A method for predicting microstructure oil film carrying capacity based on finite element simulation and RSM test

Abstract

The application discloses a kind of microstructure oil film carrying capacity prediction method based on finite element simulation and RSM test, belong to oil film carrying capacity prediction field;The application includes the following steps: constructing the three-dimensional model of mechanical induction microstructure, and the finite element analysis of oil flowing through microstructure under different working conditions;The relationship between the change of microstructure surface pressure and oil film carrying capacity is defined;The parameter data set of the change relationship between microstructure and oil film carrying capacity is constructed, and the data is preliminarily screened and normalized;Response surface method test design and analysis are carried out on the parameter data set to obtain a prediction model;The prediction software of microstructure oil film carrying capacity is obtained by using matlab program design.The application can predict the influence of different design parameters and the number of microstructure on oil film carrying capacity, construct the positive path of microstructure parameters and function, and significantly improve the surface performance of material.

Description

Technical Field

[0001] This invention belongs to the field of oil film load-bearing capacity prediction, and particularly relates to a method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM testing. Background Technology

[0002] With the increasing number of applications for hydraulic actuators, different environments are placing increasingly higher demands on their performance. Especially in the aerospace and military fields, problems such as oil leakage, easy wear, fatigue, and short lifespan are becoming increasingly apparent, seriously affecting the performance and reliability of hydraulic actuators. The piston rod, as the core component of the actuator, is crucial for ensuring reliable performance, safety, and long service life. Traditional methods such as laser-processed multi-hole end-face sealing and reciprocating sealing have been found to be ineffective and prone to failure under high pressure, while also offering limited lubrication and failing to address the wear problem of the piston rod against the wall surface. How to effectively improve the sealing and lubrication effects of the piston rod has become a major challenge in the design and manufacturing of high-performance piston rods.

[0003] Research has revealed that the plastic formation of continuous microstructures on the piston rod surface, when filled with oil, provides significantly better lubrication and sealing than traditional methods in various environments, thereby greatly improving its service performance. Since this method's mechanism is based on hydrodynamic pressure, the ability of the microstructured oil film to stably separate the piston rod from the wall surface in the working environment is a prerequisite for this effect. The strength of the oil film's load-bearing capacity directly affects the level of lubrication and sealing.

[0004] Currently, the design of microstructures all adopts a trial-and-error method, which involves processing microstructures of different simple shapes on the surface of the target component and then conducting experimental tests. However, this method is highly random, has huge design and manufacturing costs, and causes serious waste. Since the oil film load-bearing capacity and lubrication and sealing performance are positively correlated, it is of great significance to explore the relationship between the morphology and quantity of microstructures and the oil film load-bearing capacity for design verification. It is urgent to solve the problem of predicting the oil film load-bearing capacity of microstructures. Summary of the Invention

[0005] This invention proposes a method for predicting the load-bearing capacity of microstructure oil film based on finite element simulation and RSM testing. It addresses the problem that microstructure design cannot be characterized and its sealing and lubrication effects cannot be verified. By drawing on the research conclusion that oil film load-bearing capacity and lubrication sealing performance are positively correlated, the method improves the effectiveness of using oil film load-bearing capacity to correlate service characteristics.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM experiments is proposed. This method involves finite element analysis of pressure changes during oil flow through a "tent model" microstructure under different operating conditions. After defining the relationship between surface pressure changes and oil film load-bearing capacity, a relevant parameter dataset is constructed, and the data is initially screened and normalized. An RSM response surface methodology experiment is then used to design and analyze the parameter dataset to obtain a prediction model. Finally, software for predicting the load-bearing capacity of microstructured oil films is designed using MATLAB. The specific steps are as follows:

[0008] Step 1: Use Catia to construct a 3D model of the mechanically induced microstructure on the simplified piston rod surface. Constrain the state of the "tent model" with 5 parameters: length a, width b, height h, bottom angle alpha, and bottom deflection angle beta.

[0009] Step 2: Import the model into the finite element fluid simulation software Ansys_Fluent, assign material and fluid properties to the model and perform unstructured mesh generation. When generating the mesh, refine the mesh in the contact area to ensure the accuracy of the simulation results.

[0010] Step 3: Perform finite element fluid-structure interaction analysis to simulate the pressure change of the liquid perpendicular to the microstructure direction as the oil flows through the surface of the test model with microstructures. The five parameters mentioned in Step 1 and the number of microstructures together constitute the group of influencing factors in this simulation experiment. By changing one of the six parameters using the single variable method, the above simulation process is repeated to construct a parameter dataset of the relationship between microstructures and oil film bearing capacity.

[0011] Step 4: Perform initial screening of the data, and organize the results data into a list according to 6 factors and 9 levels. The average surface pressure p of the microstructure is used as the experimental response value. Randomly divide the data in the dataset into training set, cross-validation set and test set in a ratio of 8:1:1.

[0012] Step 5: Construct a predictive model for the bearing capacity of microstructured oil film using RSM response surface analysis. First, conduct single-factor experiments to determine the degree and relationship of influence of each input parameter on the average output pressure p. Based on the results of the single-factor experiments, use the Box-Behnken design principle, with the retention term as the factor to be examined and the average output pressure p as the index to be examined.

[0013] Step 6: Use the test set to test and verify the accuracy of the prediction model and evaluate it with the corresponding evaluation index. When the accuracy meets the standard, conduct a response surface experiment. Use Design-Expert software to fit the regression model of p and the retention parameter. Optimize the fitted formula through analysis of variance, factor interaction validation and significance analysis, and output the response surface analysis prediction model. When the accuracy does not meet the standard, repeat steps 4 and 5, re-divide the dataset and analyze it again until the accuracy meets the standard. After the accuracy meets the standard, output the RSM response surface analysis prediction model, which is the prediction model of the microstructure oil film bearing capacity.

[0014] Step 7: Develop software to predict the load-bearing capacity of microstructured oil films using the Matlab toolbox, and translate the model into a computer language to enable human-computer interaction.

[0015] In the steps described above, the key part of the preprocessing in step 1 is as follows: using a model simplification method, the surface texture of the piston rod is magnified, revealing that the location of the microstructure can be approximated as a plane. Since it is difficult to construct tent-shaped microstructures on the cylindrical wall, the piston rod and the inner wall model of the actuator cylinder are simplified into rectangular thin-walled units of 4mm*4mm*4mm. A section of substrate surface without microstructures is left at the oil inlet and outlet ends to simulate the flow of oil from a smooth area to a textured unit area in real-world conditions. To avoid generating turbulent flow at the rectangular right angles that do not exist in actual closed environments, the corners are rounded, and the corners are refined during the planar mesh generation process to improve simulation accuracy and effectiveness.

[0016] In step 2, the finite element analysis is performed: fluid-structure interaction analysis is carried out, with one end of the substrate set as the inlet and the other end as the outlet; the k-epsilon model is selected for the viscous model, and the simulation process is set to pressure-driven; when setting the boundary conditions, the outer thin wall is set as a fixed constraint, the initial inlet pressure is p0, the initial outlet pressure is p1, and the simulation residual accuracy is set to 0.0001.

[0017] In step 3, the post-processing stage outputs the average pressure value. The output pressure distribution cloud map is set, and the average surface pressure is calculated from the output pressure cloud map data file as the test result of the model.

[0018] The dataset consists of the following parameters: the description of the microstructure tent model has 5 parameters: tent model length b, tent model width a, tent model height h, tent model bottom angle alpha, and tent model deflection angle (angle with the x-axis) beta; the number of microstructure tent models is 1 parameter; the simulation of the oil has 4 parameters: fluid density, fluid viscosity, initial inlet pressure p0, and initial outlet pressure p1, for a total of 10 input parameters. By changing the 5 construction parameters and 1 number parameter of the microstructure tent model, the radial pressure values ​​perpendicular to the surface of the microstructure and at the corresponding positions after the oil flows through the microstructure are obtained, for a total of 1 output parameter.

[0019] The aforementioned response surface methodology (RSM) experimental design and analysis yielded a predictive model, which was designed and analyzed using the Design-Expert software platform. By setting independent variables, units, and their ranges (tent model length b, tent model width a, tent model height h, tent model bottom angle alpha, tent model deflection angle beta, and the number of tent models n), setting the objective function and units, an RSM response surface experiment was conducted. The Design-Expert software was used to fit a regression model of p and the retained parameters. The fitted formula was optimized and iterated through analysis of variance, factor cross-validation, and significance analysis to finally obtain the predictive model.

[0020] The aforementioned MATLAB program is used to design software for predicting the load-bearing capacity of microstructured oil films. An app interface is used to achieve good human-computer interaction. The prediction model is computerized by programming in MATLAB language. The oil film load-bearing capacity can be predicted by inputting parameters such as tent model length b, tent model width a, tent model height h, tent model bottom angle alpha, tent model deflection angle beta, and the number of tent models n.

[0021] Beneficial Effects: This invention provides a method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM testing. It breaks away from the traditional manual trial-and-error method of microstructure design, improves the efficiency of material surface performance optimization, reduces testing costs, and significantly shortens design time and iteration cycles. By using microstructures with different design parameters and quantities, it predicts the impact of microstructures on oil film load-bearing capacity, constructs a model of the relationship between microstructure parameters and function, and improves operability through software-based human-computer interaction. This method can solve similar problems related to oil film load-bearing capacity. Fields such as aerospace and military, which face harsh environments, place higher demands on different aspects of material service characteristics, and this method can be used for prediction and characterization. Furthermore, this method is not limited to predicting oil film load-bearing capacity; the same approach can be applied to the prediction and analysis of other performance or technical indicators, enabling rapid design optimization. Attached Figure Description

[0022] Figure 1 This is a three-dimensional structural diagram of the "tent model" in an embodiment of the present invention;

[0023] Figure 2 The simulation structure model is simplified in the embodiments of the present invention;

[0024] Figure 3 This is a flowchart of the prediction method in an embodiment of the present invention;

[0025] Figure 4 This is a flowchart of the RSM surface response analysis method in an embodiment of the present invention;

[0026] Figure 5 This is a functional design diagram of the MATLAB software in an embodiment of the present invention. Detailed Implementation

[0027] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments:

[0028] A method for predicting the oil film carrying capacity of microstructures based on finite element simulation and RSM testing is proposed. Finite element simulation is used to obtain pressure cloud maps of the microstructure before and after simulation under different pressures and flow rates. The relationship between the surface pressure change of the microstructure and the oil film carrying capacity is defined (i.e., the magnitude of the pressure value perpendicular to the microstructure surface in the simulation results corresponds to the level of oil film carrying capacity in the sealing gap). The average surface pressure is calculated from the simulation data file, and a parameter dataset relating the microstructure and the oil film carrying capacity is further processed. The dataset is filtered, and a regression model of p and the retained parameters is fitted using Design-Expert software. After optimization of the fitted formula through analysis of variance, factor cross-validation, and significance analysis, a prediction model is output. When the accuracy is insufficient, the dataset is re-divided and re-analyzed until the accuracy is achieved, and then the final RSM response surface analysis prediction model is output. An industrial program with good human-computer interaction is designed using MATLAB. This prediction model can predict the oil film carrying capacity of microstructures with different design parameters and numbers. Using this prediction method, a positive pathway between microstructure parameters and function is constructed, which can significantly improve the surface properties of materials.

[0029] Processing Figure 1 Taking a microstructure with a model length of 16μm, a model width of 9μm, a model height of 2μm, a base angle of 40°, and a beta angle of 20° as an example, as shown... Figure 1 As shown, the specific implementation steps are as follows (e.g.) Figure 3 As shown):

[0030] Using Catia, a microstructure with a model length of 16μm, a width of 9μm, a height of 2μm, a base angle of 40°, and a beta deflection angle of 20° was constructed. Based on this, a piston rod with a length of 254mm and a diameter of 24mm was further constructed. Continuous microstructures were then built on the curved surface of the piston rod, such as... Figure 2 As shown, the number of microstructures is 2000;

[0031] The model was imported using the finite element simulation software ANASYS. Material properties were assigned to the workpiece and fixture, and meshing was performed. During meshing, the mesh of the contact area was refined to ensure the accuracy of the simulation results.

[0032] Fluid-structure interaction analysis was performed, with one end of the substrate designated as the inlet and the other as the outlet; the RNGk-epsilon model was selected as the viscosity model.

[0033]

[0034]

[0035] Where k is the turbulent pulsating kinetic energy, ε is the dissipation rate of the turbulent pulsating kinetic energy; ρ is the fluid density; p is the fluid pressure; x i x j These are the displacement components of the fluid particles along the x and y axes, respectively; this experiment was conducted at a high Reynolds number Re, so the fluid dynamic viscosity is: Correction items G k G is the turbulent kinetic energy generated by the average velocity gradient. b Y is the turbulent kinetic energy generated by buoyancy. M The effect of wave expansion on turbulent dissipation rate; α k and α ε S are the reciprocals of the Prandtl number for turbulent kinetic energy and turbulent dissipation rate, respectively; k and S ε All are user-defined source items; C 1ε C 2ε C 3ε C μ η0 and β are constant coefficients; this model uses the modified turbulent kinetic energy transport equation and turbulent dissipation rate transport equation, where the constant coefficients are semi-empirical values. The empirical values ​​obtained from experiments have good accuracy and are not changed.

[0036] The simulation process was set to pressure-driven; when setting boundary conditions, the outer thin wall was set as a fixed constraint, the initial inlet pressure was p0, the initial outlet pressure was p1, and the simulation residual accuracy was set to 0.0001.

[0037] The relationship between the surface pressure change of the microstructure and the oil film bearing capacity is defined. The pressure value perpendicular to the surface of the microstructure in the simulation results is also the pressure acting on the piston rod surface of the actuator cylinder according to the principle of action and reaction. The magnitude of the pressure value also represents the level of the oil film bearing capacity of the sealing gap. The model length, model width, model height, model bottom angle, model deflection angle and the number of microstructures together constitute the group of influencing factors of this simulation experiment. The single variable method is used to change one of the six parameters respectively, and the above simulation process is repeated to construct the parameter dataset of the relationship between microstructure and oil film bearing capacity. The specific process is as follows: Step 1: Set the initial values ​​of the six influencing factors of the "tent model" length a, width b, height h, bottom angle alpha, bottom deflection angle beta, and number of microstructures per unit area n, respectively: length a is 16um, width b is 9um, height h is 2um, bottom angle alpha is 40°, bottom deflection angle beta is 20°, and number of microstructures per unit area n is 2000.

[0038] Step 2: Experiment again by changing the value of length 'a'. This parameter is biased within a 50% range based on the initial setting, with a total of nine levels.

[0039] Step 3: Repeat the steps in step 2 and apply the same treatment to the other five influencing factors to obtain a total of 54 sets of experimental results, thus completing the construction of the experimental dataset.

[0040] Set up an output pressure distribution cloud map, filter the data from the output pressure cloud map data file, remove the 3% of data with the highest pressure value and the 3% of data with the lowest pressure value, and list the results data according to 6 factors and 9 levels (some experimental results are shown below). The average pressure p on the microstructure surface is used as the experimental response value. Randomly divide the data in the dataset into training set and test set at a ratio of 8:1.

[0041] The specific implementation steps are as follows: number each group of experiments, select 6 groups as the test set using random numbers, and use the remaining groups as the training set (some groups in the training set are shown in Table 1);

[0042] Table 1 Training set group data

[0043] length width high included angle of the base Bottom deflection angle Number of models μm μm μm ° ° n 16 12 2 40 20 2000 22 9 2 32 20 2000 16 9 2 32 20 1500 16 6 1.5 40 15 1500 12 6 1.5 45 15 2000 16 9 1.5 40 20 2000 16 9 3 40 24 2400

[0044] Calculate the average surface pressure as the output test result of the model;

[0045] The Design-Expert software was used to fit a regression model between the output parameter p and the retained parameter. After optimization through analysis of variance (ANOVA), factor cross-validation, and significance analysis, a predictive model was output. Specifically, the optimization process involved selecting different factors and analysis modes (linear analysis, 2FI model, second-order model, third-order model, etc.) to obtain different ANOVA results. ANOVA yields two parameters: p-value and F-value, which are then used to analyze significance. A p-value < 0.05 indicates a significant difference, and a p-value < 0.01 indicates a highly significant difference. The p-value of a factor indicates its significant effect, including not only single-factor significant effects. There may also be significant factors with interactions such as AB, BC, and CD. The F-value represents the influence of each factor on the response value; it is the influence of a single factor. The larger the F-value, the greater the influence of that single factor on the response value. Note that only the "model" column is significant, while the "failure to fit" column is not significant. Furthermore, the linear terms (e.g., A, B, C, D, E, F), the two-factor interaction terms (e.g., AB, AC, BC), and the quadratic terms (e.g., A², B², C², D²) should be as highly significant as possible. This indicates good fitting accuracy, and the response surface approximation model can be used for subsequent optimization design. If the above requirements are not met, it is necessary to reselect the factors and conduct an analysis of variance. This implementation plan sets six factors for consideration, and the specific ranges and values ​​are shown in the table below.

[0046] Table 2. Factors to be considered

[0047] factor name unit lower limit upper limit A Model length μm 8 24 B Model width μm 6 12 C Model height μm 1 3 D included angle of the base ° 20 60 E Deflection angle ° 10 30 F Number of models indivual 1000 3000

[0048] The dependent variable, stress, was set in MPa. Linear ANOVA was performed using simulation data. After optimization, the model's p-value was <0.001, and the lack-of-fit term p-value = 0.0136 < 0.01, resulting in a quadratic multinomial regression model. The quadratic multinomial regression model is as follows:

[0049] P=+3.37-0.1292A-0.0583B-0.0333C-0.0708D-0.025E+0.4375F+0.0125AB-0.1375AC-0.05AD-0.0375AE+0.062 5AF+0BC-0.0375BD+0.01875BE-0.0125BF-0.1375CD-0.025CE-0.075CF-0.0875DE+0.0375DF+0.0375EF-0.2444A 2 +0.0347B 2 +0.1639C 2 +0.3556D 2-0.0153E 2 -0.1236F 2 ;

[0050] If the accuracy is not met, the dataset is re-partitioned and analyzed again until the accuracy is met, at which point the final RSM response surface analysis prediction model is output, such as... Figure 4 As shown; an industrial program with good human-computer interaction was designed using MATLAB. The program's functional requirements are as follows: Figure 5 As shown.

[0051] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM testing, characterized in that, Includes the following steps: Step 1: Use Catia to construct a 3D model of the mechanically induced microstructure on the simplified piston rod surface, and constrain the state of the 3D model through multiple parameters; Step 2: Import the model using the finite element fluid simulation software Ansys_Fluent, assign material and fluid properties to the model and perform unstructured mesh generation, refining the mesh in the contact area during mesh generation; Step 3: Perform finite element fluid-structure interaction analysis to simulate the pressure change of the liquid perpendicular to the microstructure during the process of the oil flowing through the surface of the test model with microstructure. The multiple parameters mentioned in Step 1 and the number of microstructure parameters together constitute the group of influencing factors in this simulation experiment. By changing one of the parameters in the group of influencing factors using the single variable method, the above simulation process is repeated to construct a parameter dataset of the relationship between microstructure and oil film bearing capacity. Step 4: Filter the data and organize the results into a list according to 6 factors and 9 levels. Use the average surface pressure p of the microstructure as the experimental response value. Randomly divide the data in the dataset into training set, cross-validation set and test set in a ratio of 8:1:

1. Step 5: Construct a predictive model for the carrying capacity of microstructured oil film using RSM response surface methodology. Fit a regression model of p and the retention parameter using Design-Expert software. Optimize the fitted formula through analysis of variance, factor cross-validation, and significance analysis to output the response surface analysis predictive model. The regression model is a quadratic multinomial regression model, which is as follows: P +3.37-0.1292A-0.0583B-0.0333C-0.0708D-0.025E+0.4375F+0.0125AB-0.1375AC-0.05AD-0.0375AE+0.0625AF+0BC-0.0375BD+0.01875BE-0.0125BF-0.1375CD-0.025CE-0.075CF-0.0875DE+0.0375DF+0.0375EF-0.2444A² + 0.0347B² + 0.1639C² + 0.3556D²- 0.0153E² - 0.1236F² Among them, factor A represents the model length, factor B represents the model width, factor C represents the model height, factor D represents the base angle, factor E represents the deflection angle, and factor F represents the number of models. Step 6: Use the test set to test and verify the accuracy of the prediction model and evaluate it with the corresponding evaluation index. When the accuracy meets the standard, conduct the response surface experiment. When the accuracy does not meet the standard, repeat steps 4 and 5, re-divide the dataset and analyze it again until the accuracy meets the standard. After the accuracy meets the standard, output the RSM response surface analysis prediction model, which is the prediction model of the microstructure oil film bearing capacity. Step 7: Develop software to predict the load-bearing capacity of microstructured oil films using the Matlab toolbox, and translate the model into a computer language to enable human-computer interaction.

2. The method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM testing according to claim 1, characterized in that, The three-dimensional model of the microstructure mentioned in step 1 is a tent model, and the constraint parameters are: tent model length b, tent model width a, tent model height h, tent model bottom angle alpha, and tent model deflection angle beta.

3. The method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM testing according to claim 2, characterized in that, In step 3, a finite element fluid-structure interaction analysis is performed, with one end of the substrate designated as the inlet and the other as the outlet. The RNG k-epsilon model is selected as the viscous model, and the simulation process is set to pressure-driven.

4. The method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM testing according to claim 3, characterized in that, The RNG k-epsilon model is: ， ， Where k is the turbulent pulsating kinetic energy. It is the dissipation rate of turbulent pulsating kinetic energy; p represents fluid pressure; , These are the displacement components of the fluid particles along the x and y axes, respectively; this experiment was conducted at a high Reynolds number Re, so the fluid dynamic viscosity is: ; Correction items ; The turbulent kinetic energy generated by the average velocity gradient; The turbulent kinetic energy generated by buoyancy; The effect of wave expansion on turbulent dissipation rate; and These are the reciprocals of the Prandtl number for turbulent kinetic energy and turbulent dissipation rate, respectively. and All are user-defined source items; , , , , , All are constant coefficients; this model uses the modified turbulent kinetic energy transport equation and turbulent dissipation rate transport equation, where the constant coefficients are semi-empirical values.

5. The method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM testing according to claim 2 or 3, characterized in that, Step 3, the construction of the dataset, specifically includes the following steps: Step 1: Set the initial values ​​for six influencing factors of the tent model: length a, width b, height h, bottom angle alpha, bottom deflection angle beta, and number of microstructures per unit area n. Step 2: Change the value of length 'a' and repeat the experiment. This parameter is biased within a 50% range based on the initial setting, with a total of nine levels. Step 3: Repeat the steps in Step 2, applying the same treatment to the remaining five influencing factors, resulting in a total of 54 sets of experimental results, thus completing the construction of the experimental dataset.

6. The method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM testing according to claim 5, characterized in that, In step 4, the data is filtered as follows: 6 groups are selected as the test set using random numbers, and the remaining groups are used as the training set.

7. The method for predicting the load-bearing capacity of microstructured oil films based on finite element simulation and RSM testing according to claim 1, characterized in that, The optimization process in step 5 is as follows: Based on the selection of different factors and analysis modes, different ANOVA results are obtained. In ANOVA, two parameters, p-value and F-value, are obtained, and then the significance is analyzed. When p-value < 0.05, the difference is significant, and when p-value < 0.01, the difference is extremely significant. The significant effect of a factor is judged based on its p-value. The F-value is the influence of each factor on the response value. The larger the F-value, the greater the influence of the single factor on the response value.

Images