d x represents the dynamic recrystallized grain radius obtained experimentally; for a specific material, x represents the value under different deformation conditions. drx and r d The value can be obtained through thermal simulation experiments, and therefore can be calculated using equation (12). The value of C is determined by the value of C.

[0035] S1-2. Static recovery is a process of microstructure and property change that occurs in metals after plastic processing at relatively low temperatures (below the recrystallization temperature). It primarily reduces the system's energy through the movement and redistribution of dislocations. An evolution model of cellular dislocation density during static recovery is established:

[0036]

[0037] In equation (13), t is the inter-pass heat preservation time. When t = 0, the dislocation density of the cell is the dislocation density stored in the cell at the end of deformation, i.e. Δt is the time increment step; k2 is the parameter representing the effect of dynamic softening on dislocation density; The residual strain stored within the cell;

[0038] S1-3. Static recrystallization nucleation is a thermally activated process that requires a certain incubation period. The grain growth process after static recrystallization nucleation is the same as that of dynamic recrystallization. The static recrystallization nucleation rate is determined as follows:

[0039]

[0040] In equation (14), C static V is a material constant; V(t) is the volume fraction of recrystallization nucleation occurring at time t; E is the storage energy; E min The minimum storage energy required for recrystallization nucleation;

[0041] The storage energy E is:

[0042] E=C0ρμb 2 V a (15)

[0043] In equation (15), C0 is a constant with a value of 0.5-1; V a ρ is the molar volume of austenite; ρ is the cell dislocation density during the static recovery process determined according to equation (13);

[0044] S1-4. Subdynamic recrystallization does not require an incubation period or a nucleation process. It is the further growth of grains that have already undergone dynamic recrystallization but have not yet had time to grow during high-temperature holding. An evolution model of the microstructure during the subdynamic recrystallization process is established.

[0045] After dynamic recrystallization occurs, the newly generated recrystallized grains have a very small dislocation density, which is significantly different from the matrix grains. This provides a driving force for the growth of recrystallized grains (grain boundary migration), causing the grain boundaries to tend to move towards the side with higher dislocation density, thereby allowing the low-energy recrystallized grains on the other side of the grain boundary to grow.

[0046] The growth rate v of recrystallized grains i for:

[0047] v i =Mf i (16)

[0048] In equation (16), v iLet M be the growth rate of recrystallized grains, M be the grain boundary mobility, and f be the grain growth driving force per unit area. i for:

[0049]

[0050] In equation (17), ρ m ρ is the dislocation density of the matrix grains; i d represents the dislocation density of the recrystallized grains. i γ is the diameter of the recrystallized grain; i This refers to the grain boundary energy between the matrix and the recrystallized grains.

[0051] This invention introduces the influence of second-phase particles on recrystallized grain growth. During recrystallized grain growth, when second-phase particles are present in the direction of grain boundary migration, the driving force for recrystallized grain growth must consider the "pinning" effect of the second-phase particles on grain boundary migration. Pinning force P Z for:

[0052]

[0053] In equation (18), γ i f is the grain boundary energy between the matrix and the recrystallized grains. v r represents the volume fraction of the precipitated secondary phase. p The average radius of the second phase;

[0054] The driving force f of recrystallization grain growth i It consists of three parts: interfacial energy, volume energy, and second-phase pinning force.

[0055]

[0056] S1-5, Grain coarsening;

[0057] After recrystallization, grains tend to coarsen during the inter-pass holding period. The driving force for grain growth differs from that of recrystallization; its primary driving force is the reduction of grain boundary energy. The driving force P for grain growth is:

[0058] P i =γ i k; (20)

[0059] In equation (20), γ i The grain boundary energy is the energy between the matrix and the recrystallized grains; the grain boundary curvature k is:

[0060]

[0061] In equation (21), A = 1.28; L is the side length of the cell; C N For the interface cell neighbors, CN =24; N i is the cell belonging to grain i among the neighboring cells; Kink is the cell belonging to grain i among the neighboring cells when the interface is assumed to be a flat interface, i.e., k = 0, Kink = 15;

[0062] S1-6. Establish a micro-organization CA model based on steps S1-1 to S1-5. The micro-organization CA model includes cells, cell states, cell space, neighbor types, and cell transformation rules, and includes the following steps:

[0063] S1-6-1, Set the material type, initial grain structure, and grain size, and set the dislocation density of the initial structure to 1×10⁻⁶. -10 μm -2 ;

[0064] S1-6-2. Input the hot deformation parameters, and calculate the number of cycles n in the simulation process based on the microstructure evolution model of the dynamic recrystallization process established in step S1-1. CA_dynamic for:

[0065]

[0066] In equation (22), ε total For the total strain; Δε CA The change in strain;

[0067] S1-6-3. After the dynamic recrystallization CA simulation cycle in step S1-6-2 is completed, calculate the number of CA cycles n for the static process. CA_static for:

[0068]

[0069] In equation (23), is the interval time for each track; Δt is the time increment step;

[0070] S1-6-4. Static recrystallization nucleation and growth only occur in regions where dynamic recrystallization has not yet occurred. The cell dislocation density during static recovery is determined according to step S1-2.

[0071] S1-6-5, Update X recry ;

[0072]

[0073] In equation (24), N recrystallization The total number of recrystallized cells, N0 is the total number of cells in the CA simulation;

[0074] If X recrystallization If the value is ≥0.98, then perform grain coarsening simulation according to step S1-5;

[0075] If X recrystallization If <0.98, then determine whether the static loop has ended: if the number of static loops is equal to the result calculated by equation (22), then the static loop ends; if the number of static loops is not equal to the result calculated by equation (22), then the static loop has not ended, and continue to calculate the static recovery density according to step S1-2, and perform the next deformation simulation until the static loop ends and the result is output.

[0076] S2, hot forging discontinuous deformation macro-micro multi-scale FE-CA coupling;

[0077] S2-1. Acquisition and Data Processing of Nodal Physical Fields: DEFORM software is commonly used to simulate the forging process of large forgings, covering processes such as upsetting and drawing. Through simulation, the metal flow trajectory can be visually presented, optimizing billet size and shape and reducing material waste. It can accurately analyze the temperature, stress, and strain distribution during the forging process, predict changes in the internal microstructure and properties of forgings, provide a strong basis for the formulation of subsequent heat treatment processes, greatly improve the quality of large forgings, reduce production costs, and promote the efficient development of the large forging manufacturing industry. This invention uses DEFORM software as an example to introduce the acquisition and data processing of nodal physical fields. After the finite element simulation is completed, the post-processing module is entered to obtain the physical field quantities of each node. The point tracking function of the software is used to view the values ​​of each physical field quantity of the element node, and the data is exported as text or spreadsheet format for subsequent CA simulation analysis.

[0078] S2-2, Simulation process flow;

[0079] 1) Construct a geometric model;

[0080] 2) Preprocessing module: First, finite element mesh generation: The constructed geometric model is assumed to be divided into several elements, and adjacent elements are connected by nodes to form an element assembly; then, object setting: Material parameters, boundary conditions, and motion settings are set for the discrete geometric model; finally, simulation settings: Finite element simulation parameters are defined.

[0081] 3) Finite element analysis solution module: First, the equilibrium differential equations, constitutive relations and boundary conditions are transformed into a nonlinear equation system through finite element discretization; then, calculations are performed using the direct iteration method and the Newton-Raphson method, and the solution results are saved in binary form; finally, users can use the software's point tracking function in the post-processor to obtain the macroscopic physical field information of the unit nodes in the hot forging process.

[0082] 4) Microstructure simulation module: Import the physical field information of each element node obtained from the finite element analysis solution module, and input the material parameters at the same time. Use the CA method to simulate the microstructure evolution of the discontinuous hot forging process of large forgings. This method can track the entire deformation process and realize the visualization of grain structure morphology.

[0083] 5) Process Analysis Module: Obtains the distribution of macroscopic physical field quantities in large forgings and the evolution law of grain structure of different material points during hot forging, and completes the prediction of mixed grain level in the discontinuous hot forging deformation process of large forgings. This invention helps to move the point of analysis and evaluation of microstructure of forgings forward, making it have the characteristics of grain structure prediction and process optimization, thereby optimizing the discontinuous hot forging deformation process of large forgings.

[0084] Furthermore, in steps S1-6, the cell, cell state, cell space, neighbor type, and cell transition rule are respectively:

[0085] 1) Cell: The initial grains adopt 2um×2um square cells;

[0086] 2) Cell State: Each cell contains 5 state variables for computation and 5 structural variables for statistics, as described below:

[0087] ① Dislocation density variable: The dislocation density variable represents the deformation storage energy of a grain;

[0088] ② Grain orientation variables: Different grains are represented by different orientations;

[0089] ③Grain boundary variable: used to record whether a cell is located at a grain boundary or within a grain;

[0090] ④ Recrystallization fraction variable: represents the proportion of each cell that undergoes recrystallization;

[0091] ⑤ Residual strain variables The trend of residual strain is consistent with the trend of dislocation density variable. The increase of residual strain within the cell is as follows:

[0092]

[0093] The following structure variables are used for statistics and display in the program:

[0094] ⑥ Grain numbering structural variable: assign a unique number to each dynamically recrystallized grain to calculate the recrystallized grain size;

[0095] ⑦ Grain color variable, recording the simulated grain morphology;

[0096] ⑧ Dynamic recrystallization number variable, "0" indicates no recrystallization. When a cell undergoes dynamic recrystallization, the number of dynamic recrystallizations of cells in the Von Neumann neighborhood that change its state must be greater than the number of recrystallizations of the cell itself;

[0097] ⑨ Static recrystallization number variable: "-1" indicates that no static recrystallization has occurred. When a cell undergoes static recrystallization, the number of static recrystallizations of cells in the Von Neumann neighborhood that change the state of that cell must be greater than its own number of recrystallizations.

[0098] ⑩ The number of cells contained in each grain is obtained by statistical analysis based on the grain number;

[0099] 3) Cellular space: 512×512 two-dimensional square cells are used, and the simulated area represents a 1024um×1024um actual sample;

[0100] 4) Neighbor type: Use Moore neighbor type;

[0101] 5) Cell transformation rule: The cell transformation rule is basically the same as that of dynamic recrystallization. When the static recrystallization grain growth and the subdynamic recrystallization grain growth collide during the inter-pass interval, the grain growth stops.

[0102] The beneficial effects of this invention are as follows:

[0103] 1. Using dislocation density as the basic variable, the change of dislocation density in the cell during discontinuous hot forging deformation was tracked, and a CA model of grain structure during the discontinuous hot forging deformation process of large forgings was established.

[0104] 2. The point tracking function of DEFORM-3D software is used to export macroscopic physical field information such as temperature, strain, and strain rate of unit nodes. This information is then coupled with the microstructure evolution CA model to integrate a multi-scale FE-CA simulation method for the discontinuous microstructure evolution of hot forging of large forgings.

[0105] 3. The FE-CA simulation method was verified by drawing process experiments. It was found that the average grain size and mixed crystal degree level of the samples from the same sampling position under the same deformation conditions were not much different from those of the simulation and the experiment, which confirmed the feasibility of the quantitative prediction method of mixed crystal degree in the discontinuous deformation process of hot forging of large forgings. Attached Figure Description

[0106] Figure 1 The stress-strain curves of 12% Cr heat-resistant steel for large forgings of ultra-supercritical rotors are shown in the preliminary state; where the strain rate in Figure (a) is 1 s. -1 The strain rate in Figure (b) is 0.1 s. -1 The strain rate in Figure (c) is 0.01 s.-1 The strain rate in Figure (d) is 0.001 s. -1 ;

[0107] Figure 2 Flowchart of CA simulation for discontinuous deformation grain structure in hot forging;

[0108] Figure 3 A simulation process flow diagram for discontinuous hot forging of large forgings;

[0109] Figure 4 This is an image of the initial grain structure morphology.

[0110] Figure 5 The figures show the grain morphology after FE-CA simulation at 1210℃ with different forging ratios; in Figures (a1) to (a3), the positions for forging ratio 1.1 are labeled 1 to 3; in Figures (b1) to (b3), the positions for forging ratio 1.3 are labeled 1 to 3; and in Figures (c1) to (c3), the positions for forging ratio 1.5 are labeled 1 to 3.

[0111] Figure 6 This is a schematic diagram of the mixed crystal grade evaluation method;

[0112] Figure 7 The images show the grain morphology after drawing experiments at 1210℃ with different forging ratios; in Figures (a1) to (a3), the positions for forging ratio 1.1 are labeled 1 to 3; in Figures (b1) to (b3), the positions for forging ratio 1.3 are labeled 1 to 3; and in Figures (c1) to (c3), the positions for forging ratio 1.5 are labeled 1 to 3.

[0113] Figure 8 A comparison chart showing the simulation and experimental results of the mixed crystallinity level at the same sampling location under various deformation conditions. Detailed Implementation

[0114] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0115] This embodiment takes the discontinuous hot forging process of 12%Cr heat-resistant steel for large forgings of ultra-supercritical rotors as an example. DEFORM-3D software is used to numerically simulate a typical multi-pass discontinuous hot forging process (drawing process) for large forgings. The anvil width ratio is controlled between 0.6 and 0.8, the V-anvil angle is 120°, and the anvil loading is controlled between 20% and 30%. The billet size is a cylinder with dimensions of Φ150mm × 100mm, the flipping method is 0°-180°-90°-180°, the forging temperature is set to 1210℃, and the forging ratios are set to 1.1, 1.3, and 1.5. The friction factor between the billet and the upper and lower dies is set to 0.7, the heat transfer coefficient is 5W / (m²·℃), and the upper and lower dies are preheated to 300℃ during the simulation. The material model used in the simulation is a dynamic recrystallization, static recrystallization, and subdynamic recrystallization kinetic model and grain size model related to 12%Cr heat-resistant steel. Based on actual forging conditions, the pressing speed is set to 5 mm / s.

[0116] A large hydraulic press was used to conduct a drawing test on 12% Cr heat-resistant steel using a flat-top, V-shaped anvil. The sample size, upper die pressing speed, forging temperature, and forging ratio settings corresponded to the Deform-3D numerical simulation. After the drawing was completed, the billet was immediately water-quenched, and after cooling, small samples of 6mm×6mm×6mm were cut for average grain size detection and mixed crystallinity grade evaluation.

[0117] like Figure 2 The method for predicting the mixed grain level in the discontinuous deformation process of hot forging of large forgings, as shown, includes the following steps:

[0118] S1. Establish a cellular automaton model of the microstructure of the discontinuous deformation process of hot forging;

[0119] S1-1. Establish an evolution model of the microstructure in the dynamic recrystallization process;

[0120] S1-1-1 Establish a model for the evolution of dislocation density;

[0121]

[0122] In equations (1) to (3), ρ ij is the cell dislocation density at coordinate (i,j); k1 is the parameter affecting dislocation density due to work hardening; k2 is the parameter affecting dislocation density due to dynamic softening; α is a constant with a value of 0.5; b is the Burgers vector; μ is the shear modulus; is the hardening rate; σ sat It is saturated stress;

[0123] In addition, the saturation stress σ sat The relationship between the yield stress σ0 and the yield stress σ0 is:

[0124]

[0125] In equation (4), σ0 is the yield stress, and ε is the strain; σ0, σ sat and θ can be derived from Figure 1 Obtained from the stress-strain curve; fitted ln(σ) sat From -σ) and ε, we can obtain k2 = 31.22, and thus k1 = 4.46 × 10 8 ;

[0126] S1-1-2. Establish a dynamic recrystallization model;

[0127] Critical dislocation density ρ c for:

[0128]

[0129] In equation (5), the dislocation mean free thread l is;

[0130]

[0131] The grain boundary mobility M is:

[0132]

[0133] The dislocation line energy τ is:

[0134]

[0135] Grain boundary energy γ between the matrix and recrystallized grains i for:

[0136]

[0137] Grain boundary energy γ at large angle grain boundaries m for:

[0138]

[0139] In equations (5) to (10), is the Burgers vector; μ refers to the shear modulus; k is a constant with a value of 10; δ is the grain boundary thickness of a specific material; D ob Q is the self-diffusion coefficient of the grain boundary at absolute zero. b K is the activation energy for grain boundary diffusion. B θ is the Boltzmann constant; i θ represents the orientation difference between recrystallized grains and adjacent grains. m Large orientation angle, θ in this embodiment m =15°; v is Poisson's ratio; see Table 1 for specific values;

[0140] Table 1. Parameters of 12% Cr heat-resistant steel

[0141]

[0142]

[0143] In equation (11), Nucleation rate; It is the strain rate; C is the material constant; m is a constant, m=1; Q act It is the nucleation activation energy; R is the gas constant, R = 8.314 J·mol⁻¹ -1 ·K -1 T is the deformation temperature, in K.

[0144] Based on the assumption of spherical grains, the experimentally obtained dynamic recrystallization volume percentage x drx for:

[0145]

[0146] In equation (12), r d x represents the dynamic recrystallized grain radius obtained experimentally; for a specific material, x represents the value under different deformation conditions. drx and r d The value can be obtained through thermal simulation experiments, and therefore can be calculated using equation (12). The value of C is determined by the value of C.

[0147] S1-2. Establish an evolution model of cell dislocation density during static recovery:

[0148]

[0149] In equation (13), t is the inter-pass heat preservation time. When t = 0, the dislocation density of the cell is the dislocation density stored in the cell at the end of deformation, i.e. Δt is the time increment step, which is equal to the time increment step of the dynamic recrystallization process in step S1-6-3; k2 is the parameter for the effect of dynamic softening on dislocation density; The residual strain stored within the cell;

[0150] S1-3. Determine the static recrystallization nucleation rate:

[0151]

[0152] In equation (14), C static As a material constant, this embodiment uses 12% Cr heat-resistant steel for ultra-supercritical rotor forgings, therefore C static =3.0161E5; V(t) is the volume fraction of recrystallization nucleation occurring at time t; E is the stored energy; E min The minimum storage energy required for recrystallization nucleation;

[0153] The storage energy E is:

[0154] E=C0ρμb 2 V a (15)

[0155] In equation (15), C0 is a constant with a value of 0.5-1; V a The molar volume of austenite is 7.2 cm³, which is taken as 7.2 cm³ in this embodiment. 3 / mol; ρ is the cell dislocation density during the static recovery process determined according to equation (13); in this embodiment, 12% Cr heat-resistant steel for ultra-supercritical rotor forgings is used, so the storage energy E corresponding to the strain of 0.1 is taken;

[0156] S1-4. Establish an evolution model of the microstructure during the subdynamic recrystallization process; the growth rate v of recrystallized grains. i for:

[0157] v i =Mf i (16)

[0158] In equation (16), v i Let M be the growth rate of recrystallized grains, M be the grain boundary mobility, and f be the grain growth driving force per unit area. i for:

[0159]

[0160] In equation (17), ρ m ρ is the dislocation density of the matrix grains; i d represents the dislocation density of the recrystallized grains. i γ is the diameter of the recrystallized grain; i The grain boundary energy between the matrix and the recrystallized grains; pinning force P Z for:

[0161]

[0162] In equation (18), γ i f is the grain boundary energy between the matrix and the recrystallized grains. v r represents the volume fraction of the precipitated secondary phase. p The average radius of the second phase;

[0163] The driving force f of recrystallization grain growth i It consists of three parts: interfacial energy, volume energy, and second-phase pinning force.

[0164]

[0165] S1-5, Grain coarsening;

[0166] The driving force P for grain growth is:

[0167] P i =γ i k; (20)

[0168] In equation (20), γ i The grain boundary energy is the energy between the matrix and the recrystallized grains; the grain boundary curvature k is:

[0169]

[0170] In equation (21), A = 1.28; L is the side length of the cell; C N For the interface cell neighbors, C N =24; N i is the cell belonging to grain i among the neighboring cells; Kink is the cell belonging to grain i among the neighboring cells when the interface is assumed to be a flat interface, i.e., k = 0, Kink = 15;

[0171] S1-6. Establish a micro-organization CA model based on steps S1-1 to S1-5. The micro-organization CA model includes cells, cell states, cell space, neighbor types, and cell transformation rules, and includes the following steps:

[0172] S1-6-1. Set the material composition, initial grain structure, and grain size. The initial grain structure morphology is as follows: Figure 4 As shown, the initial dislocation density of the structure is set to 1×10⁻⁶. -10 μm -2 ;

[0173] S1-6-2. Input the hot deformation parameters, and calculate the number of cycles n in the simulation process based on the microstructure evolution model of the dynamic recrystallization process established in step S1-1. CA_dynamic for:

[0174]

[0175] In equation (22), ε total For the total strain; Δε CA The change in strain;

[0176] S1-6-3. After the dynamic recrystallization CA simulation cycle in step S1-6-2 is completed, calculate the number of CA cycles n for the static process. CA_static for:

[0177]

[0178] In equation (23), is the interval time for each track; Δt is the time increment step;

[0179] S1-6-4. Determine the cell dislocation density during the static recovery process according to step S1-2;

[0180] S1-6-5, Update X recry ;

[0181]

[0182] In equation (24), N recrystallization The total number of recrystallized cells, N0 is the total number of cells in the CA simulation;

[0183] If X recrystallization If the value is ≥0.98, then perform grain coarsening simulation according to step S1-5;

[0184] If X recrystallization If <0.98, then determine whether the static loop has ended: if the number of static loops is equal to the result calculated by equation (22), then the static loop ends; if the number of static loops is not equal to the result calculated by equation (22), then the static loop has not ended, and continue to calculate the static recovery density according to step S1-2, and perform the next deformation simulation until the static loop ends and the result is output.

[0185] S2, hot forging discontinuous deformation macro-micro multi-scale FE-CA coupling;

[0186] S2-1. Acquisition and Data Processing of Nodal Physical Quantities: This invention uses DEFORM software as an example to introduce the acquisition and data processing of nodal physical quantity fields, such as... Figure 3 As shown. After the finite element simulation is completed, the post-processing module is entered to obtain the physical field quantities of each node. The point tracking function of the software is used to view the values ​​of each physical field quantity of the element node, and the data is exported as text or spreadsheet format for subsequent CA simulation analysis;

[0187] S2-2, Simulation process flow: The CA model of microstructure evolution during the discontinuous hot forging process of 12%Cr heat-resistant steel established above is combined with the relevant modules of DEFORM-3D numerical simulation software to integrate a multi-scale simulation system to simulate the microstructure evolution during the discontinuous hot deformation process of large forging materials. Figure 3 The simulation process flow diagram for discontinuous hot forging of large forgings is shown below.

[0188] 1) Constructing the geometric model: This invention uses PROE 5.0 for geometric modeling, and the exported format is STL.

[0189] 2) Pre-processing Module: Pre-processing is a key step in finite element analysis, as many definitions are performed during this stage. This invention uses DEFORM-3D software for finite element analysis. First, finite element mesh generation: The constructed geometric model is assumed to be divided into several elements, with adjacent elements connected by nodes to form an element assembly. Then, object setup: Material parameters, boundary conditions, and motion settings are set for the discrete geometric model. Finally, simulation setup: Finite element simulation parameters are defined.

[0190] 3) Finite Element Analysis Solving Module: When the Deform-3D software is running, firstly, the equilibrium differential equations, constitutive relations and boundary conditions are transformed into a nonlinear equation system through finite element discretization; then, calculations are performed using the direct iteration method and the Newton-Raphson method, and the results are saved in binary form; finally, users can use the software's point tracking function in the post-processor to obtain macroscopic physical field information of the unit nodes in the hot forging process.

[0191] 4) Microstructure Simulation Module: Import the physical field information of each element node obtained from the finite element analysis solution module, and input the material parameters at the same time. Use the CA method to simulate the microstructure evolution of the discontinuous hot forging process of large forgings.

[0192] 5) Process Analysis Module: Obtains the distribution of macroscopic physical field quantities of large forgings and the evolution law of grain structure of different material points during hot forging, and completes the prediction of mixed grain level in the discontinuous deformation process of hot forging of large forgings. Further, in steps S1-6, the cell, cell state, cell space, neighbor type, and cell transformation rule are respectively:

[0193] 1) Cell: The initial grains adopt 2um×2um square cells;

[0194] 2) Cell States: Each cell contains 5 state variables for calculation and 5 structural variables for statistics. Each cell has two state variables: one representing orientation, expressed as an integer between 1 and 180; the other is a grain boundary variable, where a value of "1" indicates the cell is located at a grain boundary, and a value of "0" indicates the cell is located within a grain. A detailed description follows:

[0195] ① Dislocation density variable: The dislocation density variable represents the deformation storage energy of a grain;

[0196] ② Grain orientation variables: Different grains are represented by different orientations;

[0197] ③Grain boundary variable: used to record whether a cell is located at a grain boundary or within a grain;

[0198] ④ Recrystallization fraction variable: represents the proportion of each cell that undergoes recrystallization;

[0199] ⑤ Residual strain variables The trend of residual strain is consistent with the trend of dislocation density variable. The increase of residual strain within the cell is as follows:

[0200]

[0201] ⑥ Grain numbering structural variable: assign a unique number to each dynamically recrystallized grain to calculate the recrystallized grain size;

[0202] ⑦ Grain color variable, recording the simulated grain morphology;

[0203] ⑧ Dynamic recrystallization number variable, "0" indicates no recrystallization. When a cell undergoes dynamic recrystallization, the number of dynamic recrystallizations of cells in the Von Neumann neighborhood that change its state must be greater than the number of recrystallizations of the cell itself;

[0204] ⑨ Static recrystallization number variable: "-1" indicates that no static recrystallization has occurred. When a cell undergoes static recrystallization, the number of static recrystallizations of cells in the Von Neumann neighborhood that change the state of that cell must be greater than its own number of recrystallizations.

[0205] ⑩ The number of cells contained in each grain is obtained by statistical analysis based on the grain number;

[0206] 3) Cellular space: 512×512 two-dimensional square cells are used, and the simulated area represents a 1024um×1024um actual sample;

[0207] 4) Neighbor type: Use Moore neighbor type;

[0208] 5) Cellular transformation rule: When static recrystallization grain growth and subdynamic recrystallization grain growth collide during the inter-pass interval, grain growth stops.

[0209] The cell transition rules and their physical significance are as follows.

[0210] Rule 1: Grain growth (grain boundary migration) is a thermally activated process. Atoms located at grain boundaries must overcome a certain energy barrier to transition to a new state. That is, the thermal energy of the unit cell at the grain boundary must exceed the material's migration activation energy Q. b Orientation changes are possible only at higher temperatures, and the higher the temperature, the easier it is for cells located at grain boundaries to overcome energy barriers and transition to a new state. At each CA simulation time increment, the ease with which a cell transitions is determined by probability P1, as expressed below:

[0211]

[0212] In the formula, C is a constant, which can be obtained by considering T = T m When P1 = 1, the value of C is obtained, and T is the austenitizing temperature. m It is the melting point of the material, T Ac1 This is the temperature at which austenitization begins. Rule 1 considers the effect of temperature on austenitized grain growth; the value of P1 is based on the temperature increasing to A. c1 At this time, austenitization begins, and as the temperature increases, the value of P1 gradually increases, meaning that the cell transformation becomes easier.

[0213] It is generally believed that for grains grown in a normal growth manner, when they reach a stable state, the grain morphology should satisfy the following conditions: the grain boundaries should be straight, and the angle between the grain boundaries should be 120°. Grain growth should follow these rules: First, curved grain boundaries tend to straighten, that is, the grain boundaries move towards the center of curvature to reduce surface area and lower surface energy; second, when the angle between the three grain boundaries is not equal to 120°, the grain boundaries always move towards the grain with the smaller angle, causing the three angles to tend towards 120°. Based on the above rules of normal grain growth, and considering the influence of grain boundary curvature on grain growth, cell transformation rules 2-4 are formulated, respectively considering the influence of nearest neighbor and second nearest neighbor cells on the state of the central cell, as described in detail below.

[0214] Rule 2: According to Moore neighbor type, if cell C i If there are 5 or more consecutive cells in the same state, then in the next time step, C i Their states are the same as their states. Rule 2 can be expressed as follows:

[0215]

[0216] Rule 3: If C i-3 C i-1 C i+1 and C i+3 If any three cells in a cell have the same state, then in the next CA simulation time increment step, C i Their states are the same as their states. Rule 3 can be expressed as follows:

[0217]

[0218] Rule 4: If C i-2 C i-4 C i+2 and C i+4 If any three cells in a cell have the same state, then in the next CA simulation time increment step, C i Their states are the same as their states. Rule 4 can be expressed as follows:

[0219]

[0220] Rule 5: If none of the rules 2-4 above are satisfied, then start from C. i A cell j is randomly selected from the eight cells in the Moore neighborhood for transformation, and the change in grain boundary energy of the system after the transformation is calculated. The change in grain boundary energy can be calculated using the following formula:

[0221]

[0222] In the formula, J is a measure of grain boundary energy, taken as J = 1; δ is the Kronecher symbol; k is the k-th neighbor of cell i; R is the total number of neighbors of cell i, in this simulation R = 8; C i C is the orientation number of the cell; k Let be the orientation of the k-th neighbor of cell i.

[0223] If ΔE i→j If ΔE < 0, the probability of transition is 1; if ΔE i→j If the value is ≥0, the probability of transition is 0. Rule 5 can be expressed as follows:

[0224]

[0225] The CA model program flow for normal grain growth is as follows: The initial grain structure remains the same for different simulation temperatures. The program starts by inputting the initial model and material parameters. At simulation time increment 0, a grain orientation is arbitrarily assigned to each cell in a 400×400 cell space. In each subsequent CA simulation increment, the cells are evaluated according to rules one through five. Finally, the average grain size is calculated. If it meets the requirements, the program exits; otherwise, it continues looping.

[0226] To demonstrate the effectiveness of this method, a verification experiment was conducted in this embodiment. The specific analysis and results of the verification experiment are as follows.

[0227] (I) Analysis of FE-CA simulation results;

[0228] 1. Grain structure morphology;

[0229] Figure 5 The grain structure morphology after FE-CA simulation of forging with a flat top and V-shaped bottom anvil at a forging temperature of 1210℃ and forging ratios of 1.1, 1.3, and 1.5 is presented. In the simulation, white represents the initial grain topology at the end of each pass, and colored represents the newly recrystallized grains formed during the deformation CA simulation. Figure 5 It can be observed that as the forging ratio increases, the degree of recrystallization gradually increases, while under the same deformation conditions, the degree of recrystallization from the core to the edge of the billet gradually decreases.

[0230] 2. Prediction of mixed crystallinity level;

[0231] This step adopts the patent document "A Universal Method for Evaluating Mixed Grain Structures of Austenite" (application number: 202110738402.1). Based on this method, the mixed grain structure of each sampling point under different deformation conditions obtained by the FE-CA simulation method is evaluated. The evaluation results are shown in Table 2. As can be seen from the table, the mixed grain structure at each location first increases and then decreases with the increase of the forging ratio. This is because when the forging ratio is small, the degree of dynamic recrystallization of the billet is low. In addition, when the forging ratio is small, the holding time between passes is also relatively short, making it difficult for static recrystallization and sub-dynamic recrystallization between passes to occur. The microstructure is mostly original coarse grains with few recrystallized small grains, so the mixed grain structure is small. With the increase of the forging ratio, the degree of dynamic recrystallization and the degree of recrystallization between passes both increase. At this time, there are more recrystallized small grains, but a large number of original coarse grains can still be observed. Therefore, coarse and fine grains coexist, and the mixed grain structure is larger. As the forging ratio further increases, the degree of dynamic recrystallization and inter-pass recrystallization is greater, and the field of view is mostly filled with small recrystallized grains, with very few original large grains, and the degree of mixed crystallinity decreases.

[0232] Table 2. Grain microstructure mixed crystallinity levels at each sampling point after FE-CA simulation under different deformation conditions.

[0233]

[0234] 3. Experimental verification

[0235] To verify the accuracy of the numerical simulation and explore the control of discontinuous hot forging process parameters on grain structure, a top-flat, bottom-V anvil drawing experiment was conducted on 12% Cr heat-resistant steel on a 500t hydraulic press. The sample size, upper die pressing speed, forging temperature, forging ratio settings, and sampling location were consistent with the numerical simulation. The experimental results are as follows: Figure 7 As shown, when the forging ratio is 1.1, samples numbered 1 to 3 were taken from the cross-section of the billet directly below the anvil. The microstructure after corrosion is shown below. Figure 7 As shown in (a1) to (a3), it can be seen that the grains from the center to the edge of the billet are relatively coarse. The original austenite grain boundaries at three locations near the edge of the billet exhibit a serrated shape, while a small amount of recrystallized small grains appear at locations 1 and 2. At a forging ratio of 1.3, the grain size at different locations on the billet is further refined, especially at locations 1 and 2, where the grains have become significantly refined. Figure 7 As shown in (b1) to (b3). But Figure 7 Larger coarse grains can still be seen in the middle. Figure 7The average grain sizes corresponding to (b1), (b2), and (b3) are 61 μm, 73 μm, and 136 μm, respectively. A comparison shows that the grain sizes at positions 1 and 2 are not significantly different, but the difference with position 3 is greater than 60 μm. This indicates a high degree of overall grain inhomogeneity across the billet cross-section from the core to the surface. This is because position 3 is located in the difficult-to-deform zone at the edge of the billet. Although the forging ratio is relatively large at this point, the increased deformation energy also leads to a longer overall deformation time, resulting in a greater drop in the billet surface temperature. Therefore, the degree of recrystallization at position 3 is lower than that at positions 1 and 2. Simultaneously, a more severe mixed-grain phenomenon is observed in the local field of view at position 3. At a forging ratio of 1.5, the grains at position 3 on the billet surface are significantly refined compared to when the forging ratio is 1.3, and the grain inhomogeneity in the local field of view is also greatly improved, with less severe mixed-grain phenomenon in the local field of view. Figure 7 As shown in (c1), (c2) and (c3), the average grain sizes at positions 1 to 3 were measured to be 35 μm, 39 μm and 54 μm respectively by the truncated chord method. The comparison shows that the average grain size at the core and edge of the billet does not differ by more than 20 μm. This indicates that the grain fineness and uniformity of the entire cross-section of the billet has been significantly improved, and the mixed grain phenomenon in the local field of view at each sampling position has also been further improved.

[0236] In summary, the trend of the mixed crystallinity level at various locations on the billet cross-section obtained in the experiment, with the change of forging ratio, is consistent with the evolution law of mixed crystal structure at the corresponding sampling locations under various deformation conditions observed after simulation. Both show an initial increase followed by a decrease with increasing forging ratio, and the mixed crystallinity level at location 3 of the billet is the highest under the same deformation conditions. The mixed crystallinity levels calculated based on the above mixed crystallinity evaluation method under various deformation conditions are shown in Table 3. The calculation results also show that the degree of mixed crystallinity is greatest at the edge of the billet. Figure 8 The figure shows a comparison between the simulation and experimental results of grain size and mixed crystallinity levels in different deformation regions. It was observed that the average grain size and mixed crystallinity level after the experiment were not much different from the simulation results, which verified the accuracy of the simulation and showed that the multi-scale FE-CA simulation method can predict the grain structure in the discontinuous deformation process of hot forging of typical large forgings.

[0237] Table 3 Comparison of mixed crystallinity levels at various sampling points under different forging ratios after experiments and simulations.

[0238]

[0239] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for predicting the mixed grain level in the discontinuous deformation process of hot forging of large forgings, characterized in that, Includes the following steps: S1. Establish a cellular automaton model of the microstructure of the discontinuous deformation process of hot forging; S1-1. Establish an evolution model of the microstructure in the dynamic recrystallization process; S1-1-1 Establish a model for the evolution of dislocation density; In equations (1) to (3), ρ ij σ is the cell dislocation density at coordinate (i,j); k1 is the parameter affecting dislocation density due to work hardening; k2 is the parameter affecting dislocation density due to dynamic softening; α is a constant with a value of 0.5; b is the Burgers vector; μ is the shear modulus; θ is the hardening rate; σ sat It is saturated stress; In addition, the saturation stress σ sat The relationship between the yield stress σ0 and the yield stress σ0 is: In equation (4), σ0 is the yield stress and ε is the strain; S1-1-2. Establish a dynamic recrystallization model; Critical dislocation density ρ c for: In equation (5), the dislocation mean free thread l is; The grain boundary mobility M is: The dislocation line energy τ is: Grain boundary energy γ between the matrix and recrystallized grains i for: Grain boundary energy γ at large angle grain boundaries m for: In equations (5) to (10), b is the Burgers vector; μ refers to the shear modulus; k is a constant with a value of 10; δ is the grain boundary thickness of a specific material; D ob Q is the self-diffusion coefficient of the grain boundary at absolute zero. b K is the activation energy for grain boundary diffusion. B θ is the Boltzmann constant; i θ represents the orientation difference between recrystallized grains and adjacent grains. m Large-angle orientation angle; v is Poisson's ratio; S1-1-3. Establish a dynamic recrystallization nucleation model; In equation (11), Nucleation rate; It is the strain rate; C is the material constant; m is a constant, m=1; Q act It is the nucleation activation energy; R is the gas constant, R = 8.314 J·mol⁻¹ -1 ·K -1 T is the deformation temperature, in K. Based on the assumption of spherical grains, the experimentally obtained dynamic recrystallization volume percentage x drx for: In equation (12), r d The radius of the dynamically recrystallized grains obtained from the experiment; S1-2. Establish an evolution model of cell dislocation density during static recovery: In equation (13), t is the inter-pass heat preservation time. When t = 0, the dislocation density of the cell is the dislocation density stored in the cell at the end of deformation, i.e. Δt is the time increment step; k2 is the parameter representing the effect of dynamic softening on dislocation density; The residual strain stored within the cell; S1-3. Determine the static recrystallization nucleation rate: In equation (14), C static V is a material constant; V(t) is the volume fraction of recrystallization nucleation occurring at time t; E is the storage energy; E min The minimum storage energy required for recrystallization nucleation; The storage energy E is: E=C0ρµb 2 V a (15) In equation (15), C0 is a constant with a value of 0.5-1; V a ρ is the molar volume of austenite; ρ is the cell dislocation density during the static recovery process determined according to equation (13); S1-4. Establish an evolution model of the microstructure during the subdynamic recrystallization process; the growth rate v of recrystallized grains. i for: v i *Mf i (16) In equation (16), v i Let M be the growth rate of recrystallized grains, M be the grain boundary mobility, and f be the grain growth driving force per unit area. i for: In equation (17), ρ m ρ is the dislocation density of the matrix grains; i d represents the dislocation density of the recrystallized grains. i γ is the diameter of the recrystallized grain; i This refers to the grain boundary energy between the matrix and the recrystallized grains. Pinning force P Z for: In equation (18), γ i f is the grain boundary energy between the matrix and the recrystallized grains. v r represents the volume fraction of the precipitated secondary phase. p The average radius of the second phase; The driving force f of recrystallization grain growth i It consists of three parts: interfacial energy, volume energy, and second-phase pinning force. S1-5, Grain coarsening; The driving force P for grain growth is: P i =c i k; (20) In equation (20), γ i The grain boundary energy is the energy between the matrix and the recrystallized grains; the grain boundary curvature k is: In equation (21), A = 1.28; L is the side length of the cell; C N For the interface cell neighbors, C N =24; N i is the cell belonging to grain i among the neighboring cells; Kink is the cell belonging to grain i among the neighboring cells when the interface is assumed to be a flat interface, i.e., k = 0, Kink = 15; S1-6. Establish a micro-organization CA model based on steps S1-1 to S1-5. The micro-organization CA model includes cells, cell states, cell space, neighbor types, and cell transformation rules, and includes the following steps: S1-6-1, Set the material type, initial grain structure, and grain size, and set the dislocation density of the initial structure to 1×10⁻⁶. -10 μm -2 ; S1-6-2. Input the hot deformation parameters, and calculate the number of cycles n in the simulation process based on the microstructure evolution model of the dynamic recrystallization process established in step S1-1. CA_dynamic for: In equation (22), ε total For the total strain; Δε CA The change in strain; S1-6-3. After the dynamic recrystallization CA simulation cycle in step S1-6-2 is completed, calculate the number of CA cycles n for the static process. CA_static for: In equation (23), t inter-pass Δt represents the time interval between each pass; Δt represents the time increment step. S1-6-4. Determine the cell dislocation density during the static recovery process according to step S1-2; S1-6-5, Update X recry ; In equation (24), N recrystallization The total number of recrystallized cells, N0 is the total number of cells in the CA simulation; If X recrystallization If the value is ≥0.98, then perform grain coarsening simulation according to step S1-5; If X recrystallization If <0.98, then determine whether the static loop has ended: if the number of static loops is equal to the result calculated by equation (22), then the static loop ends; if the number of static loops is not equal to the result calculated by equation (22), then the static loop has not ended, and continue to calculate the static recovery density according to step S1-2, and perform the next deformation simulation until the static loop ends and the result is output. S2, hot forging discontinuous deformation macro-micro multi-scale FE-CA coupling; S2-1. Obtaining and processing the physical field quantities at the nodes: Use the point tracking function of the software to view the values ​​of each physical field quantity at the unit nodes, and export the data as text or spreadsheet format for subsequent CA simulation analysis. S2-2, Simulation process flow; 1) Construct a geometric model; 2) Preprocessing module: First, finite element mesh generation: The constructed geometric model is assumed to be divided into several elements, and adjacent elements are connected by nodes to form an element assembly; then, object setting: Material parameters, boundary conditions, and motion settings are set for the discrete geometric model; finally, simulation settings: Finite element simulation parameters are defined. 3) Finite element analysis solution module: First, the equilibrium differential equations, constitutive relations and boundary conditions are transformed into a nonlinear equation system through finite element discretization; then, calculations are performed using the direct iteration method and the Newton-Raphson method, and the solution results are saved in binary form; finally, users can use the software's point tracking function in the post-processor to obtain the macroscopic physical field information of the unit nodes in the hot forging process. 4) Microstructure Simulation Module: Import the physical field information of each element node obtained from the finite element analysis solution module, and input the material parameters at the same time. Use the CA method to simulate the microstructure evolution of the discontinuous hot forging process of large forgings. 5) Process Analysis Module: Obtain the distribution of macroscopic physical field quantities of large forgings and the evolution law of grain structure of different material points during hot forging, and complete the prediction of mixed crystal level in the discontinuous deformation process of hot forging of large forgings.

2. The method for predicting the mixed grain level in the discontinuous deformation process of hot forging of large forgings according to claim 1, characterized in that, In steps S1-6, the cell, cell state, cell space, neighbor type, and cell transition rule are respectively: 1) Cell: The initial grains adopt 2um×2um square cells; 2) Cell State: Each cell contains 5 state variables for computation and 5 structural variables for statistics, as described below: ① Dislocation density variable: The dislocation density variable represents the deformation storage energy of a grain; ② Grain orientation variables: Different grains are represented by different orientations; ③Grain boundary variable: used to record whether a cell is located at a grain boundary or within a grain; ④ Recrystallization fraction variable: represents the proportion of each cell that undergoes recrystallization; ⑤ Residual strain variables The trend of residual strain is consistent with the trend of dislocation density variable. The increase of residual strain within the cell is as follows: ⑥ Grain numbering structural variable: assign a unique number to each dynamically recrystallized grain to calculate the recrystallized grain size; ⑦ Grain color variable, recording the simulated grain morphology; ⑧ Dynamic recrystallization number variable, "0" indicates that no recrystallization has occurred; when a cell undergoes dynamic recrystallization, the number of dynamic recrystallizations of cells in the Von Neumann neighborhood that change the state of that cell must be greater than its own number of recrystallizations. ⑨ Static recrystallization number variable, "-1" indicates that no static recrystallization has occurred. When a cell undergoes static recrystallization, the number of static recrystallizations of cells in the Von Neumann neighborhood that change the state of that cell must be greater than its own number of recrystallizations. ⑩ The number of cells contained in each grain is obtained by statistical analysis based on the grain number; 3) Cellular space: 512×512 two-dimensional square cells are used, and the simulated area represents a 1024um×1024um actual sample; 4) Neighbor type: Use Moore neighbor type; 5) Cellular transformation rule: When static recrystallization grain growth and subdynamic recrystallization grain growth collide during the inter-pass interval, grain growth stops.

Images