Ball bearing interference fit dataset construction method, apparatus, electronic device, and medium

Abstract

The application discloses a ball bearing interference fit dataset construction method and device, electronic equipment and medium, and relates to the technical field of vehicle simulation. The method comprises the following steps: determining the input parameters, output parameters and parameter intervals of the input parameters of the ball bearing interference fit dataset, and constructing the size chain of the shaft and the bearing; based on a preset parameterized modeling method, establishing the parameterized geometric model of the bearing inner ring and the shaft neck of the ball bearing according to the input parameters, the output parameters and the size chain; based on the parameter intervals and a preset sampling method, performing sampling processing on the input parameters to generate multiple input parameter combinations; based on the multiple input parameter combinations, reconstructing the parameterized geometric model of the bearing inner ring and the shaft neck of the ball bearing to obtain a solving header file corresponding to each input parameter combination; calculating the solving header file corresponding to each input parameter combination respectively, and extracting the output parameters corresponding to each input parameter combination from the calculation results to generate the ball bearing interference fit dataset.

Description

Technical Field

[0001] This application relates to the field of vehicle simulation technology, and in particular to a method for constructing a ball bearing interference fit dataset, a device for constructing a ball bearing interference fit dataset, a computer-readable storage medium, and an electronic device. Background Technology

[0002] The performance analysis, simulation verification, and machine learning model training of ball bearing interference fits all rely heavily on the support of high-quality datasets. The representativeness, consistency, and accuracy of the datasets directly determine the subsequent analysis results and optimization efficiency.

[0003] Currently, the construction of such datasets relies heavily on manual operation. A typical process includes manually defining parameters, manual sampling, non-parametric geometric modeling, manually setting finite element boundary conditions and contact parameters, manually submitting calculations, and extracting results group by group. This approach results in extremely low efficiency; processing a single sample takes three to six hours, and for datasets with hundreds of samples, the total time often reaches weeks or even months, severely lagging behind subsequent analysis and training. Furthermore, the samples are of poor quality and lack representativeness. Random sampling easily leads to uneven parameter space coverage, with key parameter combinations being missed and samples being redundant, making it difficult to form a comprehensive and balanced data foundation. In addition, data consistency and accuracy are difficult to guarantee. The correlation between dimensional parameters is difficult to decouple effectively during manual processing, easily leading to logical contradictions. The high dependence of finite element modeling accuracy, boundary condition setting, and result extraction on human experience further introduces uncontrollable human error. Moreover, the overall construction cost is high, consuming significant human resources and requiring long-term use of high-end computing hardware and commercial finite element software licenses, resulting in high time and resource costs. Finally, the process is fragmented and has poor reusability. The lack of standardized connections and unified data formats in each link and the independent and scattered manual operations make it difficult to reuse the dataset or extend it to different models of ball bearings for different application scenarios. Summary of the Invention

[0004] This application aims to at least partially address one of the technical problems in related technologies. To this end, the first objective of this application is to propose a method for constructing a ball bearing interference fit dataset, comprising: determining the input parameters, output parameters, and parameter ranges of the input parameters for the ball bearing interference fit dataset, and constructing a dimension chain between the shaft and the bearing; establishing a parametric geometric model of the inner ring and journal of the ball bearing based on a preset parametric modeling method, according to the input parameters, output parameters, and dimension chain; sampling the input parameters based on the parameter range and a preset sampling method to generate multiple combinations of input parameters; reconstructing the parametric geometric model of the inner ring and journal of the ball bearing based on the multiple combinations of input parameters to obtain a solution header file corresponding to each combination of input parameters; calculating the solution header file corresponding to each combination of input parameters, and extracting the output parameters corresponding to each combination of input parameters from the calculation results to generate the ball bearing interference fit dataset.

[0005] This application seamlessly integrates parametric modeling, intelligent sampling, finite element preprocessing, batch calculation, and result extraction through a fully automated scripting process, completely solving the pain points of low efficiency, long cycle, and high cost of traditional methods. It reduces the construction time of hundreds of datasets from weeks to hours. At the same time, by constructing a size chain to decouple variable relationships and adopting Latin hypercube sampling, it ensures the consistency of data logic and the coverage balance of the sample space from the source, eliminates human error, and ensures the accuracy and representativeness of the output dataset. It can be quickly reused for different types of bearings by simply adjusting the input parameter range, realizing the high-quality, high-efficiency, and standardized construction of ball bearing interference fit datasets.

[0006] The second objective of this application is to propose a device for constructing a ball bearing interference fit dataset.

[0007] The third objective of this application is to provide a computer-readable storage medium.

[0008] The fourth objective of this application is to propose an electronic device.

[0009] To achieve the above objectives, the first aspect of this application proposes a method for constructing a ball bearing interference fit dataset. This method involves determining the input parameters, output parameters, and parameter ranges of the input parameters for the ball bearing interference fit dataset, and constructing a dimensional chain between the shaft and the bearing. Based on a preset parametric modeling method, a parametric geometric model of the inner ring and journal of the ball bearing is established according to the input parameters, output parameters, and dimensional chain. The input parameters are sampled based on the parameter range and a preset sampling method to generate multiple combinations of input parameters. Based on these multiple combinations of input parameters, the parametric geometric model of the inner ring and journal of the ball bearing is reconstructed to obtain a solution header file corresponding to each input parameter combination. The solution header file corresponding to each input parameter combination is calculated, and the output parameters corresponding to each input parameter combination are extracted from the calculation results to generate the ball bearing interference fit dataset.

[0010] According to one embodiment of this application, based on multiple input parameter combinations, the parametric geometric model of the inner ring and journal of a ball bearing is reconstructed using a preset script to obtain a solution header file corresponding to each input parameter combination. This includes: reconstructing the parametric geometric model of the inner ring and journal based on multiple input parameter combinations to generate a reconstructed geometric model corresponding to each input parameter combination; traversing the faces and edges of the reconstructed geometric model, selecting cylindrical surfaces with a diameter equal to the diameter of the journal's cylindrical surface as journal mating surfaces, selecting cylindrical surfaces with a diameter equal to the inner diameter of the bearing's inner ring as bearing inner ring mating surfaces, and selecting circular edge lines with a diameter equal to the diameter of the journal's cylindrical surface as the boundary lines at both ends of the journal; discretizing the journal mating surfaces into a set of elements and setting them as primary mating surfaces, discretizing the bearing inner ring mating surfaces into a set of elements and setting them as secondary mating surfaces, and discretizing the boundary lines at both ends of the journal into a set of nodes, and applying three-dimensional freedom constraints to this set of nodes; and generating a solution header file corresponding to each input parameter combination based on a preset contact type, primary mating surfaces, secondary mating surfaces, and the set of nodes with applied three-dimensional freedom constraints.

[0011] According to one embodiment of this application, the solver header file corresponding to each input parameter combination is calculated, and the output parameters corresponding to each input parameter combination are extracted from the calculation results. This includes: batch submitting the solver header file corresponding to each input parameter combination to the solver for calculation, generating an output data file corresponding to each input parameter combination; determining the corresponding interference fit surface node set, bearing inner ring raceway surface node set, and journal bore wall node set based on the output data file corresponding to each input parameter combination; and determining the corresponding output parameters based on the interference fit surface node set, bearing inner ring raceway surface node set, and journal bore wall node set.

[0012] According to one embodiment of this application, a parametric geometric model of the inner ring and journal of a ball bearing is established based on a preset parametric modeling method, according to input parameters, output parameters, and a dimensional chain. This includes: determining a first relationship between the journal geometry and the input parameters, and a second relationship between the inner ring and the input parameters, based on the dimensional chain of the shaft and the bearing; generating a parametric geometric model of the journal based on the journal bore diameter and the first relationship; and generating a parametric geometric model of the inner ring of the bearing based on the mating surface of the journal and the bearing and the second relationship.

[0013] According to one embodiment of this application, the input parameters include geometric parameters, interference parameters, and material parameters.

[0014] According to one embodiment of this application, the output parameters include bearing clamping force, maximum radial deformation of the bearing inner ring raceway, and maximum radial deformation of the journal bore surface.

[0015] According to one embodiment of this application, the preset sampling method includes Latin hypercube sampling or orthogonal experimental design.

[0016] To achieve the above objectives, a second aspect of this application proposes a device for constructing a ball bearing interference fit dataset, comprising: a determination module for determining the input parameters, output parameters, and parameter ranges of the input parameters for the ball bearing interference fit dataset, and constructing a dimension chain between the shaft and the bearing; a modeling module for establishing a parametric geometric model of the inner ring and journal of the ball bearing based on a preset parametric modeling method, according to the input parameters, output parameters, and dimension chain; a sampling module for sampling the input parameters based on the parameter range and a preset sampling method to generate multiple combinations of input parameters; a reconstruction module for reconstructing the parametric geometric model of the inner ring and journal of the ball bearing based on the multiple combinations of input parameters to obtain a solution header file corresponding to each combination of input parameters; and a calculation module for calculating the solution header file corresponding to each combination of input parameters, and extracting the output parameters corresponding to each combination of input parameters from the calculation results to generate a ball bearing interference fit dataset.

[0017] To achieve the above objectives, a third aspect of this application provides a computer-readable storage medium storing a ball bearing interference fit dataset construction program, which, when executed by a processor, implements the aforementioned ball bearing interference fit dataset construction method.

[0018] To achieve the above objectives, a fourth aspect of this application provides an electronic device, including a memory, a processor, and a ball bearing interference fit dataset construction program stored in the memory and capable of running on the processor. When the processor executes the ball bearing interference fit dataset construction program, it implements the aforementioned ball bearing interference fit dataset construction method.

[0019] According to the ball bearing interference fit dataset construction method, apparatus, electronic device, and medium of this application embodiment, the input parameters, output parameters, and parameter ranges of the input parameters of the ball bearing interference fit dataset are determined, and the dimension chain of the shaft and bearing is constructed; based on a preset parametric modeling method, a parametric geometric model of the inner ring and journal of the ball bearing is established according to the input parameters, output parameters, and dimension chain; the input parameters are sampled based on the parameter range and a preset sampling method to generate multiple input parameter combinations; based on the multiple input parameter combinations, the parametric geometric model of the inner ring and journal of the ball bearing is reconstructed to obtain the solution header file corresponding to each input parameter combination; the solution header file corresponding to each input parameter combination is calculated, and the output parameters corresponding to each input parameter combination are extracted from the calculation results to generate the ball bearing interference fit dataset. This application seamlessly integrates parametric modeling, intelligent sampling, finite element preprocessing, batch calculation, and result extraction through a fully automated scripting process, completely solving the pain points of low efficiency, long cycle, and high cost of traditional methods. It reduces the construction time of hundreds of datasets from weeks to hours. At the same time, by constructing a size chain to decouple variable relationships and adopting Latin hypercube sampling, it ensures the consistency of data logic and the coverage balance of the sample space from the source, eliminates human error, and ensures the accuracy and representativeness of the output dataset. It can be quickly reused for different types of bearings by simply adjusting the input parameter range, realizing the high-quality, high-efficiency, and standardized construction of ball bearing interference fit datasets. Attached Figure Description

[0020] Figure 1 This is a flowchart of a method for constructing a ball bearing interference fit dataset according to some embodiments of this application; Figure 2 This is a schematic diagram showing the dimensional annotations of the interference fit of a ball bearing according to some embodiments of this application; Figure 3 This is a schematic diagram of the dimensional chain construction of a ball bearing interference fit model according to some embodiments of this application; Figure 4 A diagram showing the results of Latin hypercube sampling according to some embodiments of this application; Figure 5 This is a schematic diagram of a ball bearing interference fit finite element model according to some embodiments of this application; Figure 6 A block diagram of a ball bearing interference fit dataset construction apparatus according to some embodiments of this application; Figure 7 This is a block diagram of an electronic device according to some embodiments of this application. Detailed Implementation

[0021] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0022] The following describes in detail, with reference to the accompanying drawings, the method, apparatus, electronic device, and medium for constructing ball bearing interference fit datasets according to embodiments of this application.

[0023] Figure 1 This is a flowchart of a method for constructing a ball bearing interference fit dataset according to some embodiments of this application. (Refer to...) Figure 1 The method for constructing a ball bearing interference fit dataset according to embodiments of this application may include the following steps: S110, determine the input parameters, output parameters, and parameter range of the input parameters for the ball bearing interference fit dataset, and construct the dimension chain of the shaft and bearing.

[0024] Specifically, before creating a ball bearing interference fit dataset, it is necessary to determine the input and output parameters of the ball bearing interference fit dataset.

[0025] To further illustrate the above embodiments, in this application embodiment, the input parameters include geometric parameters, interference parameters, and material parameters.

[0026] Specifically, refer to Figure 2 The ball bearing includes a journal 1 and a bearing 2. The geometric parameters include bearing parameters and journal parameters. The bearing parameters include the inner diameter of the bearing inner ring D2, the outer diameter of the bearing inner ring D4, the radius of the rolling element R, the rolling diameter of the rolling element D0, and the width of the inner ring W. The journal parameters include the diameter of the journal bore wall d, the diameter of the retaining ring groove D3, the journal diameter D2 (the journal diameter and the outer diameter of the bearing inner ring are mating surfaces and have the same dimensions), the shoulder diameter D1, the journal length L, and the width of the retaining ring groove d.

[0027] Interference parameters refer to the radius interference between the bearing inner ring and the journal; material parameters include the elastic modulus and Poisson's ratio of the bearing inner ring and journal materials.

[0028] To further illustrate the above embodiments, in the embodiments of this application, the output parameters include bearing clamping force, maximum radial deformation of the bearing inner ring raceway, and maximum radial deformation of the journal bore surface.

[0029] Specifically, the output parameters include the bearing clamping force F, the maximum radial deformation of the bearing inner ring raceway ZC_U1, and the maximum radial deformation of the journal bore surface Z_U1. The bearing clamping force F refers to the resultant force of the radial contact forces generated by the interference fit, and is a performance indicator measuring the bearing's axial load-carrying capacity, measured in N. The maximum radial deformation of the bearing inner ring raceway ZC_U1 is used to check whether the rolling clearance of the bearing after interference fit meets the requirements, measured in mm. The maximum radial deformation of the journal bore surface Z_U1 is used to check whether the assembly allowance of the journal bore surface after interference fit meets the requirements, measured in mm.

[0030] Based on mechanical design manuals, industry standards, and product design requirements, it is also necessary to ensure that each input parameter is within a reasonable value range, forming a sample distribution space to ensure that the sample data conforms to the actual engineering application scenario. Therefore, the parameter range of the input parameters is also required.

[0031] For example, the bearing parameters are: bearing inner ring inner diameter D2=50-60mm, bearing inner ring outer diameter D4=70-80mm, rolling element radius R=3-5mm, rolling element rolling diameter D0=60-70mm, inner ring width W=15-25mm; bearing parameters are: bearing inner ring inner diameter D2=50-60mm, bearing inner ring outer diameter D4=70-80mm, rolling element radius R=3-5mm, rolling element rolling diameter D0=60-70mm, inner ring width W=15-25mm; interference parameter: radius interference CLEARANCE=0.005-0.05mm; material parameters: the bearing inner ring and journal are both made of steel, with an elastic modulus of 210000Mpa and a Poisson's ratio of 0.3.

[0032] Furthermore, referring to Figure 3 This approach decouples the design dimensions of the shaft and bearing into intermediate variables that facilitate modeling, resolving the correlation between dimensional variables in Design of Experiments (DOE) design and providing a data logic foundation for parametric modeling. For example, if the diameter of the journal bore is defined as d, then the diameter of the journal retainer groove, D3, is equal to d + 2 * H1 (where H1 is the wall thickness of the retainer groove and the journal bore). Similarly, all design dimensions of the bearing inner ring and the journal are constructed into independent dimensional chains, ensuring data logic consistency during parameter adjustments. In this way, a dimensional chain for the shaft and bearing is established.

[0033] The parameterization definition and dimension chain construction method of this application is universal. By adjusting the input parameter type, value range and dimension chain relationship, it can adapt to the data set construction needs of ball bearings of different models and specifications, solve the problem of poor reusability of traditional datasets, and significantly improve the applicability of the technical solution.

[0034] S120, based on a preset parametric modeling method, establishes a parametric geometric model of the inner ring and journal of the ball bearing according to the input parameters, output parameters and dimensional chain.

[0035] Specifically, geometric modeling typically uses non-parametric methods, but using non-parametric methods for ensemble modeling makes geometric reconstruction difficult. Therefore, this application employs a parametric modeling method. Based on a preset parametric modeling method (3D modeling software), it establishes a parametric geometric model of the inner ring and journal of a ball bearing with interference fit according to input parameters, output parameters, and dimensional chains, and automatically assembles them, providing a geometric foundation for subsequent automated finite element modeling.

[0036] For example, firstly, several independent variables are selected from the input parameters as modeling root parameters. These root parameters are the bottom-level variables in the dimension chain that do not depend on other parameters. Then, starting from the modeling root parameters, the values ​​of each derived dimension are calculated step by step according to the geometric association rules represented by the dimension chain. During the geometric generation process, the cylindrical surface corresponding to the journal bore diameter is used as the initial geometric primitive. Based on the derived dimensions, structural features such as the journal retaining groove, shoulder, and end face are generated sequentially to form the journal geometry. Then, using the journal mating surface as the reference datum, the bearing inner ring bore surface is generated according to the relationship between the bearing inner ring inner diameter and the journal diameter in the dimension chain. Based on the derived dimensions such as the inner ring outer diameter, raceway position, and width, the structural features of the bearing inner ring are generated sequentially. During the generation process, the journal geometry and the bearing inner ring geometry automatically establish an assembly relationship through coaxial constraints, and the interference fit in the assembly relationship is directly controlled by the radius interference in the input parameters. When the input parameters change, the modeling root parameters are updated, each derived dimension in the dimension chain is recalculated synchronously, and the geometric model is automatically reconstructed while maintaining the assembly constraint relationship unchanged.

[0037] To further illustrate the above embodiments, in this application embodiment, based on a preset parametric modeling method, a parametric geometric model of the bearing inner ring and journal of a ball bearing is established according to input parameters, output parameters, and a dimensional chain. This includes: determining a first relationship between the journal geometry and the input parameters, and a second relationship between the bearing inner ring and the input parameters, based on the dimensional chain of the shaft and the bearing; generating a parametric geometric model of the journal based on the journal bore diameter and the first relationship; and generating a parametric geometric model of the bearing inner ring based on the mating surface of the journal and the bearing and the second relationship.

[0038] Specifically, before modeling, the dimensional chain between the shaft and bearing is analyzed to clarify the implicit geometric association rules. These rules include the first relationship between the journal geometry and the input parameters, and the second relationship between the bearing inner ring and the input parameters. The first relationship characterizes which input parameters determine each dimension of the journal (e.g., retainer groove diameter, shoulder diameter, journal length) and how they are calculated. The second relationship characterizes which input parameters determine each dimension of the bearing inner ring (e.g., inner diameter, outer diameter, raceway position) and how they are calculated.

[0039] Furthermore, a parametric geometric model of the journal is generated based on the first relation. Centered on the origin of the global coordinate system and using the journal bore diameter as the reference dimension, the journal bore diameter is substituted into the first relation to calculate all other dimensions of the journal step by step, thereby generating a complete journal geometric model. The shape of the entire journal is automatically determined by the first relation and input parameters, without the need for manual specification of each dimension.

[0040] Furthermore, a parametric geometric model of the bearing inner ring is generated based on the second relation and automatically assembled. Using the pre-generated journal mating surface as a positioning reference, the relevant input parameters are substituted into the second relation to calculate all dimensions of the bearing inner ring and generate a geometric model. Simultaneously, the inner bore surface of the bearing inner ring and the journal mating surface are automatically coaxially constrained to ensure correct assembly relationship, and the interference fit is directly reflected by the radius interference fit in the input parameters.

[0041] When the input parameters change, the geometric models of the journal and bearing inner ring are automatically updated based on the first and second relationships, the assembly relationship remains unchanged, and the whole process does not require remodeling or manual adjustment.

[0042] S130, the input parameters are sampled based on the parameter range and a preset sampling method to generate multiple combinations of input parameters. To further illustrate the above embodiments, in this application embodiment, the preset sampling method includes Latin hypercube sampling or orthogonal experimental design, and no specific limitation is made here.

[0043] In the following description, the Latin hypercube sampling method is used as an example, but this is not intended to limit the scope of this application.

[0044] Specifically, the Latin hypercube sampling method is used to sample the input parameters within the parameter range to achieve DOE design, generating N sets of input parameter combinations (e.g., N is 400), ensuring that the samples are evenly distributed in the parameter space and covering all key parameter combinations.

[0045] For example, the parameter range of each input parameter is divided into N equally probable intervals (e.g., the interference δ = 0-0.1 mm is divided into 400 intervals, each with a width of 0.00025 mm). One sample point is randomly selected from the N intervals of each parameter to ensure that each interval is selected only once. Then, the sample points of each input parameter are randomly combined to generate 400 independent combinations of input parameters.

[0046] Thus, compared to traditional random sampling, Latin hypercube sampling can cover a wider parameter space with the same sample size, reduce sample redundancy, and improve the representativeness of the dataset, such as... Figure 4 As shown, this is a low-dimensional visualization result of the Latin hypercube sampling sample in the application embodiment. The sample uniformly covers the parameter space. At the same time, by constructing a size chain to decouple the parameter correlation, data logic contradictions are avoided to a certain extent, which solves the problems of poor sample quality and insufficient consistency in traditional methods, and provides high-quality data support for subsequent model training or performance analysis.

[0047] S140 reconstructs the parametric geometric model of the inner ring and journal of the ball bearing based on multiple input parameter combinations to obtain the solution header file corresponding to each input parameter combination.

[0048] Specifically, based on the secondary development of finite element simulation software, automated scripts are written to automate the entire process of model reconstruction, mesh generation, feature selection, contact settings, boundary condition settings, local coordinate system establishment, solver settings, and header file export, without manual intervention. The specific implementation method involves using Python to pass instructions to the SimLab finite element analysis software via XML statements to achieve automated workflow. Figure 5 The image shows an automated finite element model of a bearing interference fit.

[0049] For example, in the above embodiment, a table (sample matrix) containing hundreds of sets of input parameters has been generated through DOE sampling, with each row of data representing a shaft and bearing design scheme. At this time, the parametric geometric model of the bearing inner ring and journal is equivalent to a general deformation template.

[0050] Next, the script begins to read each row of data in the table in a loop. When it reads the first set of parameters, the script automatically updates the parametric geometric model of the bearing inner ring and journal in the 3D software. The journal bore diameter automatically changes to the value specified by that set of parameters, and all related dimensions such as the retaining ring groove, shoulder, and bearing inner ring thickness are also updated synchronously according to the dimensional chain logic. This completes one model reconstruction, generating a 3D geometry that perfectly corresponds to that set of parameters.

[0051] Finally, the script did not directly use the 3D model for calculation. Instead, it automatically performed a series of finite element preprocessing operations, including mesh generation, material assignment, contact settings, interference fit matching, and boundary condition fixing. After processing, all settings were packaged and exported as a solver header file (i.e., a command sheet submitted to the finite element solver).

[0052] Once this set of parameters is processed, the script immediately reads the next row of data in the table and repeats the above process. In the end, it produces hundreds of solution header files that correspond one-to-one with the combination of input parameters, thus preparing all the input materials for subsequent batch calculations.

[0053] To further illustrate the above embodiments, in this application embodiment, based on multiple input parameter combinations, the parametric geometric model of the inner ring and journal of the ball bearing is reconstructed using a preset script to obtain a solution header file corresponding to each input parameter combination. This includes: reconstructing the parametric geometric model of the inner ring and journal based on multiple input parameter combinations to generate a reconstructed geometric model corresponding to each input parameter combination; traversing the faces and edges of the reconstructed geometric model, selecting cylindrical surfaces with a diameter equal to the diameter of the journal's cylindrical surface as journal mating surfaces, selecting cylindrical surfaces with a diameter equal to the inner diameter of the bearing's inner ring as bearing inner ring mating surfaces, and selecting circular edge lines with a diameter equal to the diameter of the journal's cylindrical surface as the boundary lines at both ends of the journal; discretizing the journal mating surfaces into a set of elements and setting them as primary mating surfaces, discretizing the bearing inner ring mating surfaces into a set of elements and setting them as secondary mating surfaces, and discretizing the boundary lines at both ends of the journal into a set of nodes, and applying three-dimensional freedom constraints to this set of nodes; and generating a solution header file corresponding to each input parameter combination based on a preset contact type, primary mating surfaces, secondary mating surfaces, and the set of nodes after applying three-dimensional freedom constraints.

[0054] Specifically, the script reads the sample matrix obtained by Latin hypercube sampling and uses the input parameter combinations therein to drive the deformation and update of the parametric geometric model of the bearing inner ring and journal. Each set of input parameters generates a corresponding new geometric model, and 400 sets of input parameters generate a total of 400 geometric models. The script also automatically generates finite element meshes for the reconstructed geometric models.

[0055] The script automatically identifies and filters the faces and edges needed for subsequent use on the newly generated geometric model, based on two criteria: shape (e.g., cylindrical surface, rounded edge, chamfered surface) and size (e.g., whether the diameter is equal to the diameter of the journal cylindrical surface or the inner diameter of the bearing inner ring in the current design). When filtering interference fit surfaces, cylindrical surfaces with a diameter equal to the diameter of the journal cylindrical surface are selected as journal mating surfaces, cylindrical surfaces with a diameter equal to the inner diameter of the bearing inner ring are selected as bearing inner ring mating surfaces, and rounded edge lines with a diameter equal to the diameter of the journal cylindrical surface are selected as the boundary lines at both ends of the journal. The filtered faces and edges are stored in a collection for subsequent contact settings, boundary condition settings, and result retrieval.

[0056] Using the set selected in the previous step, set the journal mating surface unit set as the main mating surface, set the bearing inner ring mating surface unit set as the secondary mating surface, set the preset contact type as surface-to-surface contact, and automatically fill in the interference parameter value determined in the sampling of the current scheme.

[0057] By using the selected set of nodes on the circular edges at both ends of the journal, three degrees of freedom constraints are applied to completely fix the model in space, preventing rigid body displacement during calculation that could lead to non-convergence.

[0058] Configure the solver parameters and specify the result variables to be output. Then package all the aforementioned settings (geometry, mesh, contact, boundary, and solver parameters) and export them as solver header files according to the sample number. Repeat this process to complete the preprocessing of all input parameter groups one by one.

[0059] This application automates the entire process of finite element model reconstruction, mesh generation, contact setting, boundary condition setting, and result extraction using Python scripts, eliminating the need for manual intervention and avoiding errors caused by manual operation to a certain extent. At the same time, the standardized data output format (CSV file) ensures data consistency, solving the problems of poor accuracy and insufficient reliability of traditional datasets, and reducing the data error rate by more than 80%.

[0060] S150 calculates the solution header file corresponding to each combination of input parameters and extracts the output parameters corresponding to each combination of input parameters from the calculation results to generate a ball bearing interference fit dataset.

[0061] Specifically, after determining the solution header file corresponding to each combination of input parameters, a Python script is used to batch submit the finite element model for calculation and automatically output the database. The model is saved to a specified folder by name. After the simulation calculation is completed, the Python script reads the Abaqus output database (.odb file), automatically extracts three types of output parameters, and finally generates a ball bearing interference fit dataset based on all combinations of input parameters and the corresponding output parameters.

[0062] To further illustrate the above embodiments, in this application embodiment, the solver header file corresponding to each input parameter combination is calculated, and the output parameters corresponding to each input parameter combination are extracted from the calculation results. This includes: batch submitting the solver header file corresponding to each input parameter combination to the solver for calculation, generating an output data file corresponding to each input parameter combination; determining the corresponding interference fit surface node set, bearing inner ring raceway surface node set, and journal bore wall node set based on the output data file corresponding to each input parameter combination; and determining the corresponding output parameters based on the interference fit surface node set, bearing inner ring raceway surface node set, and journal bore wall node set.

[0063] Specifically, the script submits hundreds of solution header files (.inp or .dat) exported from the above embodiments to the finite element solver in batches, which then performs iterative calculations one by one. After completing the calculation for each scheme, the solver outputs an output data file (such as a .odb file) containing all the physical field information (displacement, stress, contact force, etc.) for that scheme, and saves it to a specified directory according to the model name. At this point, all calculation results are still scattered across hundreds of independent binary files, and a directly usable dataset has not yet been formed.

[0064] The script traverses the folder, opening each output data file sequentially according to its sample number. Since the nodes at key locations were pre-named and saved with the result files during the reconstruction phase, the script can directly call three pre-defined node sets by name: the interference fit surface node set representing the contact surface, the bearing inner ring raceway surface node set representing the bearing raceway position, and the journal bore wall surface node set representing the journal bore wall. After loading the local cylindrical coordinate system, the script extracts key data from these three node sets: the normal contact force from the interference fit surface node set, and the radial displacement components from the latter two node sets.

[0065] Summing the normal contact forces of all nodes on the contact surface yields the final bearing clamping force for this scheme. The maximum radial displacement of all nodes on the bearing raceway surface and the journal bore wall surface is extracted to obtain the maximum radial deformation of the bearing inner ring raceway and the maximum radial deformation of the journal bore surface.

[0066] After processing each file, the script writes the three values ​​along with the corresponding input parameters into a row of a table. This process is repeated until all output data files have been processed, ultimately bringing together the key results scattered across the files into a complete data table containing hundreds of input-output mapping relationships, thus completing the automated construction of the dataset.

[0067] This application achieves unified management of batch computing tasks through automated scripts, enabling dataset construction to be completed on ordinary computers without relying on high-end hardware. Furthermore, the dataset construction process only requires the automated calling authorization of finite element software, eliminating the need for manual operation by professional personnel. This lowers the software usage threshold and cost. Compared to traditional methods, the cost of dataset construction is reduced by more than 90%, enabling small and medium-sized enterprises to easily obtain high-quality datasets.

[0068] In summary, this application integrates the entire process of parametric geometric modeling, Latin hypercube sampling, automated finite element model building, batch calculation submission, and automated result extraction into a scripted workflow. This eliminates inefficient steps such as manually defining parameters, manually modeling, setting boundary conditions group by group, and manually extracting results. The construction time for a single sample group is reduced from several hours to minutes, and the total time for building a complete dataset containing hundreds of samples is shortened from several weeks to several hours, significantly improving dataset construction efficiency. Furthermore, by constructing a dimension chain between shafts and bearings and decoupling input parameters into independent variables, the consistency of the model's geometric logic during parameter adjustment is ensured. This avoids data inconsistencies that can easily arise when manually handling dimension relationships, thus ensuring the accuracy of the data. The accuracy and consistency of the dataset are ensured. Furthermore, the Latin hypercube sampling method is used to uniformly stratify and sample the parameter space, ensuring comprehensive coverage of key parameter combinations without redundancy, significantly improving the representativeness of the dataset. Moreover, the entire construction process is script-driven and requires no manual intervention, reducing the reliance on operational experience in finite element modeling, avoiding the introduction of human error, and achieving standardization and reusability of the construction process. Only the input parameter range needs to be adjusted to adapt to the dataset construction requirements of different ball bearing models. Finally, the script automatically extracts key results directly from the output database file and summarizes them into structured tables, forming a high-quality input-output mapping dataset that can be directly used for performance analysis, simulation verification, and machine learning model training.

[0069] Corresponding to the above embodiments, this application also proposes a device for constructing a ball bearing interference fit dataset.

[0070] Reference Figure 2 The ball bearing interference fit dataset construction device 200 includes: a determination module 210, a construction module 220, a sampling module 230, a reconstruction module 240, and a calculation module 250.

[0071] The system comprises the following modules: Determination module 210 determines the input parameters, output parameters, and parameter ranges of the input parameters for the ball bearing interference fit dataset, and constructs the dimension chain between the shaft and the bearing. Establishment module 220 establishes a parametric geometric model of the ball bearing's inner ring and journal based on a preset parametric modeling method, according to the input parameters, output parameters, and dimension chain. Sampling module 230 samples the input parameters based on the parameter range and a preset sampling method to generate multiple input parameter combinations. Reconstruction module 240 reconstructs the parametric geometric model of the ball bearing's inner ring and journal based on multiple input parameter combinations to obtain a solution header file corresponding to each input parameter combination. Calculation module 250 calculates the solution header file corresponding to each input parameter combination and extracts the output parameters corresponding to each input parameter combination from the calculation results to generate the ball bearing interference fit dataset.

[0072] According to one embodiment of this application, the reconstruction module 240 is specifically used to reconstruct the parametric geometric model of the bearing inner ring and journal based on multiple input parameter combinations to generate a reconstructed geometric model corresponding to each input parameter combination; traverse the faces and edges of the reconstructed geometric model, select cylindrical surfaces with a diameter equal to the diameter of the journal cylindrical surface as journal mating surfaces, select cylindrical surfaces with a diameter equal to the inner diameter of the bearing inner ring as bearing inner ring mating surfaces, and select circular edge lines with a diameter equal to the diameter of the journal cylindrical surface as the boundary lines at both ends of the journal; discretize the journal mating surfaces into a set of elements and set them as primary mating surfaces, discretize the bearing inner ring mating surfaces into a set of elements and set them as secondary mating surfaces, and discretize the boundary lines at both ends of the journal into a set of nodes, and apply three-dimensional freedom constraints to the set of nodes; and generate a solution header file corresponding to each input parameter combination based on the preset contact type, primary mating surfaces, secondary mating surfaces, and the set of nodes after applying three-dimensional freedom constraints.

[0073] According to one embodiment of this application, the calculation module 250 is specifically used to: submit the solution header file corresponding to each input parameter combination to the solver in batches for calculation, and generate an output data file corresponding to each input parameter combination; determine the corresponding interference fit surface node set, bearing inner ring raceway surface node set, and journal bore wall node set based on the output data file corresponding to each input parameter combination; and determine the corresponding output parameters based on the interference fit surface node set, bearing inner ring raceway surface node set, and journal bore wall node set.

[0074] According to one embodiment of this application, the reconstruction module 240 is specifically used to: determine a first relationship between the journal geometry and the input parameters and a second relationship between the bearing inner ring and the input parameters based on the dimension chain of the shaft and the bearing; generate a parametric geometric model of the journal based on the journal bore diameter and the first relationship; and generate a parametric geometric model of the bearing inner ring based on the mating surface of the journal and the bearing and the second relationship.

[0075] According to one embodiment of this application, the input parameters include geometric parameters, interference parameters, and material parameters.

[0076] According to one embodiment of this application, the output parameters include bearing clamping force, maximum radial deformation of the bearing inner ring raceway, and maximum radial deformation of the journal bore surface.

[0077] According to one embodiment of this application, the preset sampling method includes Latin hypercube sampling or orthogonal experimental design.

[0078] It should be noted that the above explanation of the embodiments and beneficial effects of the ball bearing interference fit dataset construction method also applies to the ball bearing interference fit dataset construction device of this application. To avoid redundancy, it will not be elaborated in detail here.

[0079] Corresponding to the above embodiments, this application also proposes a computer-readable storage medium.

[0080] The present application provides a computer-readable storage medium storing a ball bearing interference fit dataset construction program, which, when executed by a processor, implements the aforementioned ball bearing interference fit dataset construction method.

[0081] It should be noted that the above explanation of the embodiments and beneficial effects of the method for constructing the ball bearing interference fit dataset also applies to the computer-readable storage medium of the embodiments of this application. To avoid redundancy, it will not be elaborated in detail here.

[0082] Corresponding to the above embodiments, this application also proposes an electronic device.

[0083] See Figure 3 As shown, the electronic device 300 of this application includes a memory 310, a processor 320, and a ball bearing interference fit dataset construction program stored in the memory 310 and capable of running on the processor 320. When the processor executes the ball bearing interference fit dataset construction program, it implements the aforementioned ball bearing interference fit dataset construction method.

[0084] It should be noted that the above explanation of the embodiments and beneficial effects of the method for constructing the ball bearing interference fit dataset also applies to the electronic devices in the embodiments of this application. To avoid redundancy, they will not be elaborated in detail here.

[0085] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0086] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0087] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0088] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0089] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "joining," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components, unless otherwise expressly limited. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0090] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for constructing a dataset of ball bearing interference fits, characterized in that, include: Determine the input parameters, output parameters, and parameter ranges of the ball bearing interference fit dataset, and construct the dimension chain between the shaft and the bearing; Based on a preset parametric modeling method, a parametric geometric model of the inner ring and journal of a ball bearing is established according to the input parameters, the output parameters, and the dimensional chain. The input parameters are sampled based on the parameter range and a preset sampling method to generate multiple combinations of input parameters. Based on the multiple input parameter combinations, the parametric geometric model of the inner ring and journal of the ball bearing is reconstructed to obtain the solution header file corresponding to each input parameter combination; The solver header file corresponding to each of the input parameter combinations is calculated, and the output parameters corresponding to each of the input parameter combinations are extracted from the calculation results to generate a ball bearing interference fit dataset.

2. The method for constructing a ball bearing interference fit dataset according to claim 1, characterized in that, Based on the multiple input parameter combinations, the parametric geometric model of the inner ring and journal of the ball bearing is reconstructed using a preset script to obtain the solution header file corresponding to each input parameter combination, including: Based on the multiple combinations of input parameters, the parametric geometric models of the bearing inner ring and journal are reconstructed to generate the reconstructed geometric model corresponding to each combination of input parameters. Traverse the faces and edges of the reconstructed geometric model, select cylindrical surfaces with a diameter equal to the diameter of the journal cylindrical surface as journal mating surfaces, select cylindrical surfaces with a diameter equal to the inner diameter of the bearing inner ring as bearing inner ring mating surfaces, and select circular edge lines with a diameter equal to the diameter of the journal cylindrical surface as the boundary lines at both ends of the journal. The journal mating surface is discretized into a set of elements and set as the master mating surface; the bearing inner ring mating surface is discretized into a set of elements and set as the slave mating surface; and the boundary lines at both ends of the journal are discretized into a set of nodes, and a three-dimensional degree of freedom constraint is applied to the set of nodes. Based on the preset contact type, the main mating surface, the secondary mating surface, and the node set after applying three-dimensional degree-of-freedom constraints, a solution header file corresponding to each input parameter combination is generated.

3. The method for constructing a ball bearing interference fit dataset according to claim 1, characterized in that, The step of calculating the solution header file corresponding to each of the input parameter combinations and extracting the output parameters corresponding to each of the input parameter combinations from the calculation results includes: The solution header file corresponding to each of the input parameter combinations is submitted to the solver in batches for calculation, generating the output data file corresponding to each of the input parameter combinations; Based on the output data file corresponding to each of the input parameter combinations, determine the corresponding interference fit surface node set, bearing inner ring raceway surface node set, and journal bore wall surface node set; The corresponding output parameters are determined based on the set of node points of the interference fit surface, the set of node points of the bearing inner ring raceway surface, and the set of node points of the journal bore wall surface.

4. The method for constructing a ball bearing interference fit dataset according to claim 1, characterized in that, The method based on preset parametric modeling establishes a parametric geometric model of the inner ring and journal of the ball bearing according to the input parameters, the output parameters, and the dimensional chain, including: Based on the dimensional chain of the shaft and bearing, a first relationship between the journal geometry and the input parameters and a second relationship between the bearing inner ring and the input parameters are determined; Based on the journal bore diameter and the first relationship, a parametric geometric model of the journal is generated; Based on the mating surfaces of the journal and the bearing and the second relationship, a parametric geometric model of the bearing inner ring is generated.

5. The method for constructing a ball bearing interference fit dataset according to claim 1, characterized in that, The input parameters include geometric parameters, interference parameters, and material parameters.

6. The method for constructing a ball bearing interference fit dataset according to claim 1, characterized in that, The output parameters include bearing clamping force, maximum radial deformation of the bearing inner ring raceway, and maximum radial deformation of the journal bore surface.

7. The method for constructing a ball bearing interference fit dataset according to claim 1, characterized in that, The preset sampling methods include Latin hypercube sampling or orthogonal experimental design.

8. A device for constructing a dataset of ball bearing interference fits, characterized in that, include: The determination module is used to determine the input parameters, output parameters, and parameter range of the ball bearing interference fit dataset, and to construct the dimension chain of the shaft and bearing; A module is established to create a parametric geometric model of the inner ring and journal of a ball bearing based on a preset parametric modeling method, according to the input parameters, the output parameters, and the dimensional chain. The sampling module is used to sample the input parameters based on the parameter range and a preset sampling method to generate multiple combinations of input parameters. The reconstruction module is used to reconstruct the parametric geometric model of the inner ring and journal of the ball bearing based on the multiple input parameter combinations, so as to obtain the solution header file corresponding to each input parameter combination; The calculation module is used to calculate the solution header file corresponding to each of the input parameter combinations, and extract the output parameters corresponding to each of the input parameter combinations from the calculation results to generate a ball bearing interference fit dataset.

9. A computer-readable storage medium, characterized in that, It stores a ball bearing interference fit dataset construction program, which, when executed by a processor, implements the ball bearing interference fit dataset construction method according to any one of claims 1-7.

10. An electronic device, characterized in that, The system includes a memory, a processor, and a ball bearing interference fit dataset construction program stored in the memory and capable of running on the processor. When the processor executes the ball bearing interference fit dataset construction program, it implements the ball bearing interference fit dataset construction method according to any one of claims 1-7.

Images