Method for improving microstructure simulation precision of casting

Abstract

The invention discloses a method for improving the microstructure simulation precision of a casting, which starts from a # imgabs0 # simulation model and principle, and specifically comprises the following steps: carrying out atomic mean square displacement simulation on the basis of lamps to obtain diffusion customs of different elements in alloy substrate elements, and meanwhile, obtaining a Gibbs-Thompson coefficient of a certain component alloy on the basis of thermodynamic calculation, so as to improve the microstructure simulation precision of the casting. Therefore, high-precision dendritic crystal growth kinetic parameters are obtained. Secondly, performing statistical analysis on EBSD simulation results to obtain exact nucleation parameters of the castings. And finally, standardized operation is conducted on the flow of mold flow analysis, a temperature field result with high credibility is obtained, high-precision casting microstructure simulation is achieved, therefore, real-time feedback is conducted on design and optimization of a process scheme, and finally effective guidance on actual production is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of casting microstructure simulation, and in particular to a method for improving the simulation accuracy of casting microstructure. Background Art

[0002] The simulation methods of macro defects such as shrinkage and shrinkage holes are relatively mature and can be efficiently completed based on various casting finite element simulation software. Microstructure simulation can be achieved based on neural networks, CA cellular automata and other methods. However, such methods can generally only be used to simulate local microstructures, and their temperature fields are generally set to uniform temperatures, which is quite different from the actual molding environment of castings. The method interpolates the temperature field of the cellular model based on the model node information of the finite element simulation results, thereby obtaining a high-precision temperature field of the casting and achieving a microstructure simulation that is closer to reality.

[0003] Procast is a widely used casting finite element simulation software that can achieve high-precision simulation of casting process flow, solidification and temperature change distribution. In addition, based on finite element and cellular automation The module can be used to simulate and predict the evolution of grains and the microstructures formed during the solidification process of castings. Among them, the growth kinetics parameters and nucleation parameters of the model have a great influence on the simulation accuracy. However, due to the diversity of casting material types and their components, it is difficult to quickly determine the growth kinetics parameters and nucleation parameters for a certain alloy or component. For this, the official solution given by Procast is to use unified default parameters. This processing method of using unified default parameters for simulation reduces the accuracy of the simulation from the source, and the modification of a series of simulation parameters such as nucleation parameters based on default parameters can only be used for qualitative analysis. The significance of microstructure simulation is not significant and cannot provide effective guidance for actual production. Summary of the invention

[0004] In view of the above-mentioned shortcomings of the prior art, the object of the present invention is to provide a method for improving the simulation accuracy of casting microstructure. Starting from the simulation model and principle, the existing processing methods are improved through experiments, calculations, molecular dynamics simulation and other means, so that high-precision growth kinetics parameters and nucleation parameters of the target alloy can be quickly obtained. At the same time, the process of mold flow analysis is standardized to obtain temperature field results with high credibility, and high-precision casting microstructure simulation is achieved. Then, real-time feedback is given to the design and optimization of the process scheme, and ultimately effective guidance for actual production is achieved, so as to solve the problems of low simulation accuracy of existing technologies, insignificant significance of microstructure simulation, and only qualitative analysis, which cannot provide effective guidance for actual production.

[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0006] A method for improving the simulation accuracy of casting microstructure, specifically comprising the following steps:

[0007] Step 1: Based on Lammps, an atomic mean square displacement model of binary alloys is constructed and molecular dynamics simulation is performed to obtain the diffusion coefficients of different elements in the alloy matrix element;

[0008] Step 2: Perform electron backscatter diffraction (EBSD) characterization on the test sample, filter the results, remove the noise, and then use the area weighting method to count the grain size and number in the area to obtain the number of planar grains. At the same time, determine the body nucleus density and surface nucleus density of the alloy based on the algorithm recommended by ASTM;

[0009] Step 3: Obtain the Gibbs-Thompson coefficient through thermodynamic calculation;

[0010] Step 4: Import the high-precision simulation parameters obtained in Steps 1 and 3 into the Procast-based The calculation module combines the Procast built-in material calculation module to complete the calculation of the dendrite growth kinetic parameters of the KGT model, and sets the nucleation parameters obtained in step 2 to Boundary conditions of the simulation;

[0011] Step 5: Use Procast to complete the setting of casting molding process parameters, boundary conditions, simulation parameters, and simulate the molding process to obtain the temperature field and flow field during the casting molding process, and then conduct Calculation is performed to finally obtain high-precision casting microstructure simulation results.

[0012] Preferably, in step 1, the specific steps of constructing the atomic mean square displacement model based on Lammps are as follows:

[0013] S1: Set the model unit, boundary type, atom type, and time step;

[0014] S2: Set the area for atomic distribution simulation, including setting the range of the atomic distribution area, merging the two areas into a box area, filling the matrix atom A and the atom B to be solved, and setting the molar masses of the two atoms respectively. After that, divide the boundary area and fix the left and right boundaries of the box area;

[0015] S3: Set up the force field, thermodynamic output information, and energy minimization model;

[0016] S4: Set the temperature initialization parameters, relaxation parameters, and heat and pressure diffusion parameters. The resulting .data file contains the basic information of the atomic mean square displacement molecular dynamics model. Combined with the .pot potential energy file and .in script file used, simulation calculations can be performed based on Lammps.

[0017] Preferably, in S3, the energy minimization model uses the conjugate gradient descent method, and the operating conditions for stopping the minimization are that the energy tolerance, the force tolerance and the number of iteration steps reach corresponding values.

[0018] Preferably, in S4, during the setting of the heat preservation and pressure preservation diffusion parameters, the calculation and result output of the atomic mean square displacement MSD are set, the diffusion coefficient is fitted according to the data after the calculation is completed, and the EAM potential function is selected for the setting of the potential energy.

[0019] Preferably, in step 1, the diffusion coefficients of atoms of different elements in the alloy matrix element are obtained using the atomic mean square displacement model and are specifically calculated as follows:

[0020]

[0021] In formula (1), Dsim is the atomic diffusion coefficient, To simulate the system dimension;

[0022] The mean square displacement of an atom is defined as:

[0023]

[0024] In formula (2), MSD is the mean square displacement of atoms, r i is the atomic position, N is the total number of atoms;

[0025] From equations (1) and (2), the relationship between the atomic diffusion coefficient and the mean square displacement is:

[0026]

[0027] According to formula (3), the diffusion coefficients of different elements in the alloy matrix elements can be fitted based on the atomic mean square displacement simulation results.

[0028] Preferably, in step 3, the Gibbs-Thompson coefficient is calculated by the following formula:

[0029]

[0030] In formula (4), Γ is the Gibbs-Thompson coefficient, in m·k; T is the melting temperature of the alloy with uniform composition, in K; σ is the surface tension of the solid-liquid interface of the alloy, in J / m 2; L is the latent heat of phase change of the alloy, in J / Kg; ρ is the density of the alloy, in Kg / m 3 The values ​​of the phase change latent heat and density at a certain temperature are calculated based on Thermo-Calc.

[0031] Preferably, the melting temperature T is calculated by the built-in material calculation module of Procast; the alloy solid-liquid interface surface tension σ is calculated based on the following equation:

[0032]

[0033] In formula (5), ΔG is the change in Gibbs free energy, and A is the area of ​​the solid-liquid interface.

[0034] Preferably, in step 2, the nucleation density is calculated by the following relationship:

[0035]

[0036] In formula (6), N V 、N A 、N L They are the number of grains per unit volume, the number of grains per unit area, and the number of grains per unit measuring line length.

[0037] Preferably, in step 4, the growth kinetics parameter is calculated as follows:

[0038]

[0039] In formula (7) and formula (8), a2 and a3 are coefficients for calculating the growth rate of the dendrite tip in the KGT model, m is the liquidus slope, C0 is the initial composition of the alloy, and k is the solute equilibrium distribution coefficient; wherein, m and k are calculated based on the Procast built-in material calculation module.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The present invention combines molecular dynamics, thermodynamic calculations, experiments and other means, starting from the principle of cellular automaton simulation of grain growth, proposes to calculate the diffusion coefficients of various elements in the alloy in the base element at different temperatures, and calculate the Gibbs-Thompson coefficient based on thermodynamics, thereby quickly obtaining high-precision dendrite growth kinetic parameters. In addition, by performing statistical analysis on the EBSD characterization results, the exact nucleation parameters of this type of alloy are determined. By comparing the simulation under default conditions, high-precision casting microstructure simulation can be achieved, thereby achieving effective guidance for process optimization and actual production. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 The present invention provides a simulation technology roadmap for a method of improving the simulation accuracy of casting microstructure.

[0043] Figure 2 This is a schematic diagram of the Lammps construction model described in the present invention.

[0044] Figure 3 This is a schematic diagram of the fitting of atomic mean square displacement and diffusion coefficient described in the present invention.

[0045] Figure 4 This is a schematic diagram of the Lammps mean square displacement simulation program of the present invention.

[0046] Figure 5 For the present invention Schematic diagram of parameter setting and calculation for simulated growth dynamics.

[0047] Figure 6 For the present invention Schematic diagram of the simulation nucleation parameter settings.

[0048] Figure 7 The default parameters and high-precision parameters of the present invention Schematic diagram of simulation results comparison; A is the high-precision parameter obtained by the method of the present invention Simulation results, B is under default parameter conditions Simulation results.

[0049] Figure 8 Schematic diagram of EBSD characterization and data processing results of Example 1 of the present invention; wherein A is the result diagram of the unprocessed image, and B is the result diagram after processing using the method of the present invention.

[0050] Fig. 9 Schematic diagram of the die-casting test specimen casting of Example 1 of the present invention; wherein, a is a front view of the die-casting test casting model, b is a side view of the die-casting test casting model, ① is a fluidity specimen, ② and ⑤ are both flat plate tensile specimens, ③ and ④ are both cylindrical tensile specimens, and ⑥ is a torque test specimen. DETAILED DESCRIPTION

[0051] The present invention will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the present invention belong to the scope of protection of the present invention.

[0052] Unless otherwise indicated in specific cases in the present invention, the numerical ranges listed herein include the upper and lower limits, and all integers and fractions within the range, and are not limited to the specific values ​​listed when defining the range.

[0053] 1. A method to improve the simulation accuracy of casting microstructure

[0054] Step 1: Based on Lammps, an atomic mean square displacement model of binary alloys is constructed and molecular dynamics simulation is performed to obtain the diffusion coefficients of different elements in the alloy matrix element;

[0055] Step 2: Perform electron backscatter diffraction (EBSD) characterization on the test sample, filter the results, remove the noise, and then use the area weighting method to count the grain size and number in the area to obtain the number of planar grains. At the same time, determine the body nucleus density and surface nucleus density of the alloy based on the algorithm recommended by ASTM;

[0056] Step 3: Obtain the Gibbs-Thompson coefficient through thermodynamic calculation;

[0057] Step 4: Import the high-precision simulation parameters obtained in Steps 1 and 3 into the Procast-based The calculation module combines the Procast built-in material calculation module to complete the calculation of the dendrite growth kinetic parameters of the KGT model, and sets the nucleation parameters obtained in step 2 to Boundary conditions of the simulation;

[0058] Step 5: Use Procast to complete the setting of casting molding process parameters, boundary conditions, simulation parameters, and simulate the molding process to obtain the temperature field and flow field during the casting molding process, and then conduct Calculation is performed to finally obtain high-precision casting microstructure simulation results.

[0059] The present invention finds that in Procast casting finite element simulation software, The growth kinetics parameters and nucleation parameters of the model in the module are officially set as unified default parameters by Procast. This processing method reduces the accuracy of the simulation from the source, and the modification of a series of simulation parameters such as nucleation parameters based on the default parameters can only be used for qualitative analysis. The significance of microstructure simulation is not significant and cannot provide effective guidance for actual production. In order to solve this technical problem, the present invention is based on the principle of CA cellular automaton to simulate grain growth, and focuses on improving the accuracy of the simulated crystal growth kinetics parameters and nucleation parameters. Figure 1 As shown in the technical roadmap, the present invention is based on Procast finite element simulation software The module improves the accuracy of the growth kinetic parameters and nucleation parameters in CAFE microstructure simulation by means of lammps simulation, thermodynamic calculation and experiment. Combined with the casting temperature field results obtained by Procast process simulation, high-precision casting microstructure simulation can be achieved. Among them, the growth kinetic parameters for improving the accuracy proposed in the present invention are based on the KGT dendrite growth kinetic model. The model describing the growth velocity of the dendrite tip in the KGT model is as follows:

[0060] v(ΔT)=a2ΔT 2 +a3ΔT 3 (9)

[0061] In formula (9), a2(m / s·K 3 ) and a3(m / s·K 2 ) is the growth kinetics parameter. In order to improve the accuracy of this coefficient, the parameters that need to be paid attention to in the present invention are the diffusion coefficient of different elements in the matrix alloy and the Gibbs-Thompson coefficient.

[0062] In step 1, the diffusion coefficient of different elements in the alloy matrix element is set to 3e-9 by default in the Procast casting software. This default value is neither representative nor truly reflects the actual production situation. Figures 2-3 As shown, the method for improving the parameter accuracy of the present invention is based on lammps software, constructing a diffusion model of a binary alloy composed of an alloy base element and a desired element, performing atomic mean square displacement simulation based on molecular dynamics, and calling out the simulation results after the calculation is completed, fitting the atomic mean square displacement, and thereby obtaining the diffusion coefficient of a certain element in the alloy base element.

[0063] In some embodiments of the present invention, in step 1, Figure 4 As shown, the specific steps for constructing the atomic mean square displacement model based on Lammps are as follows:

[0064] S1: Set basic information such as model units, boundary types, and atom types;

[0065] S2: Set the area for atomic distribution simulation, including setting the range of the atomic distribution area, merging the two areas into a box area, filling the matrix atom A and the atom B to be solved, and setting the molar masses of the two atoms respectively. After that, divide the boundary area and fix the left and right boundaries of the box area;

[0066] S3: Set up the force field, thermodynamic output information, and energy minimization model;

[0067] S4: Set the temperature initialization parameters, relaxation parameters and heat preservation and pressure diffusion parameters. The data file thus formed contains the basic information of the atomic mean square displacement molecular dynamics model. Combined with the .pot potential energy file and the .in script file used, simulation calculations can be performed based on Lammps. Among them, in S4, the calculation and result output of the atomic mean square displacement MSD are set during the heat preservation and pressure diffusion parameter setting process and saved as a .data file. After the calculation is completed, the diffusion coefficient can be fitted based on the data in this file in combination with EXCEL, MATLAB, Origin and other software. The potential energy setting selects the EAM potential function, and the energy minimization model uses the conjugate gradient descent method. The operating conditions for stopping the minimization are that the energy tolerance, force tolerance and number of iterations reach the corresponding values.

[0068] In some embodiments of the present invention, in step 1, the diffusion coefficients of atoms of different elements in the alloy matrix element are obtained using the atomic mean square displacement model and are specifically calculated as follows:

[0069]

[0070] In formula (1), Dsim is the atomic diffusion coefficient, To simulate the system dimension;

[0071] The mean square displacement of an atom is defined as:

[0072]

[0073] In formula (2), MSD is the mean square displacement of atoms, r i is the atomic position, N is the total number of atoms;

[0074] From equations (1) and (2), the relationship between the atomic diffusion coefficient and the mean square displacement is:

[0075]

[0076] According to formula (3), the diffusion coefficients of different elements in the alloy matrix elements can be fitted based on the atomic mean square displacement simulation results.

[0077] In some embodiments of the present invention, in step 3, the Gibbs-Thompson coefficient is calculated by the following formula:

[0078]

[0079] In formula (4), Γ is the Gibbs-Thompson coefficient, in m·k; T is the melting temperature of the alloy with uniform composition, in K; σ is the surface tension of the solid-liquid interface of the alloy, in J / m 2; L is the latent heat of phase change of the alloy, in J / Kg; ρ is the density of the alloy, in Kg / m 3 The melting temperature T is calculated by the built-in material calculation module of Procast; the values ​​of the latent heat of phase change and density at a certain temperature are calculated based on Thermo-Calc; the surface tension σ of the alloy solid-liquid interface is calculated based on the following equation, which is used to describe the change in surface tension caused by the difference in intermolecular interaction energy at the solid-liquid interface.

[0080]

[0081] In formula (5), ΔG is the change in Gibbs free energy, and A is the area of ​​the solid-liquid interface.

[0082] In some embodiments of the present invention, step 2 is specifically as follows: Figure 8 As shown, in In the model simulating grain growth, the nucleation density can be analyzed based on the EBSD (electron backscatter diffraction) technology to analyze the crystal of the material. The orientation difference is determined to be a certain threshold, and the detection area is divided into different grains. The EBSD results are analyzed and counted with the help of OIM Analysis, Channel 5, AztecCrystal and other software to determine the number of grains in a certain area. Multiple samples from different areas of the casting are then taken for EBSD characterization, and the nucleation density is determined based on the ASTM recommended algorithm.

[0083] In step 4, the nucleation density is calculated using the following relationship:

[0084]

[0085] In formula (6), N V 、N A 、N L They are the number of grains per unit volume, the number of grains per unit area, and the number of grains per unit measurement line length, respectively. Therefore, the body nucleus density and the surface nucleus density can be determined simultaneously based on the number of planar grains counted by EBSD results.

[0086] In some embodiments of the present invention, step 2 specifically includes: based on the EBSD characterization results, the size and number of grains in a certain area are counted. Considering that the image quality is different and there are areas with poor quality, the present invention proposes that the results need to be filtered before statistics to remove noise and improve the credibility of the EBSD characterization results. The specific method is to set a certain threshold pixel count, filter low-precision measurement results, and for results with both large and small grains, due to the large standard deviation of the grain size, the results are counted in an area-weighted manner to obtain more realistic statistical results.

[0087] In step 4, if Figure 5 As shown, the growth kinetic parameters a2 and a3 can be automatically calculated based on the Procast built-in calculation module. The calculation formula is as follows:

[0088]

[0089] In formula (7) and formula (8), a2 and a3 are coefficients for calculating the growth rate of the dendrite tip in the KGT model, m is the liquidus slope, C0 is the initial composition of the alloy, and k is the solute equilibrium distribution coefficient; wherein, m and k are calculated based on the Procast built-in material calculation module.

[0090] 2. Examples and Comparative Examples

[0091] Embodiment 1:

[0092] See also Fig. 9 The simulation object is a die-casting test casting of a certain Al-Si alloy, which specifically includes a fluidity test specimen (1), a flat tensile specimen (2, 5), a cylindrical tensile specimen (3, 4), and a torque test specimen (6). The simulation process adopts vacuum high-pressure die-casting.

[0093] According to the technical route of the present invention, considering that the pouring temperature in the process parameters is set to 700°C, the diffusion coefficients of different elements in aluminum at 700°C are first calculated, and the steps are as follows:

[0094] In the first step, the mean square displacement simulation models of alloy elements and Al matrix elements were constructed respectively, and then molecular dynamics simulation calculations were performed to export the atomic mean square displacement MSD data in the .data log file. Finally, based on the Origin software, the time step was used as the horizontal coordinate and the atomic mean square displacement result was used as the vertical coordinate for fitting to obtain the atomic diffusion coefficient. The results are as follows:

[0095] Table 1

[0096] element Element mass fraction (%) Diffusion coefficient in aluminum at 700℃ Si 9.033 9.417e-09 Fe 0.097 5.567e-09 Cu 0.009 8.167e-09 Mg 0.229 5.601e-09 Mn 0.510 8.683e-09 Ti 0.135 2.933e-09 Cr 0.001 1.717e-09 Ni 0.004 3.117e-09 Sn 0.002 1.215e-09 Zn 0.015 2.451e-09

[0097] The second step is to obtain a high-precision alloy Gibbs-Thompson coefficient based on the aforementioned thermodynamic calculation principle, which specifically includes calculating the alloy thermodynamic parameters based on Thermo-Calc to obtain the phase change latent heat and density of the target composition alloy, followed by calculating the surface tension of the alloy, and finally obtaining the Gibbs-Thompson coefficient Γ = 6.91446e-7m·k.

[0098] The third step is the calculation of nucleation parameters, see Figure 8 ,in Figure 8A is the result of unprocessed image processing, where a large number of impurity pixels can be observed, which has an adverse effect on the statistics of subsequent results. Figure 8 B is the result of using the above processing method. Figure 8 A, impurity pixels are effectively removed. The distribution and number of grains in a certain area can be obtained by processing the EBSD results. There are 79 grains in the area shown in the figure. The EBSD results of metallographic samples in other parts are processed in the same process, and the body core density is determined to be 6e+13 (1 / m 3 ), the surface nucleus density is 1.8e+9(1 / m 2 ).

[0099] The fourth step is to set the parameters of die-casting finite element simulation after completing the geometric model processing and meshing in Procast pre-processing.

[0100] See also Figure 4 First, calculate the material thermal properties based on the Procast built-in material calculation module in the material setting interface, then modify the default parameters to the calculated atomic diffusion coefficient and Gibbs-Thompson coefficient, and finally click compute Complete the calculation of growth kinetic parameters a2 and a3, such as Figure 4 .

[0101] a2=7.145e-7m / s·K 2 , a3=3.034e-7m / s·K 3 .

[0102] See also Figure 6 , complete the material setting, and continue to set the interface heat transfer coefficient and process parameters. The nucleation parameters are set in the process parameter setting link, and the nucleation undercooling, nucleation undercooling standard deviation and nucleation density of the body nucleation parameters and surface nucleation parameters are set respectively.

[0103] Regarding the simulation operating parameters, based on the die casting simulation, it is necessary to The module sets the grid parameters for the cellular automaton calculation, and finally sets the number of calculation cores and checks the sub-step, that is, The solver starts running.

[0104] See also Figure 6 , Figure 7 , Figure 8 After the run is completed, the result is post-processed based on the Procast viewer module. The results are taken from Fig. 9 The middle position of the No. 3 tensile specimen of the die-cast specimen shown corresponds to Figure 8EBSD characterization position. After conversion, the simulated grain size is 36.54μm, and the grain equivalent circle diameter area weighted value of EBSD characterization statistics is 35.3μm. The simulation results are in good agreement with the measured experimental results. Therefore, it can be considered that the proposed method of the present invention has good feasibility and practicality, can effectively improve the accuracy of simulation, and has good application prospects.

[0105] Comparative Example 1

[0106] This comparative example is used to illustrate the differences between the growth kinetic parameters of the KGT model calculated based on the method proposed in the present invention and the parameters under default conditions, and to show the comparison of simulation results under the two parameter conditions.

[0107] Based on Example 1, the test sample used is exactly the same as that of Example 1. The diffusion coefficient obtained by the method proposed in the present invention is shown in Table 1. The atomic diffusion coefficient under default conditions in Comparative Example 1 is uniformly 3e-9; the Gibbs-Thompson coefficient Γ=6.91446e-7m·k calculated by the present invention based on the thermodynamic principle, while the Gibbs-Thompson coefficient under default conditions in Comparative Example 1 is 2e-7m·k. The growth kinetic parameters of the KGT model calculated thereby are compared as shown in the following table:

[0108] Table 2

[0109]

[0110] See also Figure 7 , Comparative Example 1, under default conditions, parameters The grain equivalent circle diameter size obtained by simulation calculation is 48.7 μm, which is much larger than the actual grain size. The grain equivalent circle diameter size calculated by high-precision growth kinetic parameters and nucleation parameters according to Example 1 is 36.54 μm, which is close to the actual size. Therefore, it can be considered that the method proposed by the present invention significantly improves The accuracy of simulation can provide effective guidance for actual production.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit the technical solution. Those skilled in the art should understand that those modifications or equivalent substitutions of the technical solution of the present invention that do not depart from the purpose and scope of the technical solution should be included in the scope of the claims of the present invention.

Claims

1. A method for improving the simulation accuracy of casting microstructure, characterized in that: The specific steps include: Step 1: Based on Lammps, an atomic mean square displacement model of binary alloys is constructed and molecular dynamics simulation is performed to obtain the diffusion coefficients of different elements in the alloy matrix element; Step 2: Perform electron backscatter diffraction characterization on the test sample, filter the results, remove the noise, and then use the area weighting method to count the size and number of grains in the area to obtain the number of planar grains. At the same time, determine the body nucleus density and surface nucleus density of the alloy based on the algorithm recommended by ASTM; Step 3: Obtain the Gibbs-Thompson coefficient through thermodynamic calculation; Step 4: Import the high-precision simulation parameters obtained in Steps 1 and 3 into the Procast-based The calculation module combines the Procast built-in material calculation module to complete the calculation of the dendrite growth kinetic parameters of the KGT model, and sets the nucleation parameters obtained in step 2 to Boundary conditions of the simulation; Step 5: Use Procast to complete the setting of casting molding process parameters, boundary conditions, simulation parameters, and simulate the molding process to obtain the temperature field and flow field during the casting molding process, and then conduct Calculation is performed to finally obtain high-precision casting microstructure simulation results.

2. The method according to claim 1, characterized in that: In step 1, the specific steps of building an atomic mean square displacement model based on Lammps are as follows: S1: Set the model unit, boundary type, atom type, and time step; S2: Set the area for atomic distribution simulation, including setting the range of the atomic distribution area, merging the two areas into a box area, filling the matrix atom A and the atom B to be solved, and setting the molar masses of the two atoms respectively. After that, divide the boundary area and fix the left and right boundaries of the box area; S3: Set up the force field, thermodynamic output information, and energy minimization model; S4: Set the temperature initialization parameters, relaxation parameters, and heat and pressure holding diffusion parameters. The resulting .data file contains the basic information of the atomic mean square displacement molecular dynamics model. Combined with the .pot potential energy file and .in script file used, simulation calculations can be performed based on Lammps.

3. The method according to claim 2, characterized in that: In S3, the energy minimization model uses the conjugate gradient descent method, and the operating conditions for stopping the minimization are that the energy tolerance, force tolerance and the number of iteration steps reach corresponding values.

4. The method according to claim 2, characterized in that: In S4, during the setting of the heat preservation and pressure preservation diffusion parameters, the calculation and result output of the atomic mean square displacement MSD are set, the diffusion coefficient is fitted according to the data after the calculation is completed, and the EAM potential function is selected for the setting of the potential energy.

5. The method according to claim 1, characterized in that: In step 1, the atomic mean square displacement model is used to obtain the diffusion coefficients of atoms of different elements in the alloy matrix element. The specific calculation is as follows: In formula (1), Dsim is the atomic diffusion coefficient, To simulate the system dimension; The mean square displacement of an atom is defined as: In formula (2), MSD is the mean square displacement of atoms, r i is the atomic position, N is the total number of atoms; From equations (1) and (2), the relationship between the atomic diffusion coefficient and the mean square displacement is: According to formula (3), the diffusion coefficients of different elements in the alloy matrix elements can be fitted based on the atomic mean square displacement simulation results.

6. The method according to claim 1, characterized in that: In step 3, the Gibbs-Thompson coefficient is calculated by the following formula: In formula (4), Γ is the Gibbs-Thompson coefficient, in m·k; T is the melting temperature of the alloy with uniform composition, in K; σ is the surface tension of the solid-liquid interface of the alloy, in J / m 2 ; L is the latent heat of phase change of the alloy, in J / Kg; ρ is the density of the alloy, in Kg / m 3 The values ​​of the phase change latent heat and density at a certain temperature are calculated based on Thermo-Calc.

7. The method according to claim 6, characterized in that: The melting temperature T is calculated by the built-in material calculation module of Procast; the alloy solid-liquid interface surface tension σ is calculated based on the following equation: In formula (5), ΔG is the change in Gibbs free energy, and A is the area of ​​the solid-liquid interface.

8. The method according to claim 1, characterized in that: In step 2, the nucleation density is calculated using the following ASTM recommended calculation method: In formula (6), N V 、N A 、N L They are the number of grains per unit volume, the number of grains per unit area, and the number of grains per unit measuring line length.

9. The method according to claim 1, characterized in that: In step 4, the growth kinetics parameters are calculated as follows: In formula (7) and formula (8), a2 and a3 are coefficients for calculating the growth rate of the dendrite tip in the KGT model, m is the liquidus slope, C0 is the initial composition of the alloy, and k is the solute equilibrium distribution coefficient; wherein, m and k are calculated based on the Procast built-in material calculation module.

Images