Construction method of gradient nanocrystalline metal atom model
By constructing a gradient nanocrystal atomic model of metal materials, the problem that grain sizes in the prior art are difficult to follow the gradient arrangement of arbitrary functions is solved, and the matching of material strength and plasticity is achieved, and technical support for molecular dynamics simulation is provided.
Patent Information
- Application Number
- CN202411634560.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-15
- Publication Date
- 2025-03-07
AI Technical Summary
The prior art is difficult to achieve the arrangement of the grain size of metal materials spatially following any functional gradients, making it difficult to take into account both the strength and plasticity of the material.
By defining the initial conditions and functional relationships, the grain size gradient function is discrete by using recursive formulas, the model space is constructed and discretized, node coordinates are randomly generated, and atomic fill is used to generate gradient nanocrystalline metal atomic model.
The construction of two-dimensional and three-dimensional gradient nanocrystal atomic models of any metal material is realized, which can accurately control the spatial distribution of grain size, break through the bottleneck that is difficult to achieve both strength and plasticity, and provides technical support for molecular dynamics simulation.
Smart Images
Figure CN119580893A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of material computational simulation, and in particular relates to a method for constructing a gradient nanometal model. Background Art
[0002] Metal materials have good mechanical properties, durability, machinability, and electrical and thermal conductivity, and are therefore widely used in engineering machinery, coal mining machinery, aerospace, automobiles, electronic products and other fields. However, with the development of society, higher requirements are placed on the performance of materials. Against the background of increasing energy shortages, reducing the use of materials themselves and improving the performance of materials, especially mechanical properties, has become the focus of researchers. Studies have found that the strength and hardness of metal materials can be significantly improved by reducing the grain size of metal materials to the nanoscale. However, as the grain size decreases, the plasticity of the material decreases rapidly, and it is difficult to achieve both strength and plasticity. Therefore, how to break the contradiction between the strength and plasticity of metal materials has always been a basic scientific issue and hot topic in this field.
[0003] Gradient nanocrystalline metal refers to a metal material whose grain size presents a gradient change in space. In most cases, the grain size increases continuously from the nanoscale on the surface to the core. Studies have shown that this unique gradient structure breaks the bottleneck of strength and plasticity, achieves the matching of strength and plasticity, and significantly enhances the surface hardness, wear resistance, and fatigue resistance of metal materials. Therefore, precise regulation of gradient nanocrystalline structure and establishment of a precise relationship between structure and performance to maximize the comprehensive performance of such materials have become hot spots and difficulties in the current technology and science fields.
[0004] Molecular dynamics simulation can simulate the physical process of complex phenomena at the atomic scale and predict the performance of materials. In the field of metal materials, molecular dynamics simulation has been successfully applied to the simulation of stretching, wear and other processes of advanced metal materials such as nanocrystals, nanowires, and nanotwins, revealing the atomic mechanism of complex processes and discovering a series of unprecedented phenomena and mechanisms. Applying molecular dynamics simulation to gradient nanocrystalline metal simulation is expected to help scholars understand the role of gradient-changing grains in the deformation process, and then optimize the spatial distribution of grain size to achieve performance improvement. At present, in molecular dynamics simulation, the Voronoi polygon method is used to construct the atomic model of nanocrystals, but it is difficult to achieve the gradient distribution of grain size, especially to achieve the gradient arrangement of grain size following arbitrary functions. Therefore, it is urgent to develop a construction technology for gradient nanocrystal atomic models to achieve precise control of grain size and provide technical support for subsequent molecular dynamics simulation of complex processes. Summary of the invention
[0005] In view of the problem that the existing gradient nanocrystalline metal atomic model cannot realize the grain size following the gradient arrangement of arbitrary functions, the present invention proposes a method for constructing a gradient nanocrystalline metal atomic model.
[0006] To achieve the above object, the present invention provides the following solutions:
[0007] One of the solutions of the present invention provides a method for constructing a gradient nanocrystalline metal atom model, comprising the following steps:
[0008] S1. Define initial conditions and functional relationships
[0009] Furthermore, the initial conditions include: initial grain size d0, final grain size d f , grain size d, gradient change direction x;
[0010] Furthermore, the grain size gradient changes satisfy the function:
[0011] d=f(x)[1] The functional relationship here can be any function, including but not limited to linear function, quadratic function, and exponential function.
[0012] S2. Discretization of gradient function
[0013] Furthermore, the recursive formula [2] is used to discretize the grain size gradient function [1]:
[0014] d i =f(x i )
[0015] x i+1 =x i +d i ,(i=1,2,...,n)[2]
[0016] The initial conditions are:
[0017] x1=0
[0018] d1=d0[3]
[0019] The termination conditions are:
[0020] |d n -d f |<|d n+1 -d f |[4]
[0021] That is, the current function value and d f The difference between the next function value and d f The difference.
[0022] S3. Define the model space size
[0023] Furthermore, according to formula [5], the length of the model space is defined as L, which is 2L for the symmetric gradient model. The length direction is the direction of grain gradient change, and the width and height are specified as needed. The model can be two-dimensional or three-dimensional.
[0024]
[0025] S4. Model space discretization
[0026] Furthermore, the model space is discretized into cubes with varying sizes, and for two-dimensional models, it is discretized into squares with varying sizes;
[0027] Furthermore, the left edge of the cube / square in the gradient direction is defined as x given by equation [2] i , the side length of the cube / square is exactly the grain size d at that location i ;
[0028] Furthermore, the center coordinate point of the cube / square is calculated.
[0029] S5. Define the discreteness of the node
[0030] Furthermore, the discreteness S of the nodes is defined. When S is 0, the nodes are located at the center of the cube / square; when S is 1, the nodes are randomly arranged in the cube / square; when S is between 0 and 1, the nodes are randomly arranged in the cube / square reduced by the ratio of S.
[0031] S6. Set node coordinates
[0032] Furthermore, according to the above discretized space model and the discreteness of the nodes, a node coordinate is randomly generated within the space range through a random number function;
[0033] Furthermore, the above process can be implemented by mathematical software such as Octave, Matlab, and programming languages such as Python and C, and what is finally obtained is a txt file with node coordinate data.
[0034] S7, fill atoms
[0035] Further, the metal crystal structure unit required in the above gradient nanostructured metal is modeled, and the crystal structure can be face-centered cubic (fcc), body-centered cubic (bcc), hexagonal (hcp) or any other crystal structure;
[0036] Further, the above crystal structure unit and the node coordinates obtained in step S6 are used to perform atomic filling using the Voronoi method. The direction of each grain can be specified as needed or randomly assigned;
[0037] Further, a Lammps recognizable file is generated;
[0038] Furthermore, the above program is implemented through the open source software Atomsk to obtain the cfg file of the gradient nanocrystalline metal atomic model.
[0039] S8. Visualization of gradient nanocrystal structure and grain boundary distribution characteristics
[0040] Furthermore, the coordination number, central symmetry parameter, common domain analysis and other methods are used to identify defective atoms, and visualization software such as Ovito and Atomeye are used to display the gradient nanocrystal structure and grain boundary distribution characteristics.
[0041] The present invention discloses the following technical effects:
[0042] The present invention realizes the construction of two-dimensional and three-dimensional gradient nanocrystalline atomic models of any metal material. In particular, it realizes the gradient arrangement of grain size following any function in the spatial direction, where the functional relationship includes but is not limited to linear function, quadratic function, and exponential function. The present invention provides technical support for molecular dynamics simulation of complex processes of gradient nanocrystalline metals. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0044] Figure 1 Atomic model of two-dimensional symmetric gradient nanocrystalline pure iron (S = 0, d = 0.5x + 30)
[0045] Figure 2 Atomic model of two-dimensional symmetric gradient nanocrystalline pure iron (S = 0.6, d = 0.5x + 30)
[0046] Figure 3 Atomic model of two-dimensional asymmetric gradient nanocrystalline pure iron (S = 0.6, d = 0.5x + 30)
[0047] Figure 4 Atomic model of three-dimensional symmetric gradient nanocrystalline pure iron (S = 0.6, d = 0.5x + 30)
[0048] Figure 5 The atomic model of two-dimensional symmetric gradient nanocrystalline pure aluminum (S = 0.6, d = 0.001x 2 +30) DETAILED DESCRIPTION
[0049] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments, but the protection scope of the present invention is not limited thereto.
[0050] Example 1: Construction of a two-dimensional symmetric gradient nanocrystalline pure iron model
[0051] (1) Define initial conditions and functional relationships
[0052] The initial grain size d0 is defined as Final grain size d f for d=0.5x+30.
[0053] (2) Discretization of gradient function
[0054] 1) The gradient function d = 0.5x + 30 is discretized by the recursive formula [2], and the initial conditions x1 = 0, d1 = 30 are introduced. Then, a series of d is obtained by combining d = 0.5x + 30. i (i=1,…,n).
[0055] 2) Determine d by formula [4] n The value of .
[0056] (3) Define the model space dimensions.
[0057] The length of the model space is calculated by formula [5]: And specify the width as
[0058] (4) Model space discretization
[0059] 1) Discretize the above model space into squares with varying gradient sizes.
[0060] 2) The left edge of the square in the gradient direction is defined as x obtained by equation [2] i , the side length of the square is exactly the grain size d at that position i
[0061] 3) Calculate the center coordinates of the square.
[0062] (5) Define the discreteness S of the node.
[0063] When S is 0, the seed only appears in the center of the grid. When S is 1, the seed can appear anywhere in the grid. When S is between 0 and 1, the nodes are randomly arranged in a cube / square that is reduced in proportion to S. In this example, S is 0.
[0064] (6) Set node coordinates
[0065] Furthermore, according to the above discretized space model and the discreteness of the nodes, a node is randomly generated within the space range through a random number function, and its coordinates are recorded.
[0066] The above program is implemented through the open source software Octave, and finally the node coordinates are obtained.
[0067] (7) Filling Atoms
[0068] 1) First, construct the crystal structure model of bccα-Fe;
[0069] 2) Based on the crystal structure model obtained in step 1), the node coordinates obtained in step (6) are used to construct a gradient nanometal model using the Voronoi method, with the orientation of the grains being random.
[0070] The above program is implemented through the open source software Atomsk.
[0071] (8) Visualization of gradient nanocrystalline structure and grain boundary distribution characteristics.
[0072] The open source software Ovito was used to visualize the gradient nanostructured metal and observe its grain size distribution.
[0073] Attached Figure 1 This is a common neighbor analysis diagram of the two-dimensional symmetrical gradient nanocrystalline bcc structure α-Fe in Example 1 when the discreteness is 0. The dark blue part is the bcc structure atom, and the white part is the grain boundary. It can be seen that the grain size is distributed in a symmetrical gradient.
[0074] Comparative Example 1: The difference between this comparative example and Example 1 is that the discreteness is increased, and the discreteness S is 0.6.
[0075] Attached Figure 2 This is a common domain analysis diagram of the two-dimensional symmetric gradient nanocrystalline bcc structure α-Fe when the discreteness is 0.6 in Example 1. From the figure, it can be seen that the grain boundaries are obviously irregular compared to when d=0, and the introduction of discreteness makes the distribution of grain boundaries more consistent with the actual distribution.
[0076] Comparative Example 2: Construction of a two-dimensional asymmetric gradient nanocrystalline pure iron model
[0077] The only difference between this comparative example and comparative example 1 is that the atoms in the symmetric part are deleted for the construction of the asymmetric gradient nanocrystal model.
[0078] Attached Figure 3 This is a common domain analysis diagram of asymmetric gradient nanocrystalline bcc structure α-Fe in comparative example 2 when the discreteness S is 0.6. It can be seen that asymmetric gradient nanocrystalline pure iron is successfully established by this method.
[0079] Comparative Example 3: Construction of a three-dimensional symmetrical gradient nanocrystalline pure iron model.
[0080] The difference between this embodiment and comparative example 1 is that the gradient nodes are distributed in three-dimensional space, and the model space size is The remaining steps are the same as those in Comparative Example 1.
[0081] Attached Figure 4 This is the grain segmentation diagram of the three-dimensional symmetric gradient nanocrystalline pure iron when S=0.6. Different from the previous example, the common domain analysis method is not used here. Instead, different grains are colored through polyhedral template matching and grain segmentation modifiers. Through this method, we can see that the grain size of the three-dimensional model presents a symmetrical gradient distribution in the length direction.
[0082] Example 2: Construction of a two-dimensional symmetrical gradient nanocrystalline pure aluminum model.
[0083] The difference between this embodiment and embodiment 1 is that the function expression is a quadratic function d=0.001x 2 +30, the discreteness is 0.6, and the model space size is The filling atoms are fcc structure pure aluminum.
[0084] Attached Figure 5 This is the common domain analysis diagram of pure aluminum with two-dimensional symmetrical gradient nanocrystalline fcc structure in Example 2 when the discreteness is 0.6. The green part is the fcc structure atom, and the white part is the grain boundary. Compared with the comparative example 1 with the same discreteness of 0.6, the grain size changes according to a linear function, the grain size of the fine grain part of Example 2 changes slowly at the beginning, and then increases rapidly.
Claims
1. A method for constructing a gradient nanocrystalline metal atomic model, characterized in that: The following steps are involved: (1) Define initial conditions and functional relationships; (2) Discreteness of gradient function; (3) Define the model space size; (4) Model space discretization; (5) Define the discreteness of the node; (6) Set node coordinates; (7) Filling atoms; (8) Visualization of gradient nanocrystalline structure and grain boundary distribution characteristics.
2. The method for constructing a gradient nanocrystalline metal atomic model according to claim 1, wherein the initial conditions include an initial grain size d0, a final grain size d f , the direction x of the grain size gradient change and the functional relationship of the grain size gradient change [1], where the functional relationship can be any function, including but not limited to linear function, quadratic function, exponential function. d=f(x)[1].
3. According to the method for constructing a gradient nanocrystalline metal atomic model according to claim 1, the discretization of the gradient function is achieved by the following steps: 1) The recursive formula [2] is used to discretize the grain size gradient function; d i =f(x i ) x i+1 =x i +d i ,(i=1,2,...,n)[2] 2)d n is determined by the termination condition described in formula [3], that is, the current function value and d f The difference between the next function value and d f The difference. d n -d f |<|d n+1 -d f |[3]。 4. The method for constructing a gradient nanocrystalline metal atomic model according to claim 1, characterized in that The definition of the model space size is specifically defined by formula [4], where the length of the model space is defined as L, which is 2L for a symmetric gradient model, and the length direction is the direction of grain gradient change. The width and height are specified as required. The model can be two-dimensional or three-dimensional.
5. The method for constructing a gradient nanocrystalline metal atomic model according to claim 1, characterized in that The model space is discretized, specifically comprising the following steps: 1) Discretize the 3D model space into cubes with varying sizes, and discretize the 2D model space into squares with varying sizes; 2) The left edge of the cube / square in the gradient direction is defined as x obtained by equation [2] i , the side length of the cube / square is exactly the grain size d at that location i ; 3) Calculate the center coordinates of the cube / square.
6. The method for constructing a gradient nanocrystalline metal atomic model according to claim 1, characterized in that The discreteness of the nodes is defined, and the discreteness S takes values between 0 and 1: when S is 0, the nodes are located at the center of the cube / square; when S is 1, the nodes are randomly arranged in the cube / square; when S takes values between 0 and 1, the nodes are randomly arranged in the cube / square reduced by the proportion of S.
7. The method for constructing a gradient nanocrystalline metal atomic model according to claim 1, characterized in that The node coordinates are set by using a random function to randomly generate the node coordinates in the cube / square space in step (4) and step (5).
8. The method for constructing a gradient nanocrystalline metal atomic model according to claim 1, wherein the characteristic step ( 1)(2)(3)(4)(5)(6) are implemented using mathematical software including but not limited to Octave, Matlab, and programming languages such as Python and C.
9. The method for constructing a gradient nanocrystalline metal atomic model according to claim 1, characterized in that The filling of atoms specifically comprises the following sub-steps: 1) Establish the crystal structure unit of metal, including but not limited to face-centered cubic (fcc), body-centered cubic (bcc), hexagonal (hcp) and other structures; 2) Using the above crystal structure unit and the node coordinates obtained in step (6), the Voronoi method is used to perform atomic filling. The direction of each grain can be specified as needed or randomly assigned.