Aluminum nitride ceramic microscopic material removal process analysis method considering roughness effect

Abstract

In order to explore the material removal mechanism of aluminum nitride ceramic in the friction chemical polishing process, molecular dynamics under a reaction force field is used, the roughness of a workpiece and a grinding disc is used as an important consideration factor during simulation, and the aluminum nitride ceramic micro material removal process analysis method considering the roughness effect is provided. The method comprises the following steps: firstly, constructing a molecular dynamics simulation model of an aluminum nitride ceramic matrix and an aluminum oxide ceramic matrix with different rough peak surfaces, and simulating a scratching process between the aluminum oxide ceramic matrix and the aluminum nitride ceramic matrix in a large-scale atom / molecule parallel simulator LAMMPS; then, the simulated track is observed and analyzed in visual software OVITO, and the influence mechanism of different roughness factors on the material removal process of scratching the aluminum nitride ceramic substrate by the aluminum oxide ceramic substrate in the friction chemical polishing process, including the number of bonds generated on the surface of the aluminum nitride ceramic under different roughness factors, the number of the bonds generated on the surface of the aluminum nitride ceramic under different roughness factors and the number of the bonds generated on the surface of the aluminum nitride ceramic under different roughness factors, is determined. The temperature change and the stress change in the scratching process are avoided.

Description

Technical Field

[0001] This invention relates to the field of precision / ultra-precision machining, and in particular to an analytical method for the removal process of aluminum nitride ceramic micromaterials that takes into account the roughness effect. Background Technology

[0002] Aluminum nitride ceramics are wurtzite crystals with strong covalent bonds and phonons that rapidly transport heat energy. This characteristic endows aluminum nitride ceramics with excellent thermal conductivity, making aluminum nitride ceramic substrates an important component in the field of power electronics. Besides the inherent properties of the material, processing precision also significantly affects the performance of aluminum nitride ceramic substrates; for example, the surface roughness of the workpiece has a significant impact on the heat dissipation of aluminum nitride ceramic substrates. Studies have shown that aluminum nitride ceramic substrates with low surface roughness can effectively reduce internal resistance and further enhance the substrate's heat dissipation capacity.

[0003] Currently, the most common processing method for aluminum nitride ceramics is chemical mechanical polishing (CMP), which is a type of free abrasive polishing and has a low material removal rate. To improve this low material removal rate, patent CN201710160779.7 proposes a method using alumina ceramic as a polishing disc to polish aluminum nitride ceramics. This method significantly improves the material removal rate during polishing and is less likely to form a damage layer on the aluminum nitride ceramic surface. The basic principle of this method is based on tribochemical polishing. The frictional heating process enhances the chemical activity of the workpiece surface, making it easier for the workpiece surface to react with the chemicals in the polishing slurry, thus forming a softening layer on the workpiece surface. This softening layer is removed under the mechanical action of the polishing disc. However, this patent does not mention the influence of the surface roughness of the polishing disc on the entire polishing process. As can be seen from the above, the frictional heating process is extremely important for tribochemical polishing. To ensure rapid heating of the polishing area during polishing, the initial surface roughness of the polishing disc should not be too low. However, the material removal process in tribochemical polishing currently operates at the micron or even nanometer level, and its microscopic material removal mechanism still needs further investigation. In view of this, many scholars have studied and analyzed the processing of materials such as aluminum nitride, silicon carbide, polycrystalline copper, and polycrystalline silicon based on molecular dynamics. However, most of them only considered mechanical effects, and the processing mechanism described may deviate from the actual situation. Secondly, most molecular dynamics simulations in the polishing process establish the model as a smooth plane, but in the actual processing, the initial surface of the workpiece is not smooth, but is composed of many rough peaks. In addition, the roughness of the grinding wheel itself is high, but the rough peaks on the surface of the grinding wheel are not taken into account in the simulation. Summary of the Invention

[0004] This invention aims to explore the material removal mechanism of aluminum nitride ceramics during tribochemical polishing. It utilizes molecular dynamics under a reactive force field and considers the roughness of the workpiece and grinding disc as important factors in the simulation, proposing an analytical method for the microscopic material removal process of aluminum nitride ceramics that takes roughness effects into account. This method simulates the scratching process between alumina and aluminum nitride ceramic substrates using the large-scale atomic / molecular parallel simulator LAMMPS, and observes and analyzes the simulated trajectory using the visualization software OVITO to clarify the influence mechanism of different roughness factors on the material removal process of the alumina ceramic substrate scratching the aluminum nitride ceramic substrate during tribochemical polishing.

[0005] The technical solution of this application is as follows:

[0006] An analytical method for removing micromaterials from aluminum nitride ceramics considering roughness effects includes the following steps: Step 1, constructing molecular dynamics simulation models of aluminum nitride and alumina ceramic matrices with different roughness peaks; Step 2, performing energy minimization on the constructed models using the large-scale atomic / molecular parallel simulator LAMMPS to bring the models to their ground state; Step 3, layering the models according to simulation requirements, dividing the aluminum nitride and alumina ceramic matrices into fixed layers and Newton layers respectively, and assigning potential functions and setting potential parameters for each atom in the model; Step 4, further refining the model to be refined. In step 5, the atoms within the simulated layer relax under canonical ensemble conditions until a set temperature is reached to bring the entire system to equilibrium. Then, the Reaxff reaction force field is set on the large-scale atomic / molecular parallel simulator LAMMPS, and the initial simulation parameters are configured. The polishing process of an alumina ceramic matrix onto an aluminum nitride ceramic matrix is ​​then simulated. Finally, the simulation results obtained in step 5 are visualized and dislocations are analyzed using OVITO software. Based on the analysis results, the effects of different surface roughness of the alumina ceramic matrix on the number of bond formations and the temperature and stress changes during polishing are determined.

[0007] Furthermore, in step 1, based on the space groups of aluminum nitride ceramics and alumina ceramics, corresponding cells are searched in an open-source database; then, the non-orthogonal cells are orthogonalized using ATOMSK software; followed by cell expansion processing to generate the initial molecular dynamics simulation model. Constructing the initial model in this manner ensures its geometric correctness, providing an ideal starting point for subsequent energy minimization and equilibrium processes. During modeling, both models can be designed as cuboids with a size of 60. 60 50 angstroms. Then, based on the dimensions, atomic number, and set roughness parameters of the initial aluminum nitride ceramic matrix model and alumina ceramic matrix model in each direction, the surface function is used to remove atoms from the model, generating rough peaks on the model surface. By introducing controllable surface inhomogeneities, the model can more closely resemble reality, thus enabling the study of more practical polishing processes.

[0008] Furthermore, in step 2, during the energy minimization process, the tolerance for force and the tolerance for energy are set to 1. 10 -16 The maximum number of iterations should be no less than 1000. High-precision settings ensure the stability of subsequent molecular dynamics simulations.

[0009] Furthermore, in step 3, the top tenth of the alumina ceramic matrix and aluminum nitride ceramic matrix model can be designated as a fixed layer, while the remaining portion can be designated as a Newton layer. The fixed layer ensures that atoms do not undergo unexpected displacements during relaxation, while the Newton layer guarantees the degrees of freedom of the atoms. During the simulation, a velocity is applied to the fixed layer of the alumina ceramic matrix, causing it to drive the Newton layer to rub against the Newton layer of the aluminum nitride ceramic matrix. This avoids conflicts between the canonical ensemble integrals in the fixed layer and the Newton layer, ensuring the accuracy of the simulation.

[0010] The formula for calculating the potential function is: E total =E bond +E over +E val +E tors +E vdwaals +E Coulomb In the formula, E bond Represents bond energy, E over E represents coordination energy. val E represents bond angle energy. tors E represents the energy generated by dihedral torsion. vdwaals E represents the energy generated by the van der Waals force. Coulomb This represents the energy generated by the Coulomb force.

[0011] Furthermore, in order to simulate the wet polishing process of an alumina ceramic substrate on an aluminum nitride ceramic substrate, in step 1, a water molecule model is additionally constructed on the large-scale atomic / molecular parallel simulator LAMMPS as a polishing medium.

[0012] Furthermore, the analysis process in step 6 includes: statistically analyzing the number of bonds generated based on the dynamic changes in the number of atoms in different regions during the simulation; statistically analyzing the temperature of the contact area during the simulation, as well as the changes in force in the x and z directions (specifically, the model is first divided into grids of known size, then the stress in each grid is statistically analyzed, the average value is calculated, and then a stress cloud map is generated based on the time change).

[0013] Compared with existing technologies, the present invention's method for analyzing the micromaterial removal process of aluminum nitride ceramics considering roughness effects first constructs molecular dynamics simulation models of aluminum nitride ceramic substrates and alumina ceramic substrates with different roughness peaks. The scratching process between the alumina ceramic substrate and the aluminum nitride ceramic substrate is simulated in the large-scale atomic / molecular parallel simulator LAMMPS. Then, the simulated trajectory is observed and analyzed in the visualization software OVITO to study the number of bonds formed on the aluminum nitride surface under different roughnesses, as well as the temperature and stress changes during the scratching process. This clarifies the influence mechanism of different roughness factors on the material removal process of the alumina ceramic substrate scratching the aluminum nitride ceramic substrate. The simulation yielded the following conclusions: In tribochemical polishing, the surface roughness of the alumina ceramic substrate has a significant impact on the polishing process. When the roughness coefficient of the alumina ceramic is large, the temperature rises rapidly during friction, and more water molecules combine with the aluminum nitride ceramic surface to form a softening layer. However, as the roughness peaks between the two are continuously eliminated, the contact probability between the roughness peaks decreases, and the temperature in the contact area decreases. Therefore, the atomic removal rate shows a gradual decreasing trend in the later stages. As can be seen from the above phenomena, when using an alumina ceramic grinding disc for tribochemical polishing of aluminum nitride ceramics, the surface of the alumina ceramic grinding disc should always have a certain roughness so that the friction can activate the ceramic surface, promote the chemical reaction, and thus enhance the material removal rate and improve the surface quality of the workpiece. Attached Figure Description

[0014] Figure 1 This is a flowchart of the analytical method for removing aluminum nitride ceramic micromaterials considering the roughness effect, as described in this application.

[0015] Figure 2 It is a model of alumina ceramics, aluminum nitride ceramics, water molecules, and combinations thereof;

[0016] Figure 3 It is the process of alumina ceramic matrix scratching aluminum nitride ceramic matrix;

[0017] Figure 4 It shows the temperature changes at the contact interface and the stress changes in the X and Z directions during the scratching process;

[0018] Figure 5It refers to the change in the number of Al-O, Al-H, and Al-N bonds throughout the entire scratching process;

[0019] Figure 6 It describes the bonding situation on the surface of the aluminum nitride ceramic substrate model during the scratching process;

[0020] Figure 7 This is a diagram showing the stress change in the aluminum nitride ceramic matrix in the Z direction after the scratching process.

[0021] Figure 8 It shows the surface morphology of the wear marks on the surface of aluminum nitride ceramic after dry and wet friction and wear tests;

[0022] Figure 9 These are the photoelectron spectroscopy results of the wear tracks after two types of friction and wear experiments;

[0023] Figure 10 These are the EDS elemental detection results for the wear-scarred and non-wear-scarred areas;

[0024] Figure 11 These are the scanning electron microscope (SEM) results of the wear marks after two types of friction and wear experiments.

[0025] Figure 12 The results show the surface morphology of aluminum nitride ceramics after polishing with alumina ceramic discs of different roughness, as well as the surface profilometer test results. Detailed Implementation

[0026] The present application will be further described below with reference to the accompanying drawings and embodiments, but this should not be construed as limiting the present application. Contents not described in detail in the following embodiments are all common knowledge in the art.

[0027] To investigate the influence of different roughness coefficients on the material removal process of an alumina ceramic substrate rubbing against an aluminum nitride ceramic substrate during tribopolishing, this invention provides an analytical method for the microscopic material removal process of aluminum nitride ceramics that considers the roughness effect. This method simulates the rubbing process between the alumina ceramic substrate and the aluminum nitride ceramic substrate in the large-scale atomic / molecular parallel simulator LAMMPS, and observes and analyzes the simulated trajectory in the visualization software OVITO to obtain the number of bonds generated on the aluminum nitride ceramic surface under different roughnesses, as well as the temperature and stress changes during the rubbing process.

[0028] Implementation Case:

[0029] See Figure 1 The analytical method for removing aluminum nitride ceramic micromaterials considering roughness effects in this implementation case includes the following steps.

[0030] Step 1: Construct molecular dynamics simulation models for aluminum nitride ceramic matrices and alumina ceramic matrices with different roughness peak surfaces.

[0031] First, aluminum nitride ceramic cells with space group P63mc and alumina ceramic cells with space group C2 / m were located in the open-source database (Material Project). Then, the non-orthogonal cells were orthogonalized using ATOMSK software. Next, cell expansion was performed to generate the initial molecular dynamics simulation model. The aluminum nitride and alumina ceramic matrix models are cuboid in shape with a size of 60. 60 50 angstroms. Simultaneously, a water molecule model was constructed on the large-scale atomic / molecular parallel simulator LAMMPS as a polishing medium.

[0032] Then, based on the dimensions, number of atoms, and set roughness parameters (the roughness parameters are set according to the desired surface roughness peak distribution) of the initial aluminum nitride ceramic matrix model and alumina ceramic matrix model in each direction, the surface function is used to remove atoms from the model and generate roughness peaks on the model surface. This function contains 6 parameters: sigma, H, Lx, m, n, and qr; where sigma represents the standard deviation; H represents the Hurst exponent (roughness index), which ranges from 0 to 1; Lx refers to the morphology in the x-direction; m refers to the number of elements in the x-direction, i.e., the number of atoms in the x-direction; n refers to the number of elements in the y-direction; and the qr parameter controls the randomness of the surface roughness peaks, with a larger value resulting in a more random distribution of roughness peaks.

[0033] In this implementation case, the specific parameters of the alumina ceramic matrix model are as follows: initial atomic number is 15300, sigma value is 4, H exponent is 0.5, Lx value is 78.182896, m and n values ​​are 16, and qr value is 0.9. Figure 2 As shown in (b). The specific parameters of the aluminum nitride ceramic matrix model are as follows: initial atomic number is 10240, sigma value is 3, H exponent is 0.4, Lx value is 65.104576, m and n values ​​are 16, qr value is 0.8, and maximum height in the z direction is 20.34518, as shown in (b). Figure 2 As shown in (a), the water molecule model is a TIP4P type model generated in LAMMPS, with 200 water molecules. Since the interaction force field between H and O atoms in the water molecule does not need to be considered in the reaction force field, the generated water molecule only contains the coordinate positions of each atom. The same applies to the alumina ceramic matrix and aluminum nitride ceramic matrix models. The final merged model is shown below. Figure 2 As shown in (c).

[0034] Step 2: Perform energy minimization on the constructed model using the large-scale atomic / molecular parallel simulator LAMMPS, setting the force tolerance and energy tolerance to 1. 10 -16 The maximum number of iterations is 1000, which puts the model in its ground state.

[0035] Step 3: Based on the simulation requirements, the model is layered. The portion of the alumina ceramic matrix model and the aluminum nitride ceramic matrix model extending 5 angstroms upwards from the bottom layer is designated as a fixed layer, while the remaining portions are designated as Newton layers (e.g., ...). Figure 2 As shown in the figure, potential functions are assigned and potential parameters are set for each atom in the model. The formula for calculating the potential function is as follows: E total =E bond +E lp +E over +E under +E val +E pen +E coa +E C2 +E tors +E conj +E H-bonds +E vdwaals +E Coulomb In the formula, E bond Represents bond energy, E lp E represents the lone pair electron energy. over E represents coordination energy. under E represents the undercoordination energy. val E represents bond angle energy. pen E represents the penalty value for alkene molecules. coa E represents the conjugate bond angle energy. C2 E represents the correction energy of a carbon-carbon double bond. tors E represents the energy generated by dihedral torsion. conj E represents the conjugate dihedral angle energy. H-bonds The energy produced by hydrogen bonding, E vdwaals The energy generated by van der Waals forces, E Coulomb This represents the energy generated by the Coulomb force.

[0036] Since the system involved in the simulation in this application is a covalent bond system, the formula can be simplified to: E total =E bond +E over +E val +E tors +E vdwaals +E Coulomb .

[0037] Step 4: Relax the atoms in the layer to be simulated under canonical ensemble conditions until the set temperature (298.15K) is reached so that the whole system is in equilibrium; the temperature damping coefficient is 20.

[0038] Step 5: Set the Reaxff reaction force field on the large-scale atomic / molecular parallel simulator LAMMPS, and set the initial simulation parameters. The nearest neighbor parameter in the simulation is set to 3.0, the simulation unit is real, the energy unit is Kcal / mol, the time unit is femtosecond, the distance unit is angstrom, the time step in the simulation is 0.2 femtosecond, the boundary condition is periodic boundary, and the atom type is full. Then simulate the process of alumina ceramic matrix scratching (polishing) aluminum nitride ceramic matrix (e.g., Figure 3 As shown in the figure; it can be seen from the figure that as the alumina ceramic substrate is scratched, the rough peaks on the surface of the aluminum nitride ceramic substrate are gradually removed.

[0039] During the rubbing process, each Newton layer is set to remain under a canonical ensemble. To avoid conflicts in the integration method, a velocity is applied to the fixed layer of the alumina ceramic substrate (using a fix-move linear method), allowing the fixed layer to drive the Newton layer to rub the Newton layer of the aluminum nitride ceramic substrate.

[0040] Step 6: Visualize and analyze the simulation results obtained in Step 5 using OVITO software, including: counting the number of bonds generated based on the dynamic changes in the number of atoms in different regions during the simulation; and counting the temperature of the contact region and the changes in forces in the x and z directions during the simulation.

[0041] Based on the analysis results, the effects of different surface roughness of the alumina ceramic matrix on the number of bonds formed and the temperature and stress changes during the polishing process can be obtained.

[0042] Use the compute command in the software to calculate the forces and temperatures during the simulation process. See also Figure 4 (a) represents the temperature change over time during the simulation; see also Figure 4 (b) represents the change of force in the X and Z directions over time during the simulation; by calculating the ratio of the forces in the x and z directions, the instantaneous friction coefficient during the simulation can be obtained. Figure 4 As can be seen, the force in the X direction increases significantly with the scratching process, and gradually decreases as atoms are removed from the surfaces of the alumina and aluminum nitride ceramic substrates. Furthermore, a significant temperature increase occurs at 1 ps, at which point... Figure 5The change in the number of bonds shows that with the increase of force and temperature in the X direction, the original Al-N chemical bonds in the system decrease significantly, while the Al-O and Al-H chemical bonds increase significantly. Therefore, the changes in force and temperature in the X direction indicate that the increase in temperature promotes the reaction on the surface of the aluminum nitride matrix.

[0043] Using the `fix bonds / reaxff` command in the software, bonding information between individual atoms during the simulation can be obtained. In this implementation example, during post-processing, Al, O, and H atoms are selected, and the generated bond level file is statistically analyzed. The bond level data in the corresponding columns is exported, showing the bond level relationships between neighboring atoms. Finally, the changes in the number of Al-O, Al-H, and Al-N bonds over time are generated, as shown below. Figure 5 As shown in the figure, as the scratching process continues, the number of Al-O and Al-H bonds in the system continuously increases, while the number of Al-N bonds decreases. This indicates that a chemical reaction occurs on the surface of the aluminum nitride ceramic substrate, forming a reaction layer as the simulation progresses. Throughout the scratching process, Al-O, Al-H, and Al-OH bonds (such as...) are formed on the surface of the aluminum nitride ceramic substrate. Figure 6 (as shown) and other types.

[0044] When statistically analyzing the force variations in the x and z directions, the model is first divided into blocks of known-size grids. Then, the stress within each grid is statistically analyzed, the average value is calculated, and finally, a stress contour plot is generated based on the time variation. The stress variations in the Z direction during the initial and final stages of the simulation are shown below. Figure 7 (a) and Figure 7 As shown in (b), it can be seen from the figure that a certain stress layer is generated in the model due to the relaxation process at the beginning of the scratching stage, and the surface residue decreases significantly after the scratching ends.

[0045] Experimental verification:

[0046] To verify the feasibility of using molecular dynamics methods to simulate the tribochemical polishing of aluminum nitride ceramics, the applicant also conducted actual tribochemical wear experiments and specific aluminum nitride ceramic polishing experiments, as detailed below.

[0047] Step 1: Friction and wear experiments were conducted on alumina ceramics and aluminum nitride ceramics using both dry and wet friction and wear methods; specifically, the rotational friction and wear method was used. In the wet friction and wear experiment, water was used as the medium, and the aluminum nitride ceramic was continuously immersed in water. The rotational speed was 240 rpm, the deflection radius was 5 mm, the constant pressure was 1 kg, and the experiment lasted 30 minutes.

[0048] Step 2: Surface morphology and photoelectron spectroscopy (PES) were performed on the aluminum nitride ceramic after the tribological test, focusing on both the wear-marked and non-wear-marked areas. The surface morphology is shown below. Figure 8 As shown, Figure 8 (a) represents dry friction wear. Figure 8 (b) represents wet friction and wear; as can be seen from the figure, the friction and wear experiment in a water-containing environment produced wider wear tracks. The energy dispersive spectroscopy (EDS) results are as follows: Figure 9 As shown, Figure 9 (a) represents dry friction wear. Figure 9 (b) illustrates wet friction wear, where a distinct non-aluminum nitride substance appears in the wear track area. Peak segmentation of the photoelectron spectroscopy reveals that the newly generated substance is likely aluminum oxide, which corresponds to the simulation results in the implementation case, further verifying the formation of a reaction layer on the workpiece surface by the tribochemical process. Furthermore, surface EDS elemental analysis was performed on the friction-wear surface, and the results are as follows: Figure 10 As shown, the testing primarily focused on the wear-scarred and non-wear-scarred areas. It was observed that under both frictional wear conditions, the oxygen content in the wear-scarred area significantly increased compared to the non-wear-scarred area, further indicating a chemical reaction occurring on the aluminum nitride ceramic surface under friction. Furthermore, the applicant also conducted surface SEM analysis on the friction-weared surface, with results as follows: Figure 11 As shown, Figure 11 (a) represents dry friction wear. Figure 11 (b) represents wet friction and wear; it can be seen that there are obvious ablation points on the surface of dry friction and wear, which may affect the surface quality of aluminum nitride ceramics after processing, while the surface of wet friction and wear does not show a similar phenomenon and the surface is relatively smooth; indicating that there is better friction and wear performance under wet conditions.

[0049] Step 3: To further illustrate the importance of surface roughness to the tribochemical polishing process and to demonstrate the accuracy of the analytical conclusions of this application, aluminum nitride ceramics were polished using alumina polishing discs with different roughnesses. Specifically, diamond with particle sizes of W2.5 and W20 was used to prepare the initial average surface roughness R. a Alumina polishing disks with values ​​of 123nm and 495nm were used to evaluate the initial average surface roughness R. a Aluminum nitride ceramics with a wavelength of 430 nm were polished using water as the polishing fluid at a rotation speed of 150 rpm and a pressure of 5 kg. The surface morphology of the polished aluminum nitride ceramics is shown below. Figure 12 (a) and Figure 12 As shown in (b), the initial surface roughness R is visible. aAn alumina ceramic polishing pad with a value of 495nm provides better polishing results, and the surface roughness R of aluminum nitride after polishing is improved. a The value reached 9.2nm, and there were no obvious pit defects on the surface.

[0050] The foregoing general description of the invention and its specific embodiments should not be construed as a limitation on the technical solution of the invention. Those skilled in the art, based on the disclosure of this application, can add, reduce, or combine the disclosed technical features in the foregoing general description and / or specific embodiments (including examples) without departing from the constituent elements of the invention, to form other technical solutions within the scope of protection of this application.

Claims

1. A method for analyzing a micro material removal process of aluminum nitride ceramic considering roughness effect, characterized in that, The method comprises the following steps: Step 1, constructing a molecular dynamics simulation model of an aluminum nitride ceramic substrate and an aluminum oxide ceramic substrate with different rough peak surfaces; Step 2, performing energy minimization processing on the constructed model on a large-scale atom / molecule parallel simulator LAMMPS to make the model in a ground state; Step 3, performing hierarchical processing on the model according to the needs of simulation, dividing the aluminum nitride ceramic substrate and the aluminum oxide ceramic substrate into fixed layers and Newton layers respectively, and performing potential function distribution and potential parameter setting on each atom in the model; Step 4, relaxing the atoms in the layer to be simulated under the condition of canonical ensemble until a set temperature is reached, so that the whole system is in an equilibrium state; Step 5, setting the Reaxff reaction force field on the large-scale atom / molecule parallel simulator LAMMPS, setting the initial parameters of the simulation, and then simulating the process of polishing the aluminum nitride ceramic substrate by the aluminum oxide ceramic substrate; Step 6, visualizing and dislocation analyzing the simulation results obtained in step 5 by OVITO software; based on the analysis results, the influence of different roughness of the aluminum oxide ceramic substrate surface on the number of bond generation and the change of temperature and stress in the polishing process is obtained.

2. The method of claim 1, wherein the method is a method of analyzing a microscale material removal process of aluminum nitride ceramic considering roughness effects, the method comprising: In the step 1, according to the space group of the aluminum nitride ceramic and the aluminum oxide ceramic, the corresponding cell is searched in the open source database; then the non-orthogonal cell is orthogonalized by the ATOMSK software; then the cell expansion processing is performed to generate the initial molecular dynamics simulation model.

3. The method of claim 2, wherein the method further comprises: According to the size, the number of atoms and the set roughness parameter of each direction of the initial aluminum nitride ceramic substrate model and the aluminum oxide ceramic substrate model, the surface function is used to remove the atoms in the model, and the rough peaks are generated on the surface of the model.

4. The method for analyzing the removal process of aluminum nitride ceramic micromaterials considering roughness effects according to claim 1, characterized in that: In the energy minimization process, the tolerance of force and the tolerance of energy are set to 1 10 -16 , and the maximum iteration steps are not less than 1000 steps.

5. The method for analyzing the removal process of aluminum nitride ceramic micromaterials considering roughness effects according to claim 1, characterized in that: In the step 3, the aluminum oxide ceramic substrate and the aluminum nitride ceramic substrate model are set as fixed layers from the bottom layer to the upper one-tenth part, and the remaining part is set as Newton layers.

6. The method of claim 5, wherein the method further comprises: The formula for calculating the potential function is as follows: E total = E bond + E over + E val + E tors + E vdwaals + E Coulomb ; in which E bond represents bond energy, E over represents over-coordination energy, E val represents bond angle energy, E tors represents dihedral angle torsion energy, E vdwaals represents energy produced by van der Waals forces, and E Coulomb represents energy produced by Coulomb forces.

7. The method of claim 1, wherein the method further comprises: determining a material removal rate of the micro-scale material removal process. In the step 5, during the simulation, the fixed layer of the aluminum oxide ceramic substrate is subjected to acceleration, so that the fixed layer drives the Newton layer to realize the scraping of the Newton layer of the aluminum nitride ceramic substrate.

8. The method of claim 1, wherein the method is for analyzing an aluminum nitride ceramic micro material removal process considering roughness effects. In the step 1, a water molecule model is constructed on the large-scale atom / molecule parallel simulator LAMMPS as a polishing medium to simulate the wet polishing of the aluminum oxide ceramic substrate on the aluminum nitride ceramic substrate.

9. The method of claim 1, wherein, The analysis process in the step 6 includes: according to the dynamic change of the number of atoms in different regions during the simulation, the number of bond generation is counted; the temperature of the contact area and the change of the force in the x direction and the z direction during the simulation are counted.

10. The method of claim 9, wherein the method further comprises: When the changes of the forces in the x and z directions are counted, the model is first divided into grids with known sizes, then the stresses in each grid are counted, the average values are calculated, and then the stress cloud map is generated according to the time change.

Images