Method for simulating three-dimensional microstructure evolution during arc welding weld pool solidification process

Abstract

This invention discloses a method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding. Specifically, it involves: Simulating the macroscopic temperature field of the welding process of T-joints in thin-walled aluminum alloy parts using the finite element method to obtain local temperature field data; interpolating the local temperature field data with time using linear interpolation to obtain microcellular temperature field data; establishing a three-dimensional dendritic growth model of the weld pool to calculate the nucleation location, growth rate, and solidity of grains in the weld pool; establishing a solute concentration field distribution model to calculate solute distribution and diffusion during the growth of interface cells; and establishing a spatial cell trapping model based on the eccentric octahedral algorithm to complete the three-dimensional microstructure evolution simulation of the weld pool solidification process. This invention employs a method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding, enabling the simulation of the three-dimensional microstructure morphology and solute concentration distribution of the weld pool during solidification. This is of great significance for a more comprehensive description of the weld pool solidification process and for improving the mechanical properties of welded joints.

Description

Technical Field

[0001] This invention belongs to the field of arc welding, specifically relating to a method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding. Background Technology

[0002] Welding is an important advanced manufacturing technology, widely used for joining and assembling thin-walled aluminum alloy parts due to its high stability and fast forming speed. During welding, the temperature in the molten pool changes rapidly, undergoing complex physical processes such as melting, solidification, and heat conduction. These processes significantly influence the microstructure and macroscopic properties of the weld joint. However, traditional sensors and detection equipment struggle to achieve real-time control and monitoring of the welding process, making it difficult to establish the correlation between transient temperature fields and the three-dimensional dendrite growth process, the microstructure morphology of the weld metal, and the performance of the weld joint.

[0003] Existing methods for simulating the microstructure of weld metal solidification processes are mainly divided into deterministic methods, probabilistic methods, and phase-field methods. Deterministic models based on solidification kinetics cannot simulate the random nucleation and random crystal orientation during dendrite growth. While phase-field methods can describe dendrite morphology, their high computational cost makes it difficult to overcome the drawbacks of small computational domains and low computational efficiency. In contrast, the cellular automata method within probabilistic methods is simple and intuitive, and can effectively describe the randomness of nucleation sites and crystal orientations, isometric crystal morphology in complex temperature fields, and larger-scale simulation domains.

[0004] Patent 201911212284.X discloses a method for simulating the solidification process of BGA tin-lead solder balls based on cellular automata. This patent can simulate dendrite growth under different supercooling conditions, but it cannot achieve arbitrary preferred orientation and cannot reveal the actual formation process of the microstructure. Patent 202311368557.6 discloses a method for simulating the microstructure of arc welds based on cellular automata. This patent can simulate the two-dimensional microstructure of the weld, but it cannot comprehensively describe the growth of dendrites in all dimensions, and its grain morphology and solute field still have a large gap with the actual situation. Summary of the Invention

[0005] The purpose of this invention is to provide a three-dimensional microstructure evolution simulation method for the solidification process of arc welding weld pool, which solves the problems in the prior art that the model cannot reflect the three-dimensional microstructure morphology and evolution process of arbitrary orientation, and the problem of non-conservation of three-dimensional solute field.

[0006] The technical solution of this invention is: a three-dimensional microstructure evolution simulation method for the solidification process of the weld pool in arc welding. During the welding process, under the influence of a heat source, a portion of the aluminum alloy thin-walled part above the liquidus temperature of the weld metal melts to form a weld pool. As the temperature decreases, the liquid phase metal in the weld pool gradually solidifies with the growth of dendrites, forming a new microstructure. The steps are as follows:

[0007] Step 1: Simulate the macroscopic temperature field of the welding process of T-joints for thin-walled aluminum alloy parts based on the finite element method, obtain local temperature field data, and proceed to Step 2.

[0008] Step 2: Based on the linear interpolation method, interpolate the local temperature field data with time to obtain the micro-cell temperature field data. Define the cells above the liquidus temperature of the material as liquid phase cells and proceed to step 3.

[0009] Step 3: Based on the non-uniform continuous nucleation model, randomly determine the nucleation location within the computational region, and proceed to Step 4.

[0010] Step 4: Based on the Gaussian distribution nucleation density model, determine whether the nucleation cell at the current total supercooling is nucleating. When the total supercooling is greater than the critical supercooling, its state changes from liquid phase cell to nucleation cell, and proceed to step 5.

[0011] Step 5: Based on solute flux balance, calculate the velocity magnitude and normal direction of the growth interface in the nucleation cell or interface cell, and then calculate the length of the cell covering the interface along the normal direction. With solid fraction increment Δf s To establish a three-dimensional molten pool dendrite growth model, proceed to step 6.

[0012] Step 6: Based on solute conservation and Fick's law, calculate the solute distribution and diffusion during the growth process of interface cells to establish a solute concentration field distribution model, and then proceed to Step 7.

[0013] Step 7: Based on the eccentric octahedral algorithm, establish a spatial cell trapping model, and calculate the three-dimensional dendrite growth orientation according to Euler angles. Transform the neighboring cell state at the vertex of the octahedron into a liquid phase cell. When the solid fraction f of this cell... s When step 1 is reached, the cells grow into solid-phase cells and capture stops. Proceed to step 8.

[0014] Step 8: Combine the non-uniform continuous nucleation model, the three-dimensional molten pool dendrite growth model, the solute concentration field distribution model, and the spatial cell trapping model, and repeat steps 3 to 7 until both the growth-state cells and liquid-phase cells in the simulation area are transformed into solid-phase cells, thereby simulating the three-dimensional microstructure of the weld metal.

[0015] Compared with the prior art, the significant advantages of this invention are:

[0016] (1) The three-dimensional molten pool dendrite growth model established in this invention can accurately calculate the length of the growing cells covered along the interface normal direction, reflect the real three-dimensional microstructure morphology, and maintain the solute field conservation during the simulation of three-dimensional microstructure evolution.

[0017] (2) The present invention uses Euler angles to represent crystal orientation and an eccentric octahedron capture algorithm, which can accurately represent the direction vector of crystal orientation in the local coordinate system, realize the growth simulation of three-dimensional dendrites with arbitrary preferred orientation, accurately calculate the distance through the cell along each crystal direction, and accurately calculate the global coordinates of the eccentric octahedron vertex. Attached Figure Description

[0018] Figure 1 This is a flowchart of a method for simulating the evolution of three-dimensional microstructure in the solidification process of arc welding weld pool, as described in this invention.

[0019] Figure 2 This is a schematic diagram of the growth of a regular hexahedral cell in an embodiment of the present invention.

[0020] Figure 3 This describes the evolution of the concentration field distribution of a single equiaxed crystal in an embodiment of the present invention.

[0021] Figure 4 This is a schematic diagram of the eccentric regular octahedral algorithm capturing cells in an embodiment of the present invention.

[0022] Figure 5 This is a schematic diagram of the three-dimensional dendrite growth orientation calculated based on Euler angles in an embodiment of the present invention.

[0023] Figure 6 These are individual equiaxed crystals with different preferred orientation angles in the embodiments of the present invention.

[0024] Figure 7 This is a three-dimensional microstructure diagram simulating the solidification process of the weld pool in an embodiment of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0026] Furthermore, the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are feasible to those skilled in the art. If the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0027] The following section will further introduce the specific implementation method, as well as the technical difficulties and inventive points of this invention, using this design example as an example.

[0028] When aluminum alloy thin-walled parts are welded to T-joints, they are affected by high-temperature heat sources. Areas above the liquidus temperature of the weld metal melt to form a molten pool. As the temperature decreases, the liquid phase metal in the molten pool begins to nucleate, and gradually solidifies as dendrites grow, forming the microstructure of the weld area. This invention uses a three-dimensional cellular automata method to simulate the microstructure evolution of the molten pool solidification process. A local area of ​​the weld is selected and divided into micro-cells. Initially, all cells are in a liquid phase state. Cells at nucleation sites nucleate and gradually grow, capturing neighboring cells as liquid phase cells according to a spatial cell capture model. Liquid phase cells and nucleated cells continue to grow and be captured. Cells that have completed growth become solid phase cells. This cycle continues until all cells have become solid phase cells, ultimately completing the three-dimensional microstructure simulation of weld solidification.

[0029] Combination Figure 1 The present invention provides a method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding, comprising the following steps:

[0030] Step 1: During welding, the metal is affected by a high-temperature heat source. Regions above the liquidus temperature of the weld metal melt to form a molten pool. As the temperature decreases, the liquid metal solidifies to form the weld. Based on the finite element method, a simulation model of the welding process is established to simulate the macroscopic temperature field of the T-joint of the aluminum alloy thin-walled part. Data on the temperature field changes over time in local areas of the molten pool are extracted, and then proceed to Step 2.

[0031] Step 2: Based on linear interpolation, interpolate the local temperature field data with time to obtain the temperature field data of the local region of the macroscopic molten pool corresponding to the microscopic cell calculation region at the microscopic time step, i.e., the microscopic cell temperature T. loc Proceed to step 3.

[0032] Step 3: Based on the non-uniform continuous nucleation model, randomly determine the nucleation location within the computational domain, as follows:

[0033] The nucleation probability P of a local region of the molten pool per unit time step v Calculated using the following formula:

[0034]

[0035] In the formula, δNv To simulate the total number of nucleated cells per unit time step within the region, N CA δn represents the number of cells in a local region of the molten pool. v V represents the increase in nucleation density in a local region of the molten pool per unit time step. CA The volume of a single microcell.

[0036] Step 4: Based on the Gaussian distribution nucleation density model, determine whether the nucleation cell at the current total undercooling has nucleated. When the total undercooling is greater than the critical undercooling, its state changes from liquid phase cell to nucleation cell, as follows:

[0037] The nucleation density model based on Gaussian distribution is used to determine whether nucleation has occurred. The expression is as follows:

[0038]

[0039] In the formula, n v For nucleation density, δn v Let ΔT be the nucleation density increment in a local region of the molten pool per unit time step, ΔT be the total undercooling, and δ(ΔT) be the undercooling increment per unit time step. Let n be the nucleation density distribution function. max For the maximum nucleation density, ΔT N ΔT is the average nucleation undercooling. σ ΔT′ represents the standard deviation undercooling; ΔT′ represents the nucleation undercooling.

[0040] The calculation of total subcooling includes thermal subcooling. Composition undercooling and curvature undercooling ΔT r =ΓK, its calculation formula is:

[0041]

[0042] In the formula, T is the liquidus temperature at the initial liquid solute concentration. loc Where is the temperature of the microscopic cell, m is the slope of the liquidus line, and C0 is the initial concentration of the solute in the liquid phase. Γ represents the solute concentration at the liquid phase equilibrium at the solid-liquid interface, Γ is the Gibbs-Thompson coefficient, and K is the interface curvature.

[0043] Step 5: Based on solute flux balance, calculate the velocity magnitude and normal direction of the growth interface in the nucleation cell or interface cell, and then calculate the length of the cell covering the interface along the normal direction. With solid fraction increment Δf s To establish a three-dimensional model of molten pool dendrite growth, the following is a detailed description:

[0044] Step 5.1: Based on solute flux balance, calculate the velocity magnitude and normal direction of the growth interface in the nucleating cell or interface cell:

[0045] During dendrite growth, the supercooling at the solid-liquid interface reaches equilibrium, from which the equilibrium concentration of the solute in the liquid phase can be calculated.

[0046]

[0047] The equilibrium concentration of the solute in the solid phase can be obtained from the local phase equilibrium at the solid-liquid interface. for:

[0048]

[0049] In the formula, k0 is the solute equilibrium distribution coefficient.

[0050] The dendrite interface growth rate is calculated based on solute flux balance, and its expression is as follows:

[0051]

[0052] In the formula, D is the normal velocity at the dendrite growth interface. L D is the diffusion coefficient of the solute in the liquid phase. S C is the diffusion coefficient of the solid solute. L C represents the concentration of the solute in the liquid phase. S This refers to the concentration of the solid phase solute. This is the normal vector of the growth interface.

[0053] The dendrite growth rate can be expressed as a finite difference form along the x, y, z axes using solidity:

[0054]

[0055]

[0056] Where i, j, and k represent element indices.

[0057] The interface curvature for three-dimensional dendrite growth is calculated as follows:

[0058]

[0059] In the formula, ε is the interfacial energy anisotropy parameter, (n x ,n y ,n z () represents the normal vector of the unit interface. It is the first partial derivative of the solid fraction in all directions.

[0060] Step 5.2: Based on the growth interface velocity and normal direction, calculate the length of the cell covered along the interface normal direction. like Figure 2 As shown, with respect to the solid fraction increment Δf s :

[0061] For the overgrown cells, i.e., the interface cells and nucleation cells, a plane parallel to the solid-liquid interface is drawn from the i-th vertex. Let d be the distance from the cell center to this plane. i Its expression is:

[0062]

[0063] Where (x0, y0, z0) are the center coordinates of the cell in the local coordinate system, (x i ,y i ,z i ) represents the coordinates of the i-th vertex of the cell in the local coordinate system.

[0064] The length of the cell covered along the interface normal direction for:

[0065]

[0066] The increment of solid fraction per unit time step is:

[0067]

[0068] In the formula, GF is the shape factor, Δt is the time increment, and V n This represents the dendrite interface growth rate.

[0069] The GF calculation rules are as follows:

[0070] (1) When there is a solid cell in the nearest neighbor cell of the growing cell, GF = 1;

[0071] (2) When there is a solid-phase cell in the second nearest neighbor cell of the growing cell.

[0072] (3) When a solid-phase cell exists in the third nearest neighbor cell of the growing cell.

[0073] Step 6: Based on solute conservation and Fick's law, calculate solute distribution and diffusion during the interfacial cell growth process to establish a solute concentration field distribution model, such as... Figure 3 As shown, the details are as follows:

[0074] When the solid fraction of a growing cell increases, if a liquid-adjacent cell exists in its neighborhood, it will expel a solute ΔC, which can be calculated as follows:

[0075]

[0076] In the formula, Δf represents the concentration of the liquid solute before the cellular state transition. s The solid fraction increment is given by k, and k0 is the solute equilibrium partition coefficient. This refers to the concentration of the liquid solute after growth. f represents the solute concentration at the liquid phase equilibrium at the solid-liquid interface. s denoted as solid fraction.

[0077] If the concentration of the liquid phase solute in adjacent cell j is... The solute concentration at the solid-liquid interface of the liquid phase equilibrium is less than that of the solute cells. Excluding the solute that will be allocated to this cell, the expression is:

[0078]

[0079] In the formula, Let be the concentration of the liquid solute in neighboring cell j at time t+1. Let be the concentration of the liquid solute in neighboring cell j at time t. Let N be the concentration of the liquid solute in the nb-th neighboring cell, and N be the total number of neighboring liquid cells.

[0080] Based on Fick's law, solute diffusion in three-dimensional microstructure evolution simulation satisfies the following equation:

[0081]

[0082]

[0083] In the formula, C L C represents the concentration of the solute in the liquid phase. S D represents the concentration of the solid phase solute. L D is the diffusion coefficient of the solute in the liquid phase. S is the diffusion coefficient of the solid solute.

[0084] Step 7: Based on the eccentric regular octahedron algorithm, establish a spatial cell capture model, such as... Figure 4 As shown, the three-dimensional dendrite growth orientation is calculated based on Euler angles, as follows. Figure 5 As shown, the neighboring cell state at the vertex of a regular octahedron is transformed into a liquid-phase cell. When the solid fraction f of this cell... s When the density reaches 1, the growth becomes a solid-phase cell and the trapping stops, thus establishing a three-dimensional molten pool dendrite growth model with arbitrary orientation, such as... Figure 6 As shown, the details are as follows:

[0085] Step 7.1: Calculate the three-dimensional dendrite growth orientation based on Euler angles:

[0086] For aluminum alloys with a face-centered cubic crystal structure, the vector representing the <100> crystal orientation after rotation can be represented by the rotation matrix R. T Calculate: In the Cartesian coordinate system XYZ, rotate the XOY plane around the Z-axis by an angle w1, then rotate the YOZ plane around the X-axis by an angle w2, and finally rotate the XOY plane around the Z-axis by an angle w3 to obtain a new coordinate system X'Y'Z', where the coordinate system XYZ and the vector (x,y,z) in the coordinate system X'Y'Z' are related. T With vector (x', y', z') T The conversion relationship between them is expressed as follows:

[0087]

[0088] Step 7.2: Based on the eccentric regular octahedron algorithm, establish a spatial cell capture model:

[0089] An eccentric regular octahedron is placed at the center of the nucleation cell. The three diagonals of the octahedron are parallel to the <100>, <010>, and <001> crystal directions of the face-centered cubic grain, respectively. The orientation vector of the crystal in the local coordinate system is u. ＜100＞ v ＜010＞ and w ＜001＞ They are respectively:

[0090] u ＜100＞ =[R 11 R 21 R 31 ]

[0091] v ＜010＞ =[R 12 R 22 R 32 ]

[0092] w ＜001＞ =[R 13 R 23 R 33 ]

[0093] Diagonal length L dia It will increase with the solid fraction increment Δf s The accumulation continues to grow, L dia The expression is:

[0094] L dia =∑Δf s L max

[0095] Maximum half-diagonal length L max The maximum distance that can be traversed through a unit cell along the crystal orientation is calculated as follows:

[0096] L max=max(L

[100] ,L

[010] ,L

[001] )

[0097] In the formula, L

[100] L

[010] and L

[001] These represent the distances traversed through the cell along the <100>, <010>, and <001> crystal directions, respectively, and their expressions are:

[0098]

[0099] An eccentric regular octahedron centered at (x0', y0', z0') has global coordinates (x0', y0', z0') at its k-th vertex. k ,y k ,z k Calculated using the following formula:

[0100]

[0101] In the formula, R1, R2, and R3 are the components of the crystal orientation direction vector along the X, Y, and Z axes, respectively, and L... R is the length of the crystal orientation direction vector.

[0102] As the diagonal grows, the vertex position of the eccentric octahedron will enter the neighboring cell. If it is a liquid phase cell, it will be captured as an interface cell, and a regular octahedron with the same deflection direction will be generated at the vertex position, thus establishing a three-dimensional dendrite growth space cell capture model.

[0103] Step 8: Combine the non-uniform continuous nucleation model, the three-dimensional molten pool dendrite growth model, the solute concentration field distribution model, and the spatial cell trapping model, and repeat steps 3 to 7 until both nucleation cells and liquid phase cells in the simulation area are transformed into solid phase cells, obtaining the three-dimensional microstructure morphology of the weld pool solidification, such as... Figure 7 As shown, the details are as follows:

[0104] After selecting a local area of ​​the macro weld as the microstructure simulation region, all cells are initialized to the liquid phase state. Based on the non-uniform continuous nucleation model, nucleation positions are randomly determined within the calculation region. Then, based on the Gaussian distribution nucleation density model, it is determined whether the nucleation position cell has nucleated under the current total undercooling. Then, based on the solute flux balance, the velocity magnitude and normal direction of the growth interface in the nucleation cell or interface cell are calculated. Then, the three-dimensional dendrite growth orientation is calculated according to Euler angles. Then, the length of the cell covering the interface normal direction and the solid fraction increment are calculated. Then, based on solute conservation and Fick's law, the solute distribution and diffusion during cell growth are calculated. Then, based on the spatial cell capture model of the eccentric octahedron algorithm, the distance through the cell along each crystal direction and the global coordinates of the eccentric octahedron vertex are calculated. Then, the growth of three-dimensional dendrites is calculated. The above process is repeated until the growth state cells and liquid phase cells in the simulation region are transformed into solid phase cells, and the three-dimensional microstructure morphology of the weld pool solidification is obtained.

Claims

1. A method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding, wherein, during the welding process, a portion of the aluminum alloy thin-walled part above the liquidus temperature of the weld metal melts to form a weld pool under the influence of a heat source; as the temperature decreases, the liquid phase metal in the weld pool gradually solidifies with the growth of dendrites, forming a new microstructure, characterized in that... The steps are as follows: Step 1: Simulate the macroscopic temperature field during the welding process of T-joints of thin-walled aluminum alloy parts using the finite element method, obtain local temperature field data, and proceed to Step 2; Step 2: Based on the linear interpolation method, interpolate the local temperature field data with time to obtain the micro-cell temperature field data. Define the cells above the liquidus temperature of the material as liquid phase cells and proceed to step 3. Step 3: Based on the non-uniform continuous nucleation model, randomly determine the nucleation location within the computational region, and proceed to Step 4; Step 4: Based on the Gaussian distribution nucleation density model, determine whether the nucleation cell at the current total supercooling is nucleated. When the total supercooling is greater than the critical supercooling, its state changes from liquid phase cell to nucleation cell, and proceed to step 5. Step 5: Based on solute flux balance, calculate the velocity magnitude and normal direction of the growth interface in the nucleation cell or interface cell, and then calculate the length of the cell covering the interface along the normal direction. With solid fraction increment Δf s To establish a three-dimensional molten pool dendrite growth model, proceed to step 6; Step 6: Based on solute conservation and Fick's law, calculate the solute distribution and diffusion during the growth process of interface cells to establish a solute concentration field distribution model, and then proceed to step 7. Step 7: Based on the eccentric octahedral algorithm, establish a spatial cell trapping model, and calculate the three-dimensional dendrite growth orientation according to Euler angles. Transform the neighboring cell state at the vertex of the octahedron into a liquid phase cell. When the solid fraction f of this cell... s Upon reaching step 1, the cell grows into a solid-phase cell and capture ceases, as detailed below: Step 7.1: Calculate the three-dimensional dendrite growth orientation based on Euler angles: For aluminum alloys with a face-centered cubic crystal structure, the vector representing the <100> crystal orientation after rotation can be represented by the rotation matrix R. T Calculate: In the Cartesian coordinate system XYZ, rotate the XOY plane around the Z-axis by an angle w1, then rotate the YOZ plane around the X-axis by an angle w2, and finally rotate the XOY plane around the Z-axis by an angle w3 to obtain a new coordinate system X'Y'Z', where the coordinate system XYZ and the vector (x,y,z) in the coordinate system X'Y'Z' are related. T With vector (x', y', z') T The conversion relationship between them is expressed as follows: Step 7.2: Based on the eccentric regular octahedron algorithm, establish a spatial cell capture model: An eccentric regular octahedron is placed at the center of the nucleation cell. The three diagonals of the octahedron are parallel to the <100>, <010>, and <001> crystal directions of the face-centered cubic grain, respectively. The orientation vector of the crystal in the local coordinate system is u. ＜100＞ v ＜010＞ and w ＜001＞ They are respectively: u ＜100＞ =[R 11 R 21 R 31 ] v ＜010＞ =[R 12 R 22 R 32 ] w ＜001＞ =[R 13 R 23 R 33 ] Diagonal length L dia It will increase with the solid fraction increment Δf s The accumulation continues to grow, L dia The expression is: L dia =∑Δf s L max Maximum half-diagonal length L max The maximum distance that can be traversed through a unit cell along the crystal orientation is calculated as follows: L max <max(L [100] ,L [010] ,L [001] ) In the formula, L [100] L [010] and L [001] These represent the distances traversed through the cell along the <100>, <010>, and <001> crystal directions, respectively, and their expressions are: The center coordinates of the eccentric regular octahedron are (x0', y0', z0'), and the global coordinates of the k-th vertex are (x0', y0', z0'). k ,y k ,z k Calculated using the following formula: In the formula, R1, R2, and R3 are the components of the crystal orientation direction vector along the X, Y, and Z axes, respectively, and L... R The length of the crystal orientation direction vector; As the diagonal grows, the vertex position of the eccentric regular octahedron will enter the neighboring cell. If it is a liquid phase cell, it will be captured as an interface cell and generate regular octahedrons with the same deflection direction at the vertex position, thereby establishing a spatial cell capture model for the growth of three-dimensional molten pool dendrites. Proceed to step 8; Step 8: Combine the non-uniform continuous nucleation model, the three-dimensional molten pool dendrite growth model, the solute concentration field distribution model, and the spatial cell trapping model, and repeat steps 3 to 7 until both the growth-state cells and liquid-phase cells in the simulation area are transformed into solid-phase cells, thereby simulating the three-dimensional microstructure of the weld metal.

2. The method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding as described in claim 1, characterized in that, In step 1, the macroscopic temperature field of the welding process of the T-joint of the thin-walled aluminum alloy part is simulated based on the finite element method to obtain local temperature field data, as follows: Based on the finite element method, a simulation model of the welding process was established to simulate the macroscopic temperature field of the T-joint of thin-walled aluminum alloy parts and extract data on the temperature field change over time in the local area of ​​the molten pool.

3. The method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding as described in claim 2, characterized in that, In step 2, based on the linear interpolation method, the local temperature field data is interpolated with time to obtain the microscopic cell temperature field data, as follows: Linear interpolation is performed over macroscopic time to obtain temperature field data at the microscopic time step, thus obtaining the temperature field data of the local region of the macroscopic molten pool corresponding to the microscopic cell computational region, i.e., the microscopic cell temperature T. loc .

4. The method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding as described in claim 3, characterized in that, In step 3, based on the non-uniform continuous nucleation model, the nucleation location is randomly determined within the computational domain, as follows: The nucleation probability P of a local region of the molten pool per unit time step v Calculated using the following formula: In the formula, δN v To simulate the total number of nucleated cells per unit time step within the region, N CA δn represents the number of cells in a local region of the molten pool. v V represents the increase in nucleation density in a local region of the molten pool per unit time step. CA The volume of a single microcell.

5. The method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding as described in claim 4, characterized in that, In step 4, based on the Gaussian distribution nucleation density model, it is determined whether the nucleation cell at the current total undercooling has nucleated. When the total undercooling is greater than the critical undercooling, its state changes from liquid phase cell to nucleation cell, as follows: The nucleation density model based on Gaussian distribution is used to determine whether nucleation has occurred. The expression is as follows: In the formula, n v For nucleation density, δn v Let ΔT be the nucleation density increment in a local region of the molten pool per unit time step, ΔT be the total undercooling, and δ(ΔT) be the undercooling increment per unit time step. Let n be the nucleation density distribution function. max For the maximum nucleation density, ΔT N ΔT is the average nucleation undercooling. σ ΔT′ represents the standard deviation undercooling; ΔT′ represents the nucleation undercooling. The calculation of total subcooling ΔT includes thermal subcooling. Composition undercooling and curvature undercooling ΔT r =ΓK, ΔT is calculated as follows: In the formula, T is the liquidus temperature at the initial liquid solute concentration. loc Where is the temperature of the microscopic cell, m is the slope of the liquidus line, and C0 is the initial concentration of the solute in the liquid phase. Γ represents the solute concentration at the liquid phase equilibrium at the solid-liquid interface, Γ is the Gibbs-Thompson coefficient, and K is the interface curvature.

6. The method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding as described in claim 5, characterized in that, In step 5, based on solute flux balance, the velocity magnitude and normal direction of the growth interface in the nucleating cell or interface cell are calculated, and then the length of the cell covering the interface along the normal direction is calculated. With solid fraction increment Δf s To establish a three-dimensional model of molten pool dendrite growth, the following is a detailed description: Step 5.1: Based on solute flux balance, calculate the velocity magnitude and normal direction of the growth interface in the nucleating cell or interface cell: During dendrite growth, the supercooling at the solid-liquid interface reaches equilibrium, from which the equilibrium solute concentration in the liquid phase at the solid-liquid interface can be calculated. The equilibrium concentration of the solute in the solid phase can be obtained from the local phase equilibrium at the solid-liquid interface. for: In the formula, k0 is the solute equilibrium partition coefficient; The dendrite interface growth rate is calculated based on solute flux balance, and the expression is: In the formula, D is the normal velocity at the dendrite growth interface. L D is the diffusion coefficient of the solute in the liquid phase. S C is the diffusion coefficient of the solid solute. L C represents the concentration of the solute in the liquid phase. S This refers to the concentration of the solid phase solute. The normal vector of the growth interface; The dendrite growth rate can be expressed as a finite difference along the X, Y, Z axes using solidity, i.e., V x (i1,j1,k1),V y (i1,j1,k1),V z (i1,j1,k1): Where i1, j1 and k1 represent element indices; The interface curvature K for three-dimensional dendrite growth is calculated as follows: In the formula, ε is the interfacial energy anisotropy parameter, (n x ,n y ,n z () represents the normal vector of the unit interface. Let Q be the first-order partial derivative of the solid fraction along the coordinate axes X, Y, Z, and Q be an intermediate variable. Step 5.2: Based on the growth interface velocity and normal direction, calculate the length of the cell covered along the interface normal direction. With solid fraction increment Δf s : For the overgrown cells, i.e., the interface cells and nucleation cells, a plane parallel to the solid-liquid interface is drawn from the i-th vertex. Let d be the distance from the cell center to this plane. i Its expression is: Where (x0, y0, z0) are the center coordinates of the cell in the local coordinate system, (x i ,y i ,z i () represents the coordinates of the i-th vertex of the cell in the local coordinate system; The length of the cell covered along the interface normal direction for: Increment of solid fraction Δf per unit time step s for: In the formula, GF is the shape factor, Δt is the time increment, and V n The dendrite interface growth rate; The GF calculation rules are as follows: 1) When a solid-phase cell exists in the nearest neighbor cell of a growing cell, GF = 1; 2) When a solid-phase cell exists in the second nearest neighbor cell of a growing cell, 3) When a solid-phase cell exists in the third nearest neighbor cell of a growing cell, 7. The method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding as described in claim 6, characterized in that, In step 6, based on solute conservation and Fick's law, the solute distribution and diffusion during the interfacial cell growth process are calculated to establish a solute concentration field distribution model, as follows: When the solid fraction of a growing cell increases, if a liquid-adjacent cell exists in its neighborhood, it will expel a solute ΔC, which can be calculated as follows: In the formula, Δf represents the concentration of the liquid solute before the cellular state transition. s The solid fraction increment is given by k, and k0 is the solute equilibrium partition coefficient. This refers to the concentration of the liquid solute after growth. f represents the solute concentration at the liquid phase equilibrium at the solid-liquid interface. s The solid fraction; If the concentration of the liquid phase solute in adjacent cell j is... The solute concentration at the solid-liquid interface of the liquid phase equilibrium is less than that of the solute cells. Excluding the solute that will be allocated to this cell, the expression is: In the formula, Let be the concentration of the liquid solute in neighboring cell j at time t+1. Let be the concentration of the liquid solute in neighboring cell j at time t. Let N be the concentration of the liquid phase solute in the nb-th neighboring cell, and N be the total number of neighboring liquid phase cells. Based on Fick's law, solute diffusion in three-dimensional microstructure evolution simulation satisfies the following equation: In the formula, C L C represents the concentration of the solute in the liquid phase. S D represents the concentration of the solid phase solute. L D is the diffusion coefficient of the solute in the liquid phase. S is the diffusion coefficient of the solid solute.

8. The method for simulating the three-dimensional microstructure evolution of the weld pool during the solidification process of arc welding as described in claim 7, characterized in that, In step 8, steps 3 through 7 are repeated by combining the non-uniform continuous nucleation model, the three-dimensional molten pool dendrite growth model, the solute concentration field distribution model, and the spatial cell trapping model, until both the nucleation cells and liquid phase cells in the simulation area are transformed into solid phase cells. The three-dimensional microstructure morphology of the weld pool is then simulated, as follows: After selecting a local area of ​​the macro weld as the microstructure simulation region, all cells are initialized to the liquid phase state. Based on the non-uniform continuous nucleation model, nucleation positions are randomly determined within the calculation region. Then, based on the Gaussian distribution nucleation density model, it is determined whether the nucleation position cell has nucleated under the current total undercooling. Then, based on the solute flux balance, the velocity magnitude and normal direction of the growth interface in the nucleation cell or interface cell are calculated. Then, the three-dimensional dendrite growth orientation is calculated according to Euler angles. Then, the length of the cell covering along the interface normal direction and the solid fraction increment are calculated. Then, based on solute conservation and Fick's law, the solute distribution and diffusion during cell growth are calculated. Then, based on the spatial cell capture model of the eccentric octahedral algorithm, the growth of three-dimensional dendrites is calculated. The above process is repeated until the growth state cells and liquid phase cells in the simulation region are all transformed into solid phase cells, thus obtaining the three-dimensional microstructure morphology of the weld pool.

Images