Micro-texturing bearing lubrication characteristic cross-scale analysis method and related equipment
By employing machine learning to predict homogenization equivalence coefficients in microtextured bearings and embedding the average Reynolds equation, combined with intelligent optimization algorithms, the problem of low calculation efficiency of lubrication characteristics in microtextured bearings is solved, achieving efficient and accurate lubrication characteristic analysis and improved load-bearing capacity.
Patent Information
- Application Number
- CN202511662340.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-13
- Publication Date
- 2026-02-17
AI Technical Summary
Existing technologies suffer from low computational efficiency and modeling difficulties in calculating the lubrication characteristics of microtextured bearings, especially in the case of high-density microtextures. Direct and indirect methods are computationally expensive, while the combination of machine learning and homogenization methods is not yet mature.
Machine learning models are used to predict homogenization equivalence coefficients and embed the average Reynolds equation to achieve macro-micro scale coupled solution of microtextured sliding bearings. Combined with intelligent optimization algorithms, the optimal combination of geometric and working parameters is automatically searched to optimize the design of texture geometry parameters.
It improves computational efficiency, reduces computational costs, maintains high-precision lubrication characteristic analysis, enhances bearing load capacity, reduces computation time by 50%-80% and 20%-40%, and optimizes the design to increase load capacity by 110%-140%.
Smart Images
Figure CN121543483A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of tribology and fluid lubrication technology, and in particular to a method and related equipment for cross-scale analysis of the lubrication characteristics of microtextured bearings. Background Technology
[0002] In recent years, surface texturing technology has become an important means of improving the lubrication performance of sliding bearings. By machining pits, grooves, or other types of patterns on the surface of the friction pair of sliding bearings, the pressure distribution within the bearing can be effectively adjusted, enhancing the bearing's load-bearing capacity and reducing the coefficient of friction. Its core principle lies in the fact that the texture can store lubricating oil, forming a "micro-oil pool," reducing direct metal-to-metal contact, and simultaneously improving oil film stiffness through hydrodynamic effects. For example, under high-speed conditions, regularly arranged pits can utilize the "secondary flow" effect to conduct heat and delay oil film rupture; while under low-speed, heavy-load conditions, the groove structure can more efficiently capture wear debris, extending bearing life. In existing technologies, the lubrication performance of textured bearings is mainly studied through the following two methods:
[0003] 1. Direct Method: This method directly solves the Reynolds equation or the Navier-Stokes equation in the computational domain, capturing the influence of texture on the lubricating film flow field by locally refining the mesh. However, when the texture distribution density is high and the individual texture size is small, the required number of meshes will increase significantly, resulting in extremely high computational costs and severely reducing the efficiency of designing and optimizing textured bearings.
[0004] 2. Indirect Method: Based on the average flow rate method proposed by Patir-Cheng, this method introduces the "pressure flow rate factor" and "shear flow rate factor" to achieve an equivalent description of the flow field effect at the microtexture, thereby simplifying the calculation. However, this method still requires independent solution for each texture unit, and if the number of textures is large, the computational efficiency remains low.
[0005] In addition, the homogenization method has demonstrated high applicability in fluid thin-film lubrication problems on rough or textured surfaces in recent years. Meanwhile, machine learning has shown powerful nonlinear modeling capabilities in friction and lubrication research, capable of predicting friction coefficients, wear rates, or flow field characteristics using existing training data. Combining machine learning with the homogenization method allows for efficient solutions to cross-scale lubrication characteristics through rapid prediction of the homogenization equivalence coefficients at various texture locations. However, a method for effectively integrating the homogenization method and machine learning is currently lacking. Therefore, it is necessary to propose a cross-scale analysis method that combines computational accuracy and high efficiency to address the problems of low computational efficiency and modeling difficulties in the lubrication characteristics of micro-textured sliding bearings. Summary of the Invention
[0006] The purpose of this invention is to provide a cross-scale analysis method and related equipment for the lubrication characteristics of microtextured bearings. This method predicts the homogenization equivalence coefficient through a machine learning model (homogenization equivalence coefficient prediction model) and embeds it into the average Reynolds equation, realizing the macro-micro scale coupled solution of microtextured sliding bearings. While ensuring the accuracy of calculation, it greatly improves the calculation efficiency and reduces the calculation cost.
[0007] To achieve the above objectives, the present invention employs the following technical solution:
[0008] In a first aspect, the present invention provides a method for cross-scale analysis of the lubrication characteristics of microtextured bearings, comprising:
[0009] Obtain structural operation data of the target bearing texture;
[0010] The structural operation data of the target bearing texture is input into the trained homogenization equivalence coefficient prediction model for processing to obtain the homogenization equivalence coefficient of the target bearing texture.
[0011] The average Reynolds equation is solved using the homogenization equivalent coefficient of the target bearing texture to obtain the lubrication characteristics of the target bearing texture; the lubrication characteristics include the pressure distribution and load-bearing capacity of the lubrication film.
[0012] Optionally, the structural operating data includes geometric parameters, distribution parameters, and operating condition parameters;
[0013] The geometric parameters include: the shape, depth, and size of the texture;
[0014] The distribution parameters include: distribution pattern, distribution density, and distribution area;
[0015] The operating parameters include: rotational speed, film thickness, film thickness tilt ratio, and bearing size.
[0016] Optionally, the training process of the homogenized equivalence coefficient prediction model includes:
[0017] Construct a sample dataset of bearing textures;
[0018] The sample dataset is divided into training set, validation set and test set according to a preset ratio;
[0019] The training set is input into the homogenization equivalence coefficient prediction model for training, and the temporary parameters of the model are obtained. Based on the temporary parameters, the temporary homogenization equivalence coefficient prediction model is obtained.
[0020] The validation set is input into the temporary homogenized equivalent coefficient prediction model for processing to obtain the prediction results of the validation set.
[0021] With the goal of minimizing the mean square error between the prediction results on the validation set and the true labels, temporary parameters are optimized to obtain preliminary optimized parameters, and a preliminary homogenized equivalent coefficient prediction model is obtained based on the preliminary optimized parameters.
[0022] The test set prediction results of the preliminary homogenized equivalence coefficient prediction model are used to evaluate the coefficient of determination index, obtain the optimal parameters of the model, and obtain the trained homogenized equivalence coefficient prediction model based on the optimal parameters of the model.
[0023] Optionally, the sample dataset for constructing the bearing texture includes:
[0024] Obtain the structural operation dataset of the bearing texture;
[0025] Utilize orthogonal experimental design structures to run parameter combinations of parameters in the dataset;
[0026] The homogenization equivalence coefficients of each parameter combination are calculated based on the finite element method, and the preliminary sample dataset of bearing texture is obtained by combining the parameter combination and the homogenization equivalence coefficients of the parameter combination.
[0027] The initial sample dataset of bearing texture is subjected to max-min normalization to obtain the sample dataset of bearing texture.
[0028] Optionally, the homogenization equivalence coefficient prediction model includes, but is not limited to: neural networks, support vector machines, random forests, and Gaussian process regression algorithms.
[0029] Optionally, the average Reynolds equation is expressed as follows:
[0030]
[0031] In the formula: x is the macroscopic coordinate of each point within the lubrication area; The gradient of x is calculated; p0(x) is the macroscopic pressure field; h0(x) is the macroscopic average film thickness; μ is the dynamic viscosity of the lubricant; u a,1 The velocity vector in the direction of bearing rotation; u a,2 This is the velocity vector in the direction perpendicular to the bearing. , , and Both represent the permeation tensor coefficients in the homogenization equivalence coefficients; and Both represent the shear flow coefficient in the homogenization equivalent coefficient.
[0032] Secondly, the present invention provides a cross-scale analysis system for the lubrication characteristics of microtextured bearings, comprising:
[0033] The data acquisition module is used to acquire the structural operation data of the target bearing texture;
[0034] The equivalent coefficient prediction module is used to input the structural operation data of the target bearing texture into the trained homogenization equivalent coefficient prediction model for processing, and obtain the homogenization equivalent coefficient of the target bearing texture.
[0035] The analysis module is used to solve the average Reynolds equation using the homogenization equivalent coefficient of the target bearing texture to obtain the lubrication characteristics of the target bearing texture; the lubrication characteristics include the pressure distribution and load-bearing capacity of the lubrication film.
[0036] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor;
[0037] The memory is used to store programs;
[0038] The processor is used to execute the program to implement the cross-scale analysis method for the lubrication characteristics of microtextured bearings as described in the first aspect.
[0039] Fourthly, the present invention provides a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the cross-scale analysis method for the lubrication characteristics of microtextured bearings as described in the first aspect.
[0040] Fifthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the cross-scale analysis method for the lubrication characteristics of microtextured bearings as described in the first aspect.
[0041] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:
[0042] This invention provides a method and related equipment for cross-scale analysis of the lubrication characteristics of microtextured bearings. The method inputs the structural operation data of the target bearing texture into a trained homogenization equivalence coefficient prediction model for processing, obtaining the homogenization equivalence coefficient of the target bearing texture; and combines this with the average Reynolds equation to obtain the lubrication characteristic results. By integrating the homogenization method with a machine learning algorithm (homogenization equivalence coefficient prediction model), cross-scale coupled analysis of microtexture and macroscopic lubrication performance is achieved. Simultaneously, it avoids the increased computation time caused by calculating the homogenization equivalence coefficient of each texture individually, significantly improving computational efficiency compared to traditional methods. It maintains prediction accuracy comparable to direct numerical solutions.
[0043] This invention provides a method and related equipment for cross-scale analysis of the lubrication characteristics of microtextured bearings. The lubrication characteristic results obtained by this method can be combined with intelligent optimization algorithms to automatically search for the optimal combination of geometric and operating parameters with the goal of maximizing load-bearing capacity, thereby achieving automatic optimization design of the texture geometry parameters and improving the bearing load-bearing capacity. Specific intelligent optimization algorithms include, but are not limited to, particle swarm optimization, genetic algorithm, Bayesian optimization algorithm or gradient search algorithm, and the optimization variables can be the geometric dimensions, distribution density, spacing or included angle of the texture, etc. Attached Figure Description
[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0045] Figure 1 The diagram shown is a schematic diagram of a cross-scale analysis process for the lubrication characteristics of a microtextured bearing according to the present invention.
[0046] Figure 2 The diagram shown is a schematic representation of the structure of a microtextured bearing in one embodiment of the present invention.
[0047] Figure 3 The diagram shown is a schematic diagram of the training process of the homogenization equivalence coefficient prediction model in one embodiment of the present invention. Detailed Implementation
[0048] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.
[0049] Example 1
[0050] like Figure 1 As shown in the figure, this invention provides a method for cross-scale analysis of the lubrication characteristics of microtextured bearings, including the following steps:
[0051] S1: Obtain structural operation data of the target bearing texture;
[0052] S2: Input the structural operation data of the target bearing texture into the trained homogenization equivalence coefficient prediction model for processing to obtain the homogenization equivalence coefficient of the target bearing texture;
[0053] S3: Solve the average Reynolds equation using the homogenization equivalent coefficient of the target bearing texture to obtain the lubrication characteristics of the target bearing texture; the lubrication characteristics include the pressure distribution and load-bearing capacity of the lubricating film.
[0054] Furthermore, this example focuses on a type of water-lubricated microtextured bearing. The established bearing surface has a microtexture, and relative motion is created by a motor driving the rotation of another smooth surface component. The lubricating medium is deionized pure water. Figure 2 The diagram shows the structure of a microtextured bearing (target bearing texture). Texture refers to minute geometrical features, either artificially designed or naturally occurring, with specific shapes, sizes, and distributions distributed on a solid surface. Figure 2 In the middle, the texture consists of circular pits regularly distributed within a rectangular area on the inner wall of the bearing.
[0055] Furthermore, the structural operation data of the target bearing texture includes its geometric parameters, distribution parameters, and operating condition parameters;
[0056] Specifically, geometric parameters include the shape, depth, and size of the texture. Common texture shapes include circles, ellipses, rectangles, wedges, and fish shapes; texture depth is the distance between the processed texture and the processed surface; the size of the texture is usually characterized by the characteristic parameters corresponding to its geometric shape. For example, a rectangular texture is described by its length and width, an elliptical texture by its major and minor axes, and so on.
[0057] Specifically, the distribution parameters include: distribution pattern, distribution density, and distribution area. The distribution pattern can be a uniform distribution in a rectangular array, or a radial distribution along the circumference, etc.; the distribution density is used to characterize the quantity density of the texture within a given distribution area; and the distribution area determines the specific distribution location of the texture on the bearing surface. For example, the texture may be distributed at 1 / 4 of the bearing circumference, or it may be a groove evenly distributed on the circumference, etc.
[0058] Specifically, the operating parameters include: rotational speed, film thickness, film thickness tilt ratio, and bearing size. The bearing size determines the dimensional difference between the texture and the bearing film thickness.
[0059] This invention introduces a cross-scale analysis method for the lubrication characteristics of microtextured bearings. By integrating homogenization and machine learning, a unified mapping relationship is established between microtexture features, fluid flow behavior, and macroscopic lubrication performance, providing an efficient, intelligent, and scalable cross-scale analysis and optimization design method for complex friction and lubrication systems.
[0060] like Figure 3 As shown, in this embodiment, the training process of the homogenized equivalent coefficient prediction model in step S2 includes:
[0061] S01: Construct a sample dataset of bearing texture;
[0062] Specifically, the sample dataset for constructing bearing textures includes:
[0063] S011: Obtain the structural operation dataset of the bearing texture;
[0064] S012: Using orthogonal experimental design structures to run parameter combinations of data in the central data set;
[0065] Specifically, the orthogonal experimental design uses parameters from the bearing texture operation data set, with each parameter having 5 levels. It constructs 81 representative experimental combinations using orthogonal arrays, and combines intuitive analysis (range method) and variance analysis to determine the significance of each parameter's influence on bearing performance indicators. Finally, it selects the parameter combination that optimizes the bearing's friction performance and load-bearing capacity. The key parameter levels can be verified and refined through simulation or physical experiments, taking into account both engineering efficiency and optimization accuracy.
[0066] S013: Calculate the homogenization equivalence coefficients of each parameter combination based on the finite element method, and obtain the preliminary sample dataset of bearing texture by combining the parameter combination and the homogenization equivalence coefficients of the parameter combination.
[0067] Specifically, step S013 includes:
[0068] Based on the bearing texture parameter combination designed by orthogonal experiments, a representative volume element (RVE) finite element model containing texture is constructed.
[0069] Boundary conditions and corresponding loads are applied to the RVE model, and the macroscopic average stress is calculated through finite element simulation.
[0070] Based on homogenization theory, the homogenization equivalence coefficients corresponding to each set of parameters are solved by the macroscopic stress-strain relationship.
[0071] The parameters of the orthogonal experiment are matched one-to-one with the calculated homogenization equivalent coefficients, and integrated into a preliminary sample dataset of bearing texture containing input parameters and output coefficients.
[0072] S014: Perform max-min normalization on the preliminary sample dataset of bearing texture to obtain the sample dataset of bearing texture.
[0073] S02: Divide the sample data set into training set, validation set and test set according to a preset ratio;
[0074] Specifically, the proportion of the training set is 60%-80%, the proportion of the validation set is 20%-10%, and the proportion of the test set is 20%-10%.
[0075] S03: Input the training set into the homogenization equivalence coefficient prediction model for training, obtain the temporary parameters of the model, and obtain the temporary homogenization equivalence coefficient prediction model based on the temporary parameters.
[0076] S04: Input the validation set into the temporary homogenized equivalent coefficient prediction model for processing to obtain the prediction results of the validation set;
[0077] S05: With the goal of minimizing the mean square error between the prediction results of the validation set and the true labels, optimize the temporary parameters to obtain the preliminary optimized parameters, and obtain the preliminary homogenized equivalent coefficient prediction model based on the preliminary optimized parameters.
[0078] S06: Using the test set, evaluate the test set prediction results output by the preliminary homogenization equivalence coefficient prediction model, obtain the optimal parameters of the model, and obtain the trained homogenization equivalence coefficient prediction model based on the optimal parameters of the model.
[0079] Furthermore, the homogenization equivalence coefficient prediction model in this embodiment includes, but is not limited to, a series of regression algorithms such as neural networks, support vector machines, random forests, and Gaussian process regression algorithms.
[0080] In this embodiment, the average Reynolds equation in step S3 is based on the homogenization method. By introducing homogenization equivalence coefficients, the influence of texture on the flow field is considered. The homogenization equivalence coefficients are used to correct the Poiseuille and Couette flow terms, thereby achieving equivalence of the textured unit for the macroscopic lubricating film pressure distribution. The textured region uses the nominal film thickness. By numerically solving the average Reynolds equation, the pressure field distribution of the bearing under different parameters can be obtained.
[0081] Specifically, the average Reynolds equation is expressed as follows:
[0082]
[0083] In the formula: x is the macroscopic coordinate of each point within the lubrication area; The gradient of x is calculated; p0(x) is the macroscopic pressure field; h0(x) is the macroscopic average film thickness; μ is the dynamic viscosity of the lubricant; u a,1 The velocity vector in the direction of bearing rotation; u a,2 This is the velocity vector in the direction perpendicular to the bearing. , , and Both represent the permeation tensor coefficients in the homogenization equivalence coefficients; and Both represent the shear flow coefficient in the homogenization equivalent coefficient.
[0084] In addition, the cross-scale analysis method for the lubrication characteristics of microtextured bearings in this embodiment can be extended to other types of lubrication systems, such as oil-lubricated sliding bearings and gas-lubricated bearings. By retraining the homogenized equivalent coefficient prediction model and correcting the fluid viscosity parameters, a universal application can be achieved under different media and loads. This method combines high accuracy, low computational cost, and transferability, providing a new cross-scale research approach for the design and lubrication performance optimization of microtextured surfaces.
[0085] In summary, this embodiment introduces a cross-scale analysis method for the lubrication characteristics of microtextured bearings. Combining homogenization and machine learning, a trained homogenization equivalence coefficient prediction model is embedded into the average Reynolds equation, forming a homogenization-machine learning coupled model (HGML-ARE model). During calculation, this model automatically calls the prediction model based on the current operating conditions to obtain the homogenization equivalence coefficients, without requiring independent calculations for each textured element. Through finite element discretization and iterative solution, the pressure distribution and load-bearing capacity of the bearing lubrication film can be quickly obtained. Compared with the results of traditional direct Reynolds equation calculations and those using only the average flow model, the HGML-ARE model is applicable to various operating conditions. By retraining the homogenization equivalence coefficient prediction model and correcting fluid parameters, it can achieve universal application across media and operating conditions. This invention reduces computation time by 50%-80% and 20%-40% respectively while maintaining the same accuracy, demonstrating its significant efficiency advantage.
[0086] Example 2
[0087] Based on the cross-scale analysis method for the lubrication characteristics of microtextured bearings introduced in Example 1, this example conducts experimental verification as follows:
[0088] An HGML-ARE model based on homogenization theory (combining homogenization and machine learning to achieve a coupled model of the analysis method in Example 1) was established to predict the pressure distribution and load-bearing capacity of bearings with microtextures. Typical operating parameters were selected to verify the model's accuracy. The bearing geometry was kept consistent with the experimental sample. Under each rotational speed condition, the oil film pressure distribution and load-bearing capacity were solved.
[0089] The numerical results were compared with the experimental data to calculate the average error of pressure distribution and bearing capacity. The stability of the model was verified by perturbation analysis: under the condition of a change of ±2% in the homogenization equivalence coefficient, the change in bearing capacity was recalculated to evaluate the sensitivity and numerical stability of the model.
[0090] Experimental verification process:
[0091] 1. Sample preparation
[0092] Stainless steel bearing rings were selected as the base material.
[0093] A regularly arranged microtexture was prepared on the inner surface of the bearing race using femtosecond laser processing technology.
[0094] The texture shape is elliptical or rectangular, with a depth of approximately 5–10 µm. The aspect ratio or length-to-width ratio is determined according to orthogonal experimental design.
[0095] The texture is distributed either circumferentially or radially, and the textured area is located within the bearing loading zone.
[0096] 2. Experimental setup and testing methods
[0097] The experiment was conducted on a motor-driven rotary test platform, with the inner ring rotated by a servo motor and the outer ring fixed.
[0098] The lubricating medium is a uniform type of lubricating oil, and the oil temperature is maintained at 25℃ under constant temperature conditions.
[0099] The thickness of the lubricating film was measured in real time using a laser displacement sensor with a sampling frequency of 1 kHz.
[0100] The experimental speeds were set to 1000 r / min, 1500 r / min and 2000 r / min, and each test was repeated three times to ensure data stability.
[0101] 3. Data Acquisition and Processing
[0102] During the steady-state operation phase, oil film thickness signals are collected, and the time-averaged film thickness is calculated.
[0103] Simultaneously, parameters such as load and friction torque are recorded to calculate the load-bearing capacity.
[0104] The experimental data were smoothed and error analyzed, and then compared with the model predictions.
[0105] 4. Result Evaluation: By comparing the experimental and computational results, it was found that the average error between the predicted and experimental results should be less than 10%. The relative error of the bearing capacity is less than 2%, indicating that the model of this invention still maintains high-precision prediction capability under high-density microtexture. Further perturbation analysis revealed that when the homogenization equivalence coefficient error is within ±2%, the change in bearing capacity is only within ±2%, indicating good model stability.
[0106] In addition, microtextures were fabricated on stainless steel bearings using femtosecond laser processing technology. A motor-driven rotary friction testing platform was employed, and the lubricating film thickness was measured using a laser displacement sensor. The results showed that the optimized sample exhibited an average increase in film thickness of 4%-7% at the same rotational speed. The average error between the experimental and theoretical calculations was approximately 10%, demonstrating good agreement after considering processing errors and vibration effects. These results validate the practical applicability of the HGML-ARE model.
[0107] Example 3
[0108] Based on the cross-scale analysis method for the lubrication characteristics of microtextured bearings introduced in Embodiment 1 of this invention, further optimization design can be performed. Using the HGML-ARE model as a foundation, a particle swarm optimization algorithm is selected to intelligently optimize the texture geometry parameters. The optimization objective is to maximize the load-bearing capacity, with constraints including the manufacturability range of the texture diameter, depth, and spacing. The bearing load-bearing capacity is calculated using COMSOL, achieving automatic optimization of the texture geometry parameters and obtaining the optimal texture design scheme. Optimization results show that the optimized bearing load-bearing capacity can be increased by 110%-140%.
[0109] Specifically, the particle swarm optimization algorithm can also be replaced by a genetic algorithm, a Bayesian optimization algorithm, or a gradient search algorithm; optimization algorithms are conventional techniques used by those skilled in the art, and will not be described in detail here.
[0110] Example 4
[0111] This invention provides a cross-scale analysis system for the lubrication characteristics of microtextured bearings, comprising:
[0112] The data acquisition module is used to acquire the structural operation data of the target bearing texture;
[0113] The equivalent coefficient prediction module is used to input the structural operation data of the target bearing texture into the trained homogenization equivalent coefficient prediction model for processing, and obtain the homogenization equivalent coefficient of the target bearing texture.
[0114] The analysis module is used to solve the average Reynolds equation using the homogenization equivalent coefficient of the target bearing texture to obtain the lubrication characteristics of the target bearing texture; the lubrication characteristics include the pressure distribution and load-bearing capacity of the lubrication film.
[0115] The specific functions of each module described above are explained in the relevant content of the method in Embodiment 1, and will not be repeated here.
[0116] Example 5
[0117] This embodiment describes an electronic device, including: a memory and a processor;
[0118] The memory is used to store programs;
[0119] The processor is used to execute the program to implement the cross-scale analysis method for the lubrication characteristics of microtextured bearings in Example 1.
[0120] This embodiment also introduces a readable storage medium storing a computer program, characterized in that, when the computer program is executed by a processor, it implements the cross-scale analysis method for the lubrication characteristics of microtextured bearings in Embodiment 1.
[0121] This embodiment also introduces a computer program product, including a computer program, characterized in that, when the computer program is executed by a processor, it implements the cross-scale analysis method for the lubrication characteristics of microtextured bearings in Embodiment 1.
[0122] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0123] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0124] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0125] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.
Claims
1. A micro-textured bearing lubrication property multiscale analysis method, characterized by, include: Obtain structural operation data of the target bearing texture; The structural operation data of the target bearing texture is input into the trained homogenization equivalence coefficient prediction model for processing to obtain the homogenization equivalence coefficient of the target bearing texture. The average Reynolds equation is solved using the homogenization equivalent coefficient of the target bearing texture to obtain the lubrication characteristics of the target bearing texture; the lubrication characteristics include the pressure distribution and load-bearing capacity of the lubrication film.
2. The micro-textured bearing lubrication property multiscale analysis method of claim 1, wherein, The structural operation data includes geometric parameters, distribution parameters, and operating condition parameters; The geometric parameters include: the shape, depth, and size of the texture; The distribution parameters include: distribution pattern, distribution density, and distribution area; The operating parameters include: rotational speed, film thickness, film thickness tilt ratio, and bearing size.
3. The micro-textured bearing lubrication property multiscale analysis method of claim 2, wherein, The training process of the homogenized equivalence coefficient prediction model includes: Construct a sample dataset of bearing textures; The sample dataset is divided into training set, validation set and test set according to a preset ratio; The training set is input into the homogenization equivalence coefficient prediction model for training, and the temporary parameters of the model are obtained. Based on the temporary parameters, the temporary homogenization equivalence coefficient prediction model is obtained. The validation set is input into the temporary homogenized equivalent coefficient prediction model for processing to obtain the prediction results of the validation set. With the goal of minimizing the mean square error between the prediction results on the validation set and the true labels, temporary parameters are optimized to obtain preliminary optimized parameters, and a preliminary homogenized equivalent coefficient prediction model is obtained based on the preliminary optimized parameters. The test set prediction results of the preliminary homogenized equivalence coefficient prediction model are used to evaluate the coefficient of determination index, obtain the optimal parameters of the model, and obtain the trained homogenized equivalence coefficient prediction model based on the optimal parameters of the model.
4. The micro-textured bearing lubrication property multiscale analysis method of claim 3, wherein, The sample dataset for constructing bearing textures includes: Obtain the structural operation dataset of the bearing texture; Utilize orthogonal experimental design structures to run parameter combinations of parameters in the dataset; The homogenization equivalence coefficients of each parameter combination are calculated based on the finite element method, and the preliminary sample dataset of bearing texture is obtained by combining the parameter combination and the homogenization equivalence coefficients of the parameter combination. The initial sample dataset of bearing texture is subjected to max-min normalization to obtain the sample dataset of bearing texture.
5. The method for cross-scale analysis of lubrication characteristics of microtextured bearings according to claim 4, characterized in that, The homogenization equivalence coefficient prediction model includes, but is not limited to: neural networks, support vector machines, random forests, and Gaussian process regression algorithms.
6. The method for cross-scale analysis of lubrication characteristics of microtextured bearings according to claim 5, characterized in that, The average Reynolds equation is expressed as follows: ； In the formula: x is the macroscopic coordinate of each point within the lubrication area; The gradient of x is calculated; p0(x) is the macroscopic pressure field; h0(x) is the macroscopic average film thickness; μ is the dynamic viscosity of the lubricant; u a,1 The velocity vector in the direction of bearing rotation; u a,2 This is the velocity vector in the direction perpendicular to the bearing. , , and Both represent the permeation tensor coefficients in the homogenization equivalence coefficients; and Both represent the shear flow coefficient in the homogenization equivalent coefficient.
7. A multi-scale analysis system for the lubrication characteristics of microtextured bearings, characterized in that, include: The data acquisition module is used to acquire the structural operation data of the target bearing texture; The equivalent coefficient prediction module is used to input the structural operation data of the target bearing texture into the trained homogenization equivalent coefficient prediction model for processing, and obtain the homogenization equivalent coefficient of the target bearing texture. The analysis module is used to solve the average Reynolds equation using the homogenization equivalent coefficient of the target bearing texture to obtain the lubrication characteristics of the target bearing texture; the lubrication characteristics include the pressure distribution and load-bearing capacity of the lubrication film.
8. An electronic device, characterized in that, include: Memory and processor; The memory is used to store programs; The processor is used to execute the program to implement the method for cross-scale analysis of the lubrication characteristics of microtextured bearings as described in any one of claims 1-6.
9. A readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for cross-scale analysis of the lubrication characteristics of microtextured bearings as described in any one of claims 1-6.
10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the method for cross-scale analysis of the lubrication characteristics of microtextured bearings as described in any one of claims 1-6.