Non-Newtonian fluid resistance loss prediction method, system, equipment and medium

Abstract

The invention discloses a non-Newtonian fluid resistance loss prediction method, system, equipment and medium, and relates to the technical field of high-speed mechanical bearing performance optimizing.The method comprises the following steps that a fluid domain model of a bearing is constructed, and mesh generation is conducted on the fluid domain model to obtain a mesh model; inputting the grid model into simulation software, setting a solver in the simulation software, selecting a corresponding flow model and a laminar flow model according to the non-Newtonian fluid, and importing a UDF function representing the physical property characteristics of the non-Newtonian fluid; and setting boundary conditions, initial conditions and iteration times, simulating the lubrication process of the non-Newtonian fluid during the operation of the bearing through the flow model and the laminar flow model, and performing iterative solution through a solver to obtain the fluid resistance loss. According to the method, the special rheological property of the lubricating grease is expressed more accurately in the form of secondary development of the UDF, and the defect that the resistance loss of the non-Newtonian fluid cannot be accurately simulated in the prior art is overcome.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of high-speed mechanical bearing performance optimization, and in particular to a non-Newtonian fluid resistance loss prediction method, system, device and medium. BACKGROUND

[0002] In the field of high-speed mechanical bearing performance optimization, reducing fluid agitation resistance loss is the key to achieving efficient lubrication. Traditional fluid resistance research mainly focuses on two aspects: one is to establish a Newtonian fluid resistance model based on the empirical test data of scholars such as Palmgren and Trippett, which can adapt to various bearing working conditions, but the empirical formula essentially limits the prediction accuracy. The second is to use the general friction torque equation provided by manufacturers such as SKF, which has a large error in prediction results and low precision.

[0003] In recent years, the breakthrough of computational fluid dynamics (CFD) technology has opened up a new way for resistance loss research. With the improvement of computing hardware performance and numerical algorithm, CFD can accurately simulate the fluid behavior in complex mechanical components. In the past decade, significant progress has been made in the study of Newtonian fluid resistance of components such as rolling element bearings (REB), but existing research has obvious limitations: first, 90% of industrial bearings use non-Newtonian fluid lubrication; second, the shear thinning characteristics of non-Newtonian fluid under high-speed high-temperature conditions make the traditional empirical formula completely invalid.

[0004] The current non-Newtonian fluid resistance loss prediction has the following technical bottlenecks: 1. Limitations of empirical formula: relying on Palmgren's empirical formula requires repeated experimental calibration, and has poor adaptability to the shear thinning characteristics of non-Newtonian fluids such as grease lubrication (error up to 30-50%), and poor applicability to different types of non-Newtonian fluids.

[0005] 2. Research gap: existing CFD research focuses on oil-lubricated bearings, and there is a lack of modeling methods for predicting the resistance loss of non-Newtonian fluid-lubricated bearings. SUMMARY

[0006] Based on the defects of the existing technology, the present application provides a non-Newtonian fluid resistance loss prediction method, system, device and medium, which solves the problem of poor adaptability of existing empirical formulas to the shear thinning characteristics of non-Newtonian fluids such as grease lubrication and the lack of modeling methods for predicting the resistance loss of non-Newtonian fluid-lubricated bearings.

[0007] The present application adopts the following technical solutions: In a first aspect, the present application provides a non-Newtonian fluid resistance loss prediction method, comprising the following steps: A fluid domain model of the bearing is constructed, and the fluid domain model is meshed to obtain a mesh model; wherein, the bearing uses non-Newtonian fluid lubrication; Input the mesh model into the simulation software, set the solver in the simulation software, select the corresponding flow model and laminar flow model according to the non-Newtonian fluid, and import the UDF function that characterizes the physical properties of the non-Newtonian fluid. By setting boundary conditions, initial conditions, and the number of iterations, the lubrication process of bearings by non-Newtonian fluids is simulated using flow models and laminar flow models. The fluid resistance loss is obtained by iteratively solving the problem using a solver.

[0008] Preferably, the UDF function is as follows: ŋ= ; ; ; In the formula, ŋ is the viscosity coefficient. For shear stress, Shear rate, k This is the consistency coefficient. n The flow index, For temperature T Dynamic viscosity at that time Viscosity at reference temperature To activate energy, This is the universal gas constant. The current absolute temperature. For reference absolute temperature, The consistency coefficient at temperature T. This is the consistency coefficient at the reference temperature.

[0009] Preferably, constructing the fluid domain model of the bearing specifically includes the following steps: Based on the structural dimensions of the bearing, geometric modeling software is used to model the bearing to obtain a geometric model; The fluid domain model is obtained by extracting the geometric model through Boolean operations.

[0010] Preferably, the solution is obtained through iterative solving using a solver, and the specific solution formula is shown below: ; ; in, For the number of periods, The torque acting on the central rolling element, The torque acting on the surface of the cage, The torque acting on the raceway surface, For fluid resistance loss, For shear stress The generated tangential force Due to pressure The generated normal force It is the lever arm vector. Let be the area of ​​a small element on the surface. This is the unit normal vector of the contact surface.

[0011] Secondly, the present invention provides a non-Newtonian fluid drag loss prediction system, comprising: A partitioning module is used to construct the fluid domain model of the bearing, and to mesh the fluid domain model to obtain a mesh model; wherein the bearing uses non-Newtonian fluid lubrication; The configuration module is used to input the mesh model into the simulation software, set the solver in the simulation software, select the corresponding flow model and laminar flow model according to the non-Newtonian fluid, and import the UDF function that characterizes the physical properties of the non-Newtonian fluid. The solver module is used to set boundary conditions, initial conditions, and the number of iterations. It simulates the lubrication process of bearings by non-Newtonian fluids using flow models and laminar flow models, and iterates through the solver to obtain the fluid resistance loss.

[0012] Thirdly, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for predicting non-Newtonian fluid resistance loss.

[0013] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for predicting non-Newtonian fluid resistance loss.

[0014] Compared with the prior art, the above-mentioned at least one technical solution adopted by the present invention can achieve the following beneficial effects: This invention first constructs a fluid domain model of the bearing, then meshes the fluid domain model to obtain a mesh model. The mesh model is input into simulation software, where a solver is set up. Based on the non-Newtonian fluid, corresponding flow and laminar flow models are selected, and UDF functions characterizing the physical properties of non-Newtonian fluids are imported. The lubrication process of the bearing by the non-Newtonian fluid is simulated using the flow and laminar flow models, and the fluid resistance loss is obtained through iterative solving by the solver. This invention employs computational fluid dynamics methods, and through secondary development of UDFs, more accurately expresses the unique rheological properties of lubricating grease. It overcomes the shortcomings of existing technologies that cannot accurately simulate the resistance loss of non-Newtonian fluids, achieving a realistic simulation of the complex flow behavior within the bearing cavity and overcoming the inaccuracy of parameter selection in empirical formulas. Attached Figure Description

[0015] 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.

[0016] Figure 1 The flowchart shows a non-Newtonian fluid drag loss prediction method based on CFD secondary development. Figure 2 This is a 3D model of the 30205 bearing. Figure 3 This is a schematic diagram of grid division; Figure 4 Distribution of fluid resistance loss of grease within the bearing cavity; Figure 5 This is a diagram showing the flow distribution of the lubricating grease. Figure 6 This is a diagram showing the shear force distribution on the roller surface. Figure 7 This is a velocity streamline diagram of the cross-section inside the bearing cavity. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 This invention proposes a method for predicting non-Newtonian fluid drag loss. This method obtains flow field characteristic information by performing a CFD simulation process (preprocessing, solving, and post-processing) on ​​a given bearing structure, and then predicts the fluid drag loss and its causes. Taking a grease-lubricated bearing as an example, the specific steps are as follows:

[0019] S1: Preprocessing.

[0020] S11: Model Building: Based on the bearing structure dimensions, use geometric modeling software (such as UG, SolidWorks, SpaceClaim, DesignModeler, etc.) to build a two-dimensional or three-dimensional geometric model.

[0021] S12: Fluid Domain Extraction / Modeling: Using modeling software, such as SpaceClaim or DesignModeler, Boolean operations are used to accurately extract the geometric model and obtain the fluid domain model.

[0022] S13: Mesh Generation and Optimization: Apply mesh generation tools (such as ANSYS Mesh, Fluent Meshing, ICEM CFD, etc.) to generate the computational domain mesh using structured or unstructured meshing methods. Use periodic boundary conditions to reduce the number of mesh elements.

[0023] S14: Mesh independence verification: By comparing the calculation results under different mesh sizes / numbers, determine the mesh configuration that meets the accuracy requirements and has the best computational efficiency.

[0024] Time step independence verification (applicable to transient calculations): By changing the time step, the calculation results are ensured to be unaffected by changes in the time step.

[0025] S2: Solve.

[0026] S21: Solver and model setup.

[0027] Choose a 2D or 3D solver (single / double precision), a pressure-based or density-based solver, and a steady-state or transient time scheme. Select a single-phase or multiphase flow model (e.g., VOF, Mixture) based on the flow physics characteristics. Select a laminar or turbulent flow model based on the Reynolds number (laminar flow is generally suitable for non-Newtonian flows). Enable the energy equations, and decide whether to enable component transport or scalar transport equations as needed.

[0028] S22: Fluid property definition: Call the built-in material database or set the non-Newtonian fluid property parameters through custom mathematical equations / UDF (user-defined function) secondary development. While inputting the grease viscosity parameters, the changes of lubricant model parameters with temperature are considered, which improves the accuracy of the model.

[0029] ŋ= (1); (2); (3); Where ŋ is the viscosity coefficient. For shear stress, Shear rate, k This is the consistency coefficient. n The flow index, For temperature T Dynamic viscosity at that time Viscosity at reference temperature To activate energy, This is the universal gas constant. The current absolute temperature. For reference absolute temperature, The consistency coefficient at temperature T. This is the consistency coefficient at the reference temperature.

[0030] The specific UDF settings are as follows: #include "udf.h" DEFINE_PROPERTY(temp_modified_hb_viscosity, cell, thread) { / / Get the physical quantities of the current unit real shear_rate = C_STRAIN_RATE_MAG(cell, thread); real temp = C_T(cell, thread); / / Temperature (K) / / Basic Herba parameters (values ​​at reference temperature T_ref=300K) real tau_0_ref = 50.0; / / Reference yield stress (Pa) real K_ref = 2.0; / / Reference consistency coefficient (Pa·s^n) real n = 0.7; / / Power-law exponent real T_ref = 300.0; / / Reference temperature (K) / / Temperature correction factor real alpha_tau = 0.02; / / Yield stress temperature coefficient (1 / K) real Ea_K = 4000.0; / / Activation energy of consistency coefficient (J / mol) real R = 8.314; / / Gas constant / / Temperature Correction Model real tau_0 = tau_0_ref * (1 - alpha_tau*(temp - T_ref)); / / Linearly corrected yield stress real K = K_ref * exp(Ea_K / R * (1 / T_ref - 1 / temp)); / / Arrhenius correction for consistency coefficient / / Herba viscosity calculation if (shear_rate > 1e-6) / / Avoid division by zero return tau_0 / shear_rate + K*pow(shear_rate, n-1); else return 1e6; / / Upper limit of viscosity at extremely low shear rates } S23: Boundary and Initialization: Set appropriate boundary conditions (such as velocity / mass flow rate / pressure inlet / outlet, wall, symmetry plane and axis, etc.) and their corresponding parameters, and initialize the flow field.

[0031] S24: Numerical scheme adjustment: Choose a pressure-velocity coupled algorithm (such as SIMPLE, SIMPLEC, PISO, Coupled) to select an appropriate spatial discretization scheme for physical quantities such as gradient, pressure, momentum, volume fraction, energy, and composition. Set a sub-relaxation factor to optimize convergence.

[0032] S25: Solution Monitoring: Set residual monitoring points and physical quantities that need to be reported (such as monitoring point data) to evaluate the calculation process.

[0033] S26: Calculation Execution: Set the number of iterations or transient calculation parameters, and start the solver to perform the calculation.

[0034] S3: Post-processing S31: Flow Field Information Extraction and Synthesis: Acquire and analyze various flow field information at a specified location or region, including (but not limited to) velocity distribution (magnitude, direction, gradient, surface velocity, update frequency, etc.), phase distribution (volume fraction, free surface location, etc.), temperature distribution, and related contour maps, vector maps, trace maps, distribution maps, histograms, animations, reports, custom functions, and scalar results.

[0035] S32: Fluid Resistance Loss Calculation (Periodic Model): The fluid resistance loss of roller bearings can be divided into two main parts: the first part is caused by the movement of the rolling elements and cage in the lubricant, which originates from the shear stress generated by the friction of the lubricant. τ and the compressive stress on the rolling elements p The direction of these stresses is opposite to the rolling direction of the rolling element, and they form a drag torque related to the bearing diameter. The second part is the churning torque, which is mainly caused by the oil film shearing effect generated when the bearing surface is immersed in oil. The surfaces involved include the ends of the rolling element, the surface of the cage, the ends of the cage, and the cage pockets, etc. This churning torque corresponds to the churning loss.

[0036] To obtain the torque (fluid resistance loss) of the cage, rolling elements, and raceways, it is necessary to integrate the shear stress and pressure on the surface, i.e.: (4); (5); in, For the number of periods, The torque acting on the central rolling element, The torque acting on the surface of the cage, The torque acting on the raceway surface, For fluid resistance loss, The tangential force is generated by shear stress. The normal force generated by pressure, It is the lever arm vector. Let be the area of ​​a small element on the surface. This is the unit normal vector of the contact surface.

[0037] S33 Visualization: The above flow field information is visualized using professional post-processing tools (such as Fluent, CFD-Post, Tecplot, Matlab, Origin, etc.).

[0038] Example 2 This embodiment takes the 30205 type tapered roller bearing as the research object and details the implementation process of the grease lubrication fluid resistance loss prediction method based on CFD secondary development. The simulation work is mainly divided into three parts: preprocessing, solving, and post-processing. See [link to documentation]. Figure 1 It should be noted that the scope of protection of this invention is not limited to the application of this specific type of bearing.

[0039] Preprocessing.

[0040] First, a 3D structural model and flow domain extraction were performed on 1 / 16 of the bearing in SolidWorks. (See...) Figure 2 The fluid domain geometry model was imported into Fluent software for mesh generation. A finer mesh was used near the inner and outer walls to ensure the narrowest gap spanned three cells. A periodic meshing method was employed to reduce the number of meshes. See the mesh generation diagram below. Figure 3 .

[0041] The initially divided meshes need to be verified for mesh independence. Several meshes of varying numbers, such as 150,000, 240,000, 350,000, and 460,000, were imported into Fluent and simulated under the same conditions. The results (fluid resistance loss and friction torque) were compared. When the number of meshes was ≥240,000, the result fluctuation was <2%. Therefore, 240,000 was selected as the optimal mesh size.

[0042] In transient flow simulations, time step independence verification is necessary to ensure the reliability of the calculation results. Based on the mesh determined by the mesh independence verification, repeated calculations are performed under different time step conditions, and the simulation results at the same flow moment are compared. By gradually decreasing the time step until the calculation results tend to stabilize and no longer change significantly, the optimal time step range is finally determined. The simulation results show that when the time step is ≤1×10⁻⁶, the optimal time step range is achieved. -5 When the step size is s, the calculation result is not affected by the step size, while also taking into account the calculation efficiency.

[0043] Please provide a solution.

[0044] The mesh obtained through the above steps was imported into Fluent for solution settings. During the numerical simulation, a three-dimensional double-precision solution mode was adopted with a 60-core parallel computing architecture. After mesh quality inspection and dimensional calibration, a pressure-based steady-state solver was selected for calculation. The multiphase flow simulation employed a VOF model combined with an implicit volume fraction algorithm, setting the gas-lipid interfacial tension coefficient to 0.02 N / m and introducing a continuous surface tension model. The energy equation was enabled. The material rheological properties were described using a user-defined function to represent the Herschel-Bulkley relationship coupled with temperature. For boundary condition settings, a velocity boundary condition (corresponding to a flow rate of 0.6 m / s at 30% lip volume) was used at the inlet, the front and rear ends were set as pressure outlets, and periodic boundary conditions were applied to both end faces. The initial flow field was set to zero gauge pressure and a lip-filled state. This configuration effectively ensured a balance between computational accuracy and efficiency. The SIMPLE pressure-velocity coupling method was selected, and the residual thresholds for continuity, velocity, and energy were set to 1 × 10⁻⁶. -6 .

[0045] In the numerical simulation, a dynamic adjustment strategy for relaxation factors was adopted, optimizing the calculation parameters within the range of 0-1 based on convergence. In the initial stage, standard relaxation factor settings were used (pressure 0.3, momentum 1, volume fraction 1, density 0.7, turbulence 0.5), maintaining default values ​​under favorable flow conditions. After determining the grease injection time parameters based on a 30% grease injection volume, a bearing motion model was established. The outer ring was set as a fixed wall, while the inner ring, cage, and rolling elements were treated using a rotating coordinate system. The surface motion characteristics of the rollers were described using a velocity synthesis method. Convergence monitoring employed the mass flow rate balance method, monitoring inlet and outlet mass flow rates and torque data in real time. Convergence was determined when both reached a stable equilibrium. Flow field initialization used a standard method, setting initial conditions from the inlet (0 Pa gauge pressure, zero velocity, pure grease phase). Transient solution parameters were set to a time step of 1×10⁻⁶. - 5 The maximum number of iterations per time step is 20. This setup ensures the stability and reliability of the computation process through reasonable parameter selection and monitoring methods.

[0046] Post-processing.

[0047] After the numerical simulation results were exported as case files and data files using Fluent software, they were visualized and analyzed using the CFD-Post post-processing tool. Four key result illustrations were generated during the processing: phase distribution characteristic map ( Figure 5 Shear force distribution diagram on roller surface () Figure 6 Velocity field distribution map ( Figure 7 ) and fluid resistance loss distribution map based on monitoring data ( Figure 4 These visualizations effectively demonstrate the flow characteristics and resistance distribution of grease within the bearing cavity, providing intuitive data support for subsequent analysis. Through a professional post-processing workflow, a systematic analysis and visualization of complex flow field data was achieved.

[0048] Based on the same concept, the present invention also provides a non-Newtonian fluid drag loss prediction system, including a partitioning module, a setting module, and a solution module.

[0049] The partitioning module is used to construct the fluid domain model of the bearing, and to mesh the fluid domain model to obtain the mesh model; the bearing uses non-Newtonian fluid lubrication.

[0050] The configuration module is used to input the mesh model into the simulation software, set the solver in the simulation software, select the corresponding flow model and laminar flow model according to the non-Newtonian fluid, and import the UDF function that characterizes the physical properties of the non-Newtonian fluid.

[0051] The solver module is used to set boundary conditions, initial conditions, and number of iterations. It simulates the bearing lubrication process of non-Newtonian fluids using flow models and laminar flow models, and iterates through the solver to obtain the fluid resistance loss.

[0052] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for predicting non-Newtonian fluid resistance loss.

[0053] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for predicting non-Newtonian fluid resistance loss.

[0054] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0055] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for predicting drag loss in non-Newtonian fluids, characterized in that, The prediction method for lubricating bearings using non-Newtonian fluids includes the following steps: A fluid domain model of the bearing is constructed, and the fluid domain model is meshed to obtain a mesh model; wherein, the bearing uses non-Newtonian fluid lubrication; Input the mesh model into the simulation software, set the solver in the simulation software, select the corresponding flow model and laminar flow model according to the non-Newtonian fluid, and import the UDF function that characterizes the physical properties of the non-Newtonian fluid. By setting boundary conditions, initial conditions, and the number of iterations, the lubrication process of non-Newtonian fluids during bearing operation is simulated using flow models and laminar flow models. The fluid resistance loss is obtained by iteratively solving the problem using a solver.

2. The method for predicting drag loss in non-Newtonian fluids as described in claim 1, characterized in that, The UDF function is as follows: ŋ= 100. ； ； In the formula, ŋ is the viscosity coefficient. For shear stress, Shear rate, k This is the consistency coefficient. n The flow index, For temperature T Dynamic viscosity at that time Viscosity at reference temperature To activate energy, This is the universal gas constant. The current absolute temperature. For reference absolute temperature, The consistency coefficient at temperature T. This is the consistency coefficient at the reference temperature.

3. The method for predicting drag loss in non-Newtonian fluids as described in claim 1, characterized in that, The construction of the fluid domain model of the bearing specifically includes the following steps: Based on the structural dimensions of the bearing, geometric modeling software is used to model the bearing to obtain a geometric model; The fluid domain model is obtained by extracting the geometric model through Boolean operations.

4. The method for predicting drag loss in non-Newtonian fluids as described in claim 1, characterized in that, The solution is obtained iteratively using a solver, and the specific solution formula is shown below: ； ； in, For the number of periods, The torque acting on the central rolling element, The torque acting on the surface of the cage, The torque acting on the raceway surface, For fluid resistance loss, For shear stress The generated tangential force Due to pressure The generated normal force It is the lever arm vector. Let be the area of ​​a small element on the surface. This is the unit normal vector of the contact surface.

5. A non-Newtonian fluid drag loss prediction system, characterized in that, include: A partitioning module is used to construct the fluid domain model of the bearing, and to mesh the fluid domain model to obtain a mesh model; wherein the bearing uses non-Newtonian fluid lubrication; The configuration module is used to input the mesh model into the simulation software, set the solver in the simulation software, select the corresponding flow model and laminar flow model according to the non-Newtonian fluid, and import the UDF function that characterizes the physical properties of the non-Newtonian fluid. The solver module is used to set boundary conditions, initial conditions, and the number of iterations. It simulates the lubrication process of bearings by non-Newtonian fluids using flow models and laminar flow models, and iterates through the solver to obtain the fluid resistance loss.

6. A computer device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the non-Newtonian fluid resistance loss prediction method according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the non-Newtonian fluid resistance loss prediction method according to any one of claims 1-4.

Images