Molecular dynamics simulation method and system for bearing steel material microscopic crack initiation in-situ tensile test
By integrating the collaborative workflow of multiple modeling software, the problem of insufficient accuracy in molecular dynamics simulation of bearing steel materials was solved, efficient and accurate material property prediction was achieved, and an economical alternative was provided, shortening the experimental cycle and reducing costs.
Patent Information
- Application Number
- CN202510653619.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-21
- Publication Date
- 2025-09-19
AI Technical Summary
When dealing with the mechanical behavior of bearing steel materials, existing molecular dynamics simulation technology has insufficient simulation accuracy and poor matching with experimental data, making it difficult to achieve efficient material performance prediction.
By integrating the functions of multiple modeling software to form a collaborative workflow, including data acquisition, processing and simulation, an accurate molecular dynamics simulation method for in-situ tensile testing of microcrack initiation in bearing steel materials is established. Combined with electron backscatter diffraction, electron probe microanalysis and molecular dynamics software, high-precision material simulation is achieved.
It achieves efficient and accurate prediction of the microstructural evolution of bearing steel materials during the tensile process, provides an economical alternative to traditional in-situ tensile tests, significantly shortens the experimental cycle and reduces costs, and provides a reliable reference for material performance prediction and design.
Smart Images
Figure CN120673865A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of bearing steel material modeling, and in particular to a molecular dynamics simulation method and system for an in-situ tensile test of microcrack initiation in a bearing steel material. Background Art
[0002] With the development of modern materials science, a deep understanding of the mechanical behavior of materials under varying conditions is becoming increasingly important. In situ tensile testing is a commonly used method for studying the micromechanical properties of metals. This test directly applies tensile stress to the metal and combines it with microscopic observation to reveal the evolution of the microstructure. However, in situ tensile testing faces multiple challenges, including expensive experimental equipment, long test cycles, and difficult sample preparation. These challenges severely impact the efficiency and cost-effectiveness of both scientific research and industrial applications.
[0003] In this context, molecular dynamics (MD) simulation has gradually become one of the effective tools for studying metal materials. MD simulation can reproduce the atomic and molecular behavior of materials under stress through computer modeling, thereby predicting the mechanical properties of materials. Compared with traditional physical experiments, MD simulation has higher operability and flexibility, can accurately control simulation conditions, and perform repeated tests at a lower cost. However, when dealing with the mechanical behavior of bearing steel materials, the existing MD simulation technology often has shortcomings such as limited simulation accuracy and low matching with experimental data. This is mainly because the microstructure of bearing steel materials is relatively complex, and it is difficult to complete the establishment of an accurate model, and effective coupling is required on large scales and multiple time scales. Summary of the Invention
[0004] To address the challenges of the existing technology, this paper provides a molecular dynamics simulation method and system for in-situ tensile testing of microcrack initiation in bearing steel. This method integrates the functionality of multiple modeling and simulation software to form an efficient collaborative workflow, enabling higher-precision mechanical simulation of bearing steel.
[0005] In order to achieve the above object, the technical solution adopted by the present invention is: a molecular dynamics simulation method for in-situ tensile test of micro crack initiation of bearing steel material, comprising the following steps:
[0006] Obtaining the matrix crystal structure, matrix crystallographic information, and crystal information containing the second phase in the area requiring in-situ tensile testing;
[0007] Obtain qualitative and quantitative analysis results of elements in the test area;
[0008] Post-processing of EBSD data is performed based on the crystal structure of the area to be tested for in-situ tensile testing, the matrix crystallographic information, and the crystal information containing the second phase to obtain the matrix grain boundary map, matrix single grain map, and second phase map of the test area;
[0009] Convert matrix grain boundary map, matrix single grain map, and second phase map into mask files;
[0010] Constructing a grain boundary model, a single grain model, and a second phase model based on the mask file and in combination with the mask;
[0011] The grain boundary model, single grain model and second phase model are combined to obtain the matrix atom model doped with the second phase;
[0012] The atomic percentage of each element obtained by EPMA was used to remove the atoms contained in the second phase and calculate the number of solid solution doping atoms of each element missing in the matrix full-atom model;
[0013] Use Lammps software to achieve solid solution doping of each element's atoms and complete in-situ tensile molecular dynamics modeling;
[0014] Stretched molecular dynamics simulations were performed using Lammps software.
[0015] Furthermore, the crystal structure of the test area in step one, the matrix crystallographic information including the crystal orientation and rotation direction, the crystal information of the second phase including the crystal type, crystal orientation, rotation direction, position, and size of the second phase, the matrix crystal structure, matrix crystallographic information, and the crystal information of the second phase are obtained using the electron backscatter diffraction method and saved in a ctf format file that can be post-processed.
[0016] Furthermore, the qualitative and quantitative analysis results of the elements in the test area were obtained using an electron probe microanalyzer, and the atomic percentage of each element was output.
[0017] Furthermore, the post-processing of the EBSD data is completed through the Matlab-MTEX module. By reading the ctf format file containing the matrix crystal structure, matrix crystallographic information, and crystal information of the second phase in the area where the in-situ tensile test is required, the grain boundary map of the matrix and the second phase phase map are output, and the image size is equal to the overall size of the test area.
[0018] Furthermore, the Matlab-MTEX module outputs matrix grain maps while preserving the overall size of the test area. Each map contains only one grain, and the number of images is equal to the number of matrix grains.
[0019] Furthermore, in step five, the Python-skimage module is used to convert the matrix grain boundary map, matrix single grain map, and second phase map of the test area into a mask file in txt format, and binarization is performed so that the mask file consists of 0 and 1, and the size is adjusted according to the required model size.
[0020] Furthermore, the Atomsk-select module is used to read the mask file in txt format, delete the atoms in the area masked by 1 character, and establish the matrix grain boundary model, matrix single grain model and second phase model.
[0021] Furthermore, the solid solution doping percentage of each element used in Lammps software was calculated by using the total number of atoms in the model, the atomic percentage of the elements obtained by EPMA, and the number of atoms of each element in the existing model.
[0022] The present invention can also provide a molecular dynamics simulation system for in-situ tensile testing of micro crack initiation in bearing steel materials, comprising a data acquisition module, a call module, and a simulation module;
[0023] The data acquisition module is used to obtain the matrix crystal structure, matrix crystallographic information, crystal information containing the second phase, and qualitative and quantitative analysis results of elements in the test area where in-situ tensile testing is required;
[0024] The calling module is used to call the simulation module to process the data or files acquired by the data acquisition module;
[0025] The simulation modules include Matlab-MTEX module, Python-skimage module, -polycrystal module and -select module of Atomsk software, and Lammps software.
[0026] Furthermore, the Matlab-MTEX module is used for EBSD data post-processing; the Python-skimage module is used to convert image files containing matrix grain boundary maps, matrix single grain maps, and second phase phase maps into mask files in txt format; the -polycrystal module and -select module of the Atomsk software are used to combine mask files to establish grain boundary models, single grain models, and second phase models through masking; and the Lammps software realizes microscopic tensile molecular dynamics simulation.
[0027] Compared with the prior art, the present invention has at least the following beneficial effects: the present invention can efficiently and accurately predict the microstructural evolution of the material during the tensile process. The molecular dynamics simulation method and system for tensile testing of bearing steel materials established according to the present invention provide an efficient and economical alternative to traditional in-situ tensile tests, and ensure experimental accuracy. It can not only help scientific researchers save a lot of time and money, but also open up new possibilities for performance prediction and material design of bearing steel materials. Through this method, not only can the experimental cycle be significantly shortened and the experimental cost be reduced, but also richer micromechanical information can be provided, providing a more reliable reference for material development and optimization. BRIEF DESCRIPTION OF THE DRAWINGS
[0028] Figure 1 The figure is a flow chart of the molecular dynamics simulation method and system for in-situ tensile testing of bearing steel materials according to the present invention. DETAILED DESCRIPTION
[0029] The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention, but should not be construed as limiting the present invention.
[0030] In this embodiment, the microstructure modeling of the M50 alloy is achieved through the following steps:
[0031] S1: Electron backscatter diffraction (EBSD) is used to scan the area planned for in-situ tensile testing to obtain crystallographic information, including the grain structure, orientation, and rotation of the matrix, as well as crystallographic information of secondary phases. The sample undergoes a series of pre-treatments, including grinding, mechanical polishing, and electrolytic polishing, to ensure a mirror-like finish and contamination-free surface. The collected EBSD data is exported as a .ctf file.
[0032] S2: Use an electron probe microanalyzer (EPMA) to perform qualitative and quantitative elemental analysis on the test area. The analysis is performed in a vacuum environment by selecting an appropriate accelerating voltage (10-20 kV) and probe current (50-100 nA). Output is the atomic percentage of each element, and the detailed distribution of each element is recorded. Analysis of silicon, aluminum, and titanium may require an inert gas atmosphere.
[0033] As a further optimization, in order to ensure that the test areas corresponding to the information of S1 and S2 remain consistent, a microhardness tester or a nanoindenter can be used to delineate the test area. A microhardness tester is used to delineate large areas. A suitable load is selected to complete the hardness measurement at the four corners of the expected test area, and the test area is circled by the indentations generated by the measurement. A nanoindenter is used to delineate small areas. Based on the large area division, a finer division is performed, and the test area size can be flexibly selected.
[0034] S3: Import the ctf format file obtained in S1 into the MTEX module of Matlab for EBSD data processing, and perform post-processing of the crystal information through the MTEX module to generate the matrix grain boundary map, single grain map and second phase map. The matrix grain boundary map and the second phase map in the output image adopt the same strategy, and the matrix grain boundary and the interior of the second phase grain are colored black, and the rest of the area is colored white. The image file is exported in jpg or png format for subsequent image processing. For the matrix single grain map, each grain is colored black in turn. It is necessary to keep the image size consistent with the overall size of the test area. If necessary, the matrix grain boundary can be colored white when output to achieve this purpose. When outputting the matrix single grain image file, the matrix crystal phase is screened and numbered in sequence according to the matrix grains, and the crystal direction and rotation direction of each matrix grain are output at the same time. The crystal direction is expressed in the form of crystal structure plus crystal band axis (such as Bcc
[001] or ), with the rotation direction represented by angles in the xyz directions (e.g., 38, 53, 137). This is saved as a txt file for identification in subsequent steps. Taking matrix grain 1 as an example, the output image is named 1.png, and the 1.txt file is also output (the data in the file is Bcc
[001] , 38, 53, 137). The second phase diagram follows the same output process as the matrix single grain image, allowing for the subsequent establishment of a molecular dynamics model for each grain.
[0035] S4: Import the image generated in S3 into the Python-skimage module. The same strategy is used for identifying the matrix grain boundary image and the individual grain image: assign the character 0 to black areas and the character 1 to white areas. Atoms in the 0-character mask area are retained in subsequent processing, while atoms in the 1-character mask area are deleted. The appropriate number of file rows and columns is selected based on the desired model size; the standard setting is 1000 rows × 1000 columns. After recognition, a mask file in txt format is output. For the second phase image, two opposing identification strategies are required. The mask file generated by the strategy of assigning black areas to 0 and white areas to 1 is subsequently used for second-phase modeling. The mask file generated by the strategy of assigning black areas to 1 and white areas to 0 is subsequently used to delete atoms in the second-phase placeholder areas in the matrix model. The number of file rows and columns in this step is adjusted according to the desired model size.
[0036] S5: 5.1 First determine the types of matrix and second phase, and download the corresponding crystal unit cell cif file from the Material Project crystal library. Multiple crystal orientations of the same crystal correspond to multiple cif files. For the matrix single grain model and the second phase model, each model needs to first establish a single crystal model of each grain's crystal orientation and rotation direction. The crystallographic information is obtained by reading the txt file obtained by S3. Use Atomsk software to complete the establishment of several matrix single crystal models. The model box sizes are equal, and the number of models is equal to the number of grains. The modeling input of each model is the crystal orientation and rotation direction txt format file of each grain output from S3. First, identify its grain crystal orientation. Different crystal orientations correspond to different cif files. Use the Atomsk-polycrystal module to complete the matrix single crystal model by copying single crystal models of different crystal orientations and assigning each model a corresponding rotation angle. The same modeling process is adopted for the second phase model. For the matrix grain boundary model, it is only necessary to establish a standard crystal orientation matrix single crystal model with the same box size.
[0037] 5.2 Use the Atomsk-select module to read the single crystal model and the mask file generated in S3, and implement modeling through masking. For the matrix grain boundary model, retain the atoms masked by the 0 character in the model and delete the remaining atoms to complete the matrix grain boundary modeling. For the single grain model, without changing the model box size, mask modeling is used to obtain several single grain models, with the number of models equal to the number of grains. To ensure that each grain has a fixed spatial position in the matrix, both grain boundary atoms and atoms within the grain are retained. The second phase model follows the same modeling process as the matrix grain boundary model.
[0038] S6: Use the Atomsk-merge module and the Lammps-read data-add module to merge the matrix grain boundary model, several single-grain models, and several second-phase single-grain models. The Atomsk-merge module is used to merge several single-grain models into a matrix full-grain model and a second-phase full-grain model, respectively. The Lammps-read data-add module is used to merge the matrix grain boundary model, the matrix full-grain model, and the second-phase full-grain model. After merging, all overlapping atoms need to be deleted in Lammps to complete the matrix model doped with the second phase.
[0039] S7: Based on the atomic ratio of each element obtained in S2, the number of atoms of each element in the model established in S6 is counted, and the number of atoms of each element missing in the current model, that is, the number of solid solution doping atoms of each element, is calculated. The calculation needs to be performed element by element to ensure that the final doping ratio is consistent with the data in S2;
[0040] S8: Use the set type / ratio and create_atom commands of Lammps software to model the solid solution doping of elements. After completing the doping of each element, count the number of existing matrix atoms and recalculate the Lammps solid solution parameters of the next element.
[0041]
[0042] p Fe ,p Mo ,p Cr ,p V is the atomic ratio of each element obtained by EPMA in S2,
[0043] are the number of Fe, Mo, Cr, V, and C atoms in the current model, which can be obtained through Lammps software output or visualization software Ovito;
[0044] N total is the sum of all atoms in the model, which can be calculated using the current parameters:
[0045]
[0046] According to the atomic ratio p of each element obtained by EPMA, the total number of Mo, Cr, V, and C atoms in the model is calculated.
[0047]
[0048] First, Mo atoms are dissolved and replaced by iron atoms to achieve substitutional solid solution. The solid solution parameters used in Lammps are:
[0049]
[0050] Next, the Cr atoms are dissolved. Since the number of existing iron atoms changes, the dissolution parameters used in Lammps are:
[0051]
[0052] Finally, the V atoms are dissolved and the dissolution parameters used in Lammps are:
[0053]
[0054] The solid solution C atoms are interstitial solid solution, and are inserted directly using the create_atom command in Lammps. Number of C atoms:
[0055]
[0056] S9: Based on the above all-atom model, complete the in situ stretching molecular dynamics simulation, simulate the tensile stress by applying velocity at the upper and lower boundaries of the model (usually 10% of the model width), compare the results of conventional macroscopic tensile tests with multiple molecular dynamics simulations, and select the appropriate velocity value according to the tensile fracture mode. The potential function is obtained from the InteratomicPotentials Repository to ensure that each element and any other element contain a potential function. Elements without existing potential functions need to use LJ potential fitting or first-principles calculations. Try to select a multivariate potential function that includes all elements to ensure simulation accuracy.
[0057] In summary, the present invention discloses a molecular dynamics simulation method and system for in-situ tensile test of micro crack initiation of bearing steel material, which obtains the crystal structure, matrix crystallographic information, and crystal information containing the second phase of the area requiring in-situ tensile test; obtains the elemental qualitative and quantitative analysis results of each tissue in the test area; imports the ctf file containing the crystal structure, matrix crystallographic information, and crystal information containing the second phase of the area requiring in-situ tensile test into the Matlab-MTEX module for EBSD data post-processing; uses the MTEX module to output the matrix grain boundary map, matrix single grain map, and second phase phase map of the test area, and outputs a picture file in jpg or png format; Import the image file into the Python-skimage module, and convert the above image file into a mask file in txt format; import the mask file into the Atomsk software, and establish the grain boundary model, single grain model, and second phase model respectively through masking; merge the grain boundary model, single grain model, and second phase model to establish a full-atom model of the matrix doped with the second phase; obtain the atomic percentage of the elements, remove the atoms contained in the second phase, and calculate the number of solid solution doping atoms of each element missing in the current model; use Lammps software to realize the solid solution doping of atoms of each element and complete in situ tensile molecular dynamics modeling; use Lammps software to complete tensile molecular dynamics simulation.
[0058] Through these detailed implementation methods, it is possible to ensure that each technical solution in this application is implemented accurately and completely, while also ensuring the operability and repeatability of the method.
[0059] The above content is only for explaining the technical idea of the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.
Claims
1. A molecular dynamics simulation method for in-situ tensile testing of micro crack initiation in bearing steel materials, characterized in that: The following steps are involved: Obtaining the matrix crystal structure, matrix crystallographic information, and crystal information containing the second phase in the area requiring in-situ tensile testing; Obtain qualitative and quantitative analysis results of elements in the test area; Post-processing of EBSD data is performed based on the crystal structure of the area to be tested for in-situ tensile testing, the matrix crystallographic information, and the crystal information containing the second phase to obtain the matrix grain boundary map, matrix single grain map, and second phase map of the test area; Convert matrix grain boundary map, matrix single grain map, and second phase map into mask files; Constructing a grain boundary model, a single grain model, and a second phase model based on the mask file and in combination with the mask; The grain boundary model, single grain model, and second phase model are combined to obtain a full-atom model of the matrix doped with the second phase; The atomic percentage of each element obtained by EPMA was used to remove the atoms contained in the second phase and calculate the number of solid solution doping atoms of each element missing in the matrix full-atom model; Use Lammps software to achieve solid solution doping of each element's atoms and complete in-situ tensile molecular dynamics modeling; Stretched molecular dynamics simulations were performed using Lammps software.
2. The molecular dynamics simulation method for in-situ tensile testing of microcrack initiation in bearing steel material according to claim 1, characterized in that: The crystal structure of the test area in step 1, the matrix crystallographic information including the crystal orientation and rotation direction, the crystal information of the second phase including the crystal type, crystal orientation, rotation direction, position, and size of the second phase, the matrix crystal structure, matrix crystallographic information, and the crystal information of the second phase are obtained using the electron backscatter diffraction method and saved in a ctf format file that can be post-processed.
3. The molecular dynamics simulation method for in-situ tensile testing of microcrack initiation in bearing steel material according to claim 1, characterized in that: The qualitative and quantitative analysis results of the elements in the test area are obtained using an electron probe microanalyzer, and the atomic percentage of each element is output.
4. The molecular dynamics simulation method for in-situ tensile testing of micro crack initiation in bearing steel material according to claim 1, characterized in that: The post-processing of EBSD data is completed through the Matlab-MTEX module. By reading the ctf format file containing the matrix crystal structure, matrix crystallographic information, and crystal information of the second phase of the area to be subjected to in-situ tensile testing, the grain boundary map of the matrix and the second phase phase map are output. The image size is equal to the overall size of the test area.
5. The molecular dynamics simulation method for in-situ tensile testing of micro crack initiation in bearing steel material according to claim 4, characterized in that: The Matlab-MTEX module outputs matrix grain maps while preserving the overall size of the test area. Each map contains only one grain, and the number of maps is equal to the number of matrix grains.
6. The molecular dynamics simulation method for in-situ tensile testing of micro crack initiation in bearing steel material according to claim 1, characterized in that: In step 5, the Python-skimage module is used to convert the matrix grain boundary map, matrix single grain map, and second phase map of the test area into a mask file in txt format, and then binarized so that the mask file consists of 0 and 1, and the size is adjusted according to the required model size.
7. The molecular dynamics simulation method for in-situ tensile testing of micro crack initiation in bearing steel material according to claim 6, characterized in that: The Atomsk-select module is used to read the mask file in txt format, delete the atoms in the area masked by 1 character, and establish the matrix grain boundary model, matrix single grain model and second phase model.
8. The molecular dynamics simulation method for in-situ tensile testing of micro crack initiation in bearing steel material according to claim 1, characterized in that: The solid solution doping percentage of each element used in Lammps software is calculated through the total number of atoms in the model, the atomic percentage of the elements obtained by EPMA and the number of atoms of each element in the existing model.
9. A molecular dynamics simulation system for in-situ tensile testing of micro crack initiation in bearing steel materials, characterized in that: Including data acquisition module, call module and simulation module; The data acquisition module is used to obtain the matrix crystal structure, matrix crystallographic information, crystal information containing the second phase, and qualitative and quantitative element analysis results of the test area in the area requiring in-situ tensile testing; The calling module is used to call the simulation module to process data and obtain the data or files obtained by the module; The simulation modules include Matlab-MTEX module, Python-skimage module, -polycrystal module and -select module of Atomsk software, and Lammps software.
10. The molecular dynamics simulation system for in-situ tensile testing of micro crack initiation in bearing steel materials according to claim 9, characterized in that: The Matlab-MTEX module is used for EBSD data post-processing; the Python-skimage module is used to convert image files containing matrix grain boundary maps, matrix single grain maps, and second phase phase maps into mask files in txt format; the -polycrystal module and -select module of the Atomsk software are used to combine mask files to establish grain boundary models, single grain models, and second phase models through masking; Lammps software realizes microscopic tensile molecular dynamics simulation.