Method for constructing finite element mesh model of material microstructure and performance simulation method thereof

Abstract

The application relates to the field of image modeling and finite element simulation analysis, and aims to solve the technical problem of insufficient grid division precision, and provides a material microstructure finite element grid model construction method and a performance simulation method thereof. A binary image is obtained by carrying out binaryzation processing on an original microstructure image; pixel grouping is carried out on the binary image according to a gray value; a dispersed phase material attribute and a matrix phase material attribute are respectively given to a first pixel group and a second pixel group; a grid is divided according to a pixel boundary of the first pixel group and the second pixel group; whenever a grid line intersects with the pixel boundary, the grid near the pixel boundary is refined, and the grid far away from the pixel boundary is not refined until the grids near all the pixel boundaries are refined, an interface grid surrounding a dispersed phase material geometric contour is formed, and a material microstructure finite element grid model is obtained. The application can consider both simulation precision and simulation efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of image modeling and finite element simulation analysis, in particular to a material microstructure finite element grid model construction method and its performance simulation method. BACKGROUND

[0002] The microstructure of a material affects its physical and chemical properties, and studying its microstructure is an important way to regulate and design materials. Currently, methods commonly used to study material properties include experiments and simulation, experiments can obtain intuitive data and results, but due to the influence of experimental environment, time and material cost, more and more simulation is used for material performance research, and finite element analysis is one of the methods.

[0003] Finite element analysis is a numerical simulation method for solving partial differential equation initial boundary value problems. The core of the method is the discretization of the structure. By discretizing the structure of the research object, the grid is divided into elements for analysis, each grid is called an "element", and the point where the elements are connected is called a "node"; then the target variable of the model structure is converted into the variable of each node, according to the provided balance equation, the corresponding algebraic equation is solved by numerical analysis method to obtain the variable result of each node, that is, the final numerical result. Finite element analysis consists of three steps: preprocessing, calculation and post-processing. Preprocessing includes establishing a geometric model, adding material properties, dividing the grid for the structure and setting the boundary conditions, the calculation process is handed over to the solver, and the post-processing mainly outputs the results of the calculation, such as force-displacement curve, stress-strain curve, etc. Establishing an accurate finite element model is a necessary condition for obtaining high-accuracy simulation results.

[0004] With the development of finite element simulation analysis technology, more and more material scholars use finite element simulation method to study the mechanical, thermal, electrical and other properties of materials. However, there are many difficulties in constructing an accurate finite element model of the material microstructure, especially the insufficient grid division accuracy, which often cannot truly reproduce the microstructure inside the material in the simulation experiment, and most of the finite element model is based on idealized assumptions, which often cannot predict the real performance of the material. SUMMARY

[0005] The purpose of the present application is to solve the problems existing in the prior art, and to provide a material microstructure finite element grid model construction method, which solves the technical problem of insufficient grid division accuracy.

[0006] The present application is realized by the following technical scheme: a material microstructure finite element grid model construction method, comprising the following steps:

[0007] Obtain the original microstructure image of the material to be tested;

[0008] The original microstructure image is binarized to obtain a binarized image, and the geometric shape of the dispersed phase material is highlighted in the matrix phase material in the binarized image;

[0009] The binarized image is pixel-grouped according to the gray value, thereby obtaining two pixel groups, a first pixel group composed of pixels corresponding to the dispersed phase material and a second pixel group composed of pixels corresponding to the matrix phase material;

[0010] The dispersed phase material attribute and the matrix phase material attribute are respectively assigned to the first pixel group and the second pixel group, thereby obtaining a material microstructure finite element geometric model;

[0011] The material microstructure finite element geometric model is divided into grids according to the pixel boundaries of the first pixel group and the second pixel group: whenever a grid line intersects with a pixel boundary, the grid near the pixel boundary is refined, and the grid far from the pixel boundary is not refined until all the grids near the pixel boundaries are refined, forming an interface grid around the geometric contour of the dispersed phase material, thereby obtaining a material microstructure finite element grid model.

[0012] Further, the original microstructure image is binarized by using a threshold comparison method: if the gray value of a pixel point is greater than or equal to a threshold value, the gray value of the pixel point is re-assigned to 255, and if the gray value of a pixel point is less than the threshold value, the gray value of the pixel point is re-assigned to 0.

[0013] Further, the threshold value for image binarization is determined as follows:

[0014] S101: Threshold initialization: the average gray value of the gray image F(x, y) of the original microstructure image is assigned to the threshold value Zt;

[0015] S102: Traverse the gray image F(x, y), and compare the gray value of each pixel point with the threshold value Zt:

[0016] If the gray value of a pixel point is greater than or equal to the threshold value Zt, the pixel point is assigned to the gray group S1, and the number N1 of pixel points with a gray value greater than or equal to the initial threshold value Zt is counted;

[0017] If the gray value of a pixel point is less than the threshold value Zt, the pixel point is assigned to the gray group S0, and the number N0 of pixel points with a gray value less than the threshold value Zt is counted;

[0018] S103: Calculate the average value Zj of the gray value of the gray group S1, and calculate the average value Zi of the gray value of the gray group S0, and calculate the average value Zt1 of Zj and Zi;

[0019] S104: calculate the absolute value of the difference between the threshold value Zt and the average value Zt1, i.e., |Zt1-Zt|;

[0020] If |Zt1-Zt|>0.1, then Zt1 is assigned to the threshold value Zt, and returns to step S102;

[0021] If |Zt1-Zt|≤0.1, then the threshold value Zt is output as the threshold value for image binarization.

[0022] Further, the original microstructure image is any one of the following images: a scanning electron microscope image, a transmission electron microscope image, an atomic force microscope image, a scanning tunneling microscope image, and an optical microscope image.

[0023] Further, the image is normalized and contrast-adjusted before pixel grouping of the binarized image.

[0024] Further, the material properties include mechanical properties, thermal properties, and electrical properties.

[0025] Further, the material properties are assigned to the divided grid in the OOF2 software.

[0026] The application also provides a material performance simulation method, which imports the material microstructure finite element grid model obtained by the material microstructure finite element grid model construction method of the application into a finite element analysis software for simulation testing.

[0027] Further, ABAQUS software is used for simulation.

[0028] Compared with the prior art, the application has the following beneficial effects:

[0029] 1. The microstructure of heterogeneous materials has different geometric shapes, and the prior art simply uniformly divides the grid without considering the interface between the heterogeneous materials. However, the grid accuracy of the interface region is crucial for the accuracy of simulation testing. The pixel boundary obtained by pixel grouping in the application reflects the interface between the heterogeneous materials (dispersed phase and matrix phase), and the interface grid obtained by grid division based on the pixel boundary greatly refines the grid in the interface region and greatly increases the grid nodes, which not only improves the accuracy of the grid model, but also is an important guarantee for the accuracy of finite element simulation.

[0030] 2. The method of the application keeps the grid in the non-interface region sparse, which is beneficial to reducing the calculation amount in the finite element simulation process and improving the simulation efficiency.

[0031] 3. The application uses a traversal loop method to optimize the threshold value for image binarization, which can more accurately distinguish the dispersed phase and the matrix phase and improve the accuracy of the pixel boundary description.

[0032] 4. The processing flow of the present application is simple, and the material microstructure finite element grid model can be obtained after the grid division, without post-processing operations such as pixel merging and image cutting.

[0033] 5. The present application has no limitation on the type of original microstructure image, and has good compatibility. BRIEF DESCRIPTION OF DRAWINGS

[0034] Figure 1 for the flowchart of the material performance simulation method in the specific embodiment;

[0035] Figure 2 for the original material microstructure image in the embodiment;

[0036] Figure 3 for the image after binarization processing in the embodiment;

[0037] Figure 4 for the finite element geometric model established based on the material microstructure image in the embodiment;

[0038] Figure 5 for the finite element grid model established based on the material microstructure image in the embodiment;

[0039] Figure 6 for Figure 5 a local enlarged view of

[0040] Figure 7 for the simulation effect diagram of stress distribution under the 100% uniaxial stretching working condition in the embodiment;

[0041] Figure 8 for the simulation effect diagram of strain distribution under the 100% uniaxial stretching working condition in the embodiment. DETAILED DESCRIPTION

[0042] Since the material performance simulation method of the present application includes the construction method of the material microstructure finite element grid model, the present specific embodiment is described by taking the material performance simulation method as an example.

[0043] The present application will be further described in detail below with reference to the accompanying drawings:

[0044] As shown in Figure 1 , a material performance simulation method comprises the following steps:

[0045] S1: obtaining an original microstructure image of a material to be tested; the original microstructure image is any one of the following images: a scanning electron microscope image, a transmission electron microscope image, an atomic force microscope image, a scanning tunneling microscope image and an optical microscope image.

[0046] S2: binarizing the original microstructure image to obtain a binarized image, in which the geometric shape of the dispersed phase material is highlighted in the matrix phase material.

[0047] The original microstructure image is binarized by threshold comparison method: if the gray value of a pixel is greater than or equal to the threshold value, the gray value of the pixel is re-assigned as 255; if the gray value of a pixel is less than the threshold value, the gray value of the pixel is re-assigned as 0.

[0048] Preferably, the threshold value for image binarization is determined as follows:

[0049] S101: threshold initialization: the average gray value of the gray image F(x, y) of the original microstructure image is assigned to the threshold value Zt;

[0050] S102: traverse the gray image F(x, y), and compare the gray value of each pixel with the threshold value Zt:

[0051] If the gray value of a pixel is greater than or equal to the threshold value Zt, the pixel is assigned to the gray group S1, and the number of pixels with gray value greater than or equal to the initial threshold value Zt is counted as N1;

[0052] If the gray value of a pixel is less than the threshold value Zt, the pixel is assigned to the gray group S0, and the number of pixels with gray value less than the threshold value Zt is counted as N0;

[0053] S103: calculate the average value Zj of the gray value of the gray group S1, and calculate the average value Zi of the gray value of the gray group S0, and calculate the average value Zt1 of Zj and Zi;

[0054] S104: calculate the absolute value of the difference between the threshold value Zt and the average value Zt1, i.e. |Zt1-Zt|;

[0055] If |Zt1-Zt|>0.1, then Zt1 is assigned to the threshold value Zt, and step S102 is returned;

[0056] If |Zt1-Zt|≤0.1, then the threshold value Zt is output as the threshold value for image binarization.

[0057] S3: import the binarized image into OOF2 software, group the pixels of the binarized image according to the gray value, thereby obtaining two pixel groups, the first pixel group is composed of pixels corresponding to the dispersed phase material, and the second pixel group is composed of pixels corresponding to the matrix phase material.

[0058] Preferably, before grouping the pixels of the binarized image, the image is standardized and the contrast is adjusted, so that the different material parts of the image show clearer contrast effect.

[0059] S4: attribute the dispersed phase material properties and the matrix phase material properties to the first pixel group and the second pixel group respectively, the material properties including mechanical properties, thermal properties and electrical properties, so as to obtain a material microstructure finite element geometric model.

[0060] S5: divide the grid of the material microstructure finite element geometric model according to the pixel boundary of the first pixel group and the second pixel group: whenever the grid line intersects with the pixel boundary, the grid near the pixel boundary is refined, and the grid far from the pixel boundary is not refined until the grid near all the pixel boundaries is refined, forming an interface grid around the geometric contour of the dispersed phase material, so as to obtain a material microstructure finite element grid model. The OOF2 software has a grid refinement function, and the grid division is completed through human-computer interaction operation until a satisfactory result is achieved, and the contour of the interface grid is basically consistent with the contour of the two-phase interface on the original microstructure image.

[0061] S6: export the material microstructure finite element grid model: change the text file exported by the OOF2 software to an inp file.

[0062] S7: import the material microstructure finite element grid model into the finite element analysis software for simulation test: open the finite element analysis software to import the inp file, so as to obtain the finite element model with divided grid, and then set the boundary conditions, load and the like in the ABAQUS software, so as to realize the simulation analysis of the required performance.

[0063] Embodiment

[0064] Taking a microstructure image of a certain rubber composite material, a scanning electron microscope image, as an example, the original microstructure image is as shown in Figure 2

[0065] First, the original microstructure image is binarized to obtain Figure 3 , and Figure 3 and Figure 2 , it can be seen that the binarized image obtained after processing has almost no structural difference in the geometric shape of the dispersed phase from the original microstructure image.

[0066] Then, the binarized image is imported into the OOF2 software, the image contrast is enhanced, the black and white pixel points of the image are selected by setting different gray value ranges, and the first pixel group BS10 and the second pixel group BS2 are established respectively. Then, the material and properties are established according to the material properties of BS10 and BS2, the properties are attributed to the material, the material is attributed to the pixel group, finally, the grid is divided, and the model is saved.

[0067] ​Finally, the text file exported by OOF2 is changed into an inp file, and the inp file is opened by using finite element analysis software, and the finite element geometric model of material microstructure and the finite element grid model of material microstructure are respectively as shown in Figure 4 and Figure 5 The local enlarged view of the finite element grid model is as shown in Figure 6 It can be seen from Figure 6 that the interface grid around the geometric contour of the dispersed phase material is very fine, and the non-interface grid is relatively sparse.

[0068] The interface grid is fine, which provides more nodes for numerical simulation of finite element analysis, and can effectively improve the simulation accuracy. The grid in the non-interface region is relatively sparse, which is beneficial to reduce the calculation amount in the finite element simulation process and improve the simulation efficiency.

[0069] The left side of the finite element grid model of material microstructure is fixed, and the model is subjected to 100% uniaxial stretching in the X-axis direction, and the stress and strain distribution under the condition of 100% uniaxial stretching are obtained by submitting the job, and are respectively as shown in Figure 7 and Figure 8

[0070] The matrix phase has a larger deformation than the dispersed phase, especially in the matrix region between the dispersed phases close to each other. Similarly, the stress value of the matrix phase between the dispersed phases close to each other also reaches the maximum, which is consistent with the result that there is a larger stress and deformation in the matrix phase region between the rubber composite material interfaces. The stress and strain distribution cloud diagram can directly display the performance difference of different phase regions of the composite material, and provide another possibility for material performance analysis in addition to experimental testing.

[0071] In summary, the present application can well realize the accurate establishment of the finite element analysis model based on the material microstructure image.

[0072] The above technical solution is only one embodiment of the present application, and for those skilled in the art, on the basis of the principle disclosed in the present application, various types of improvements or modifications can be easily made, and the technical solution described in the above embodiment is not limited to the present application, therefore, the above description is only preferred, and does not have the meaning of limitation.​

Claims

1. A method of constructing a microstructure finite element mesh model of a material, characterized by, The method comprises the following steps: obtaining an original microstructure image of a material to be tested; performing binaryzation on the original microstructure image to obtain a binaryzation image, in which the geometric shape of the dispersed phase material is highlighted in the matrix phase material; grouping the pixels of the binaryzation image according to the gray values, thereby obtaining two pixel groups, a first pixel group composed of pixels corresponding to the dispersed phase material and a second pixel group composed of pixels corresponding to the matrix phase material; assigning the properties of the dispersed phase material and the properties of the matrix phase material to the first pixel group and the second pixel group, respectively, thereby obtaining a microstructure finite element geometric model of the material; dividing the microstructure finite element geometric model of the material into grids according to the pixel boundaries of the first pixel group and the second pixel group: whenever a grid line intersects with a pixel boundary, the grids near the pixel boundary are refined, while the grids far from the pixel boundary are not refined until all the grids near the pixel boundaries are refined, forming interface grids around the geometric contour of the dispersed phase material, thereby obtaining a microstructure finite element grid model of the material; performing binaryzation on the original microstructure image by using a threshold comparison method: if the gray value of a pixel is greater than or equal to a threshold value, the gray value of the pixel is re-assigned as 255, and if the gray value of a pixel is less than the threshold value, the gray value of the pixel is re-assigned as 0; the threshold value for image binaryzation is determined in the following manner: S101: threshold initialization: the average gray value of the gray image F(x, y) of the original microstructure image is assigned to the threshold value Zt; S102: traverse the gray image F(x, y) and compare the gray value of each pixel with the threshold value Zt: if the gray value of a pixel is greater than or equal to the threshold value Zt, the pixel is assigned to the gray group S1, and the number of pixels with a gray value greater than or equal to the initial threshold value Zt is counted as N1; if the gray value of a pixel is less than the threshold value Zt, the pixel is assigned to the gray group S0, and the number of pixels with a gray value less than the threshold value Zt is counted as N0; S103: calculate the average value Zj of the gray values of the gray group S1, and calculate the average value Zi of the gray values of the gray group S0, and calculate the average value Zt1 of Zj and Zi; S104: Calculate the absolute value of the difference between the threshold value Zt and the average value Zt1, i.e. ; If Zt1 is assigned to the threshold value Zt, and the process returns to step S102. If then output the threshold Zt as the threshold for image binarization; assign material properties and divide grids in the OOF2 software.

2. The method of constructing a material microstructure finite element mesh model according to claim 1, wherein, The original microstructure image is any one of the following images: a scanning electron microscope image, a transmission electron microscope image, an atomic force microscope image, a scanning tunneling microscope image, and an optical microscope image.

3. The method of claim 1, wherein Before grouping the pixels of the binaryzation image, perform standardization processing and contrast adjustment on the image.

4. The method of claim 1, wherein The material properties include mechanical properties, thermal properties, and electrical properties.

5. A material property simulation method characterized by, The microstructure finite element grid model of the material obtained by the method for constructing a microstructure finite element grid model of a material according to any one of claims 1-4 is imported into a finite element analysis software for simulation testing.

6. The material performance simulation method of claim 5, wherein, ABAQUS software is used for simulation.

Images