Method and system for testing oil pick-up efficiency of oil supply under radial ring of high-speed bearing for aviation

Abstract

The present application relates to the technical field of intelligent detection, and more particularly to a kind of aviation high-speed bearing radial ring under oil supply lubrication oil collection efficiency test method and system, method includes adjusting nozzle angle and nozzle distance to initial value respectively;Set oil collection efficiency threshold and basic parameters;After selecting target parameter oil supply pressure and oil temperature, start oil supply and timing, whether the oil collection efficiency calculated by real-time acquisition actual parameter meets the standard, if it meets the standard, switch target parameter, if it does not meet the standard, calculate the optimal oil collection efficiency in target parameter range;Multiple target data sets are input into the coupling model based on machine learning to obtain a balanced solution, and the oil collection efficiency corresponding to the balanced solution is output;Obtain the oil collection efficiency test data set of radial oil collection ring.The present application tests the oil collection efficiency of radial oil collection ring under multiple parameters, comprehensively evaluates the performance of radial oil collection ring, realizes the oil collection efficiency test of radial oil collection ring under multiple parameters, and can improve the test accuracy.

Description

Technical Field

[0001] This invention relates to the field of intelligent detection technology, and in particular to a method and system for testing the efficiency of radial ring lubrication and oil recovery in high-speed aerospace bearings. Background Technology

[0002] Aero gas turbine engines are the primary power units of aircraft, renowned as the "crown jewel of modern industry" due to their complex structure, high reliability requirements, and integration of numerous cutting-edge technologies. With the rapid development of science and technology and the continuous improvement of aviation industry processes in my country, the performance and output power of aero engines are constantly increasing, leading to a significant increase in spindle speeds, reaching ultra-high speed levels. This results in increasingly higher DN values ​​for spindle bearings, generating more heat due to friction between the various moving parts of the bearings. The high-speed, stable operation of the spindle depends on the reliable operation of the spindle pivot rolling bearings (spindle bearings), placing extremely high demands on their lubrication and cooling performance. Poor lubrication and cooling of the spindle bearings can lead to reduced bearing life, increased engine vibration, and even catastrophic damage such as bearing seizure, seriously threatening flight safety.

[0003] Under-ring lubrication consists of two main parts: oil delivery and oil collection. Depending on the oil collection method, it can be divided into two structural forms: axial oil collection and radial oil collection. With a radial oil collection structure, the radial arrangement of the nozzles relative to the rotating shaft reduces the axial space requirement of the bearing cavity. Simultaneously, since radial oil collection eliminates the need for an axially positioned oil dam structure, it helps reduce the inner diameter of the main shaft bearing and improves the reliability of the engine main shaft bearing. However, during the oil collection process of the under-ring lubrication structure, especially in the radial oil collection process, the lubricating oil flow path is complex, and the motion process is not visible. Observation and control methods are limited by space constraints and complex operating conditions, making it difficult to carry out. How to conduct multi-parameter matching design of the radial oil collection device under high-speed conditions presents significant technical challenges. Currently, experiments on radial under-ring lubrication for high-speed bearings in aero-engines are difficult to conduct.

[0004] Existing technologies include methods for testing radial oil collection rings, such as the impeller-type radial oil collection ring test method for civil aircraft engine lubrication systems disclosed in CN109506940. Although this method records the effects of different oil supply pressures, oil supply temperatures, nozzle installation angles, and test speeds on the fuel supply efficiency of aircraft engines, it suffers from low efficiency and low accuracy of optimal parameters due to the use of measuring cups to record the collected oil volume and manual adjustment of experimental parameters. Another example is the experimental rig and test method for the radial ring lubrication and oil collection characteristics of high-speed aircraft bearings disclosed in CN116539851A. This method involves photographing the oil arrows and droplets formed by the interaction between the lubricating oil jet and the radial oil collection ring, and calculating the ratio of the total return oil mass flow rate to the measured value of the oil supply system flow meter. However, it lacks information on the impact of operating parameters on oil collection efficiency, and the method of collecting the return oil through two return ports results in insufficient accuracy in the oil collection volume. Summary of the Invention

[0005] The purpose of this invention is to propose a method for testing the oil collection efficiency of the radial oil collection ring in high-speed aerospace bearings, and to comprehensively evaluate the performance of the radial oil collection ring by testing the oil collection efficiency of the radial oil collection ring under multiple parameters.

[0006] The purpose of this invention is to propose a test system for the oil collection efficiency of radial oil collection rings in high-speed aerospace bearings, which enables the testing of the oil collection efficiency of radial oil collection rings under multiple parameters and improves the test accuracy.

[0007] To achieve this objective, the present invention adopts the following technical solution:

[0008] A method for testing the oil supply and recovery efficiency of radial ring under-circuit lubrication in high-speed aerospace bearings includes the following steps:

[0009] S1: Adjust the nozzle angle θ and nozzle distance L to their initial values ​​respectively;

[0010] S2: Set the oil recovery efficiency threshold and basic parameters, including oil temperature T, oil supply pressure P, radial oil recovery ring rotation speed n range, nozzle angle θ range, and nozzle distance L range;

[0011] S3: After selecting the target parameters of oil supply pressure and oil temperature, start oil supply and start timing, and collect actual parameter data in real time until the oil supply ends;

[0012] S4: Calculate the oil recovery efficiency based on the collected data, compare it with the oil recovery efficiency threshold to determine whether it meets the standard. If it meets the standard, switch the target parameters. If it does not meet the standard, use the target parameters of oil supply pressure and oil temperature as input, and calculate the radial oil recovery ring speed, nozzle angle and nozzle distance when the oil recovery efficiency is optimal through the Gaussian process proxy model optimized by the differential evolution algorithm.

[0013] S5: Using the radial oil-collecting ring speed, nozzle angle and nozzle distance calculated in step S4, repeat step S3. Repeat steps S3 and S4 to obtain the optimal oil-collecting efficiency within the target parameter range and construct a multi-objective dataset.

[0014] S6: Input the multi-objective dataset into the machine learning-based coupled model, with the maximum oil recovery efficiency as the core objective and the minimum nozzle distance as the auxiliary objective, and calculate the equilibrium solution [T, n, P, θ, L];

[0015] S7: Repeat step S3 using the equilibrium solution to calculate the oil recovery efficiency and output the oil recovery efficiency corresponding to the equilibrium solution.

[0016] S8: Output the multi-objective dataset and the oil recovery efficiency corresponding to the equilibrium solution to obtain the oil recovery efficiency test dataset of the radial oil recovery ring.

[0017] Furthermore, in step S6, the step of calculating the equilibrium solution based on the coupled machine learning model includes:

[0018] S61: The input features of the features in the multi-target dataset are x=[T, n, P, θ, L]. The input features are standardized and the target oil recovery efficiency η is normalized to [0, 1].

[0019] S62: Using Gaussian regression model and gradient boosting tree model as base learners, output the oil recovery efficiency prediction values ​​respectively, which are combined into a meta-feature vector [η1, η2];

[0020] S63: Random forest is used as the meta-learner. The meta-feature vector is input into the meta-learner to obtain the final predicted oil recovery efficiency value η. final ;

[0021] S64: Define the core objective as maximizing the oil recovery efficiency f1=η final The auxiliary objective is to minimize the radial distance of the nozzle, f2=l L By iteratively searching using the NSGA-II algorithm, the Pareto front solution is used to select the equilibrium solution set between oil recovery efficiency and nozzle distance.

[0022] S65: Use TOPSIS to balance and filter the equilibrium solution set, calculate the relative proximity and sort them, and output the optimal parameter equilibrium solution [η]. final [n, θ, L], output result [n, θ, L].

[0023] Furthermore, the training method for the machine learning-based coupled model includes:

[0024] Collect experimental data, construct a training dataset, standardize the input features x=[T,n,P,θ,L], normalize the target oil recovery efficiency η to [0,1], and divide the training dataset into a training set and a validation set;

[0025] The input training set is used to train two base learners, Gaussian regression model and gradient boosting tree model, respectively, using five-fold cross-validation. The predicted feature matrix of the dual output is used as the input feature of the meta-learner to train the meta-learner.

[0026] Two base learners are input with the validation set, and the output meta-learning vectors are input with the meta-learners to output the predicted oil recovery efficiency. Then, the Pareto front solution is output using the NSGA-II algorithm, and the equilibrium solution is output using the TOPSIS decision. The performance of the coupled machine learning-based model after training is evaluated.

[0027] Furthermore, in step S2, the oil temperature is 80~100℃, the oil supply pressure is 0.2MPa~0.5MPa, the radial oil receiving ring speed is 15000rpm~20000rpm, the nozzle angle is 10°~15°, and the nozzle distance is 10mm~20mm.

[0028] The nozzle angle refers to the angle between the centerline of the nozzle and the axis of the radial oil-collecting ring, and the nozzle distance refers to the horizontal distance between the nozzle tip and the radial oil-collecting ring.

[0029] A system for testing the lubrication efficiency of radial ring under-circle oil supply in high-speed aircraft bearings is provided to implement the aforementioned method for testing the lubrication efficiency of radial ring under-circle oil supply in high-speed aircraft bearings.

[0030] The system includes a testing system, an experimental platform, and a lubricating oil system. The experimental platform includes a radial oil-receiving ring drive device and a nozzle drive device.

[0031] The test control system is used to control the start and stop of the experimental platform and lubricating oil system, as well as adjust the radial oil receiving ring speed, nozzle angle, nozzle distance, oil supply pressure and oil temperature;

[0032] The test and control system is also used to set the oil recovery efficiency threshold and basic parameters, and to collect actual parameter data in real time. Based on the collected data, the oil recovery efficiency is calculated. If the oil recovery efficiency meets the standard, the lubricating oil system is controlled to adjust the oil supply pressure and oil temperature. If it does not meet the standard, the oil recovery ring speed, nozzle angle, and nozzle distance calculated by the Gaussian process surrogate model optimized by the differential evolution algorithm are collected again in real time. A multi-objective dataset is constructed using the optimal oil recovery efficiency under each target parameter. A machine learning-based coupled model processes the multi-objective dataset to output an equilibrium solution, and it is determined whether the oil recovery efficiency corresponding to the equilibrium solution meets the standard. The experimental process of calculating the oil recovery efficiency and adjusting the oil supply pressure and oil temperature is repeated to finally obtain the oil recovery efficiency test dataset of the radial oil recovery ring.

[0033] Furthermore, the radial oil slugging ring drive device includes an oil slugging seat, an oil slugging simulation shaft, and a high-speed motor. The oil slugging seat is provided with an oil collection chamber, and the oil slugging simulation shaft is used to support the radial oil slugging ring.

[0034] The oil collection simulation shaft is rotatably mounted on the oil collection seat. One end of the oil collection channel of the oil collection simulation shaft is connected to the oil collection groove of the radial oil collection ring, and the other end is connected to the oil collection cavity.

[0035] The high-speed motor is driven by the simulated oil recovery shaft.

[0036] Furthermore, the oil collection simulation shaft has a support platform for a radial oil collection ring, the oil inlets of the plurality of oil collection channels are located on the support platform, and the oil outlets of the oil collection channels are located on the end face of the oil collection simulation shaft.

[0037] The oil collecting channel is inclined toward the outer peripheral wall of the oil collection simulation shaft along the direction of oil flow.

[0038] Furthermore, the simulated oil-collecting bushing is provided with a locking ring, which positions the radial oil-collecting ring against the abutment.

[0039] The end of the oil recovery simulation shaft is detachably connected to the output end of the high-speed motor.

[0040] Furthermore, the nozzle driving device includes a worm gear mechanism and a lead screw guide mechanism. The nozzle is fixed to the worm gear of the worm gear mechanism via a mounting base, and the worm gear mechanism is mounted on the slide plate of the lead screw guide mechanism.

[0041] The worm gear mechanism is used to adjust the angle of the nozzle, and the lead screw guide mechanism is used to adjust the distance of the nozzle.

[0042] Furthermore, the lubricating oil system includes an oil tank, an oil supply line, and an oil return line, wherein the oil supply line delivers the oil from the oil tank to the nozzle;

[0043] The oil outlet of the oil collecting chamber has an oil collection tank. The radial oil collection ring drive device and the nozzle drive device are located on the experimental platform. The experimental platform has an oil return port, and the oil return port has a corresponding oil return tank. The oil collection tank and the oil return tank are both connected to the oil return circuit.

[0044] The technical solution provided by this invention may include the following beneficial effects:

[0045] The method of this invention performs global optimization on the basic data to obtain the corresponding basic parameters for the radial oil recovery efficiency of the radial oil recovery ring, providing data reference for the performance and applicable operating conditions of the radial oil recovery ring. The experiment is conducted automatically, improving experimental efficiency. Simultaneously, the system of this invention enables efficient and high-precision testing of the oil supply and lubrication recovery efficiency of the radial ring in high-speed aerospace bearings. Attached Figure Description

[0046] Figure 1 This is a flowchart illustrating an embodiment of the oil recovery efficiency testing method of the present invention;

[0047] Figure 2 This is a schematic diagram of a test system for the oil supply and lubrication efficiency of radial ring under-bearings in high-speed aircraft bearings.

[0048] Figure 3 is a schematic diagram of the experimental platform;

[0049] Figure 4 is a schematic diagram of the experimental platform after the transparent cover has been removed;

[0050] Figure 5 This is a cross-sectional schematic diagram of the radial oil recovery ring drive device;

[0051] Figure 6 This is a schematic diagram showing the engagement of the oil recovery simulation shaft and the oil recovery seat;

[0052] Figure 7 This is a schematic diagram from another perspective showing the engagement of the oil recovery simulation shaft and the oil recovery seat;

[0053] Figure 8 This is a schematic diagram of the nozzle drive device;

[0054] Among them, 1-Lubricating oil system;

[0055] 11-Oil tank, 111-Level gauge, 112-Circulating pump, 113-Heater, 114-Oil tank temperature sensor, 12-Oil supply line, 121-Oil supply pump set, 122-Oil supply filter, 123-Throttle valve, 124-Flow meter, 125-Pressure sensor, 13-Return oil line, 131-Return oil temperature sensor, 132-Return oil pump set, 133-Air-cooled radiator, 134-Return oil filter, 14-Oil collection bucket, 15-Return oil tank;

[0056] 2-Experimental platform;

[0057] 21-Radial oil collection ring, 22-Oil collection simulation shaft, 221-Locking ring, 222-Oil collection channel, 223-Abutting platform, 23-Oil collection seat, 231-Oil collection chamber, 24-High-speed motor, 241-Motor mounting base, 242-Coupling, 25-Nozzle, 26-Mounting base, 27-Worm gear assembly, 271-Worm gear, 272-Worm, 28-Screw guide rail mechanism, 281-Slide plate, 282-Guide rail, 283-Motion fixing seat, 29-Experimental platform, 291-Oil return port, 292-Transparent cover, 293-Oil delivery pipe. Detailed Implementation

[0058] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that the specific embodiments described herein are for illustrative and explanatory purposes only and are not intended to limit the present invention.

[0059] An embodiment of the present invention provides a method for testing the oil recovery efficiency of radial ring under-circuit lubrication in a high-speed aerospace bearing, comprising the following steps:

[0060] S1: Adjust the nozzle angle θ and nozzle distance L to their initial values ​​respectively;

[0061] S2: Set the oil recovery efficiency threshold and basic parameters, including oil temperature T, oil supply pressure P, radial oil recovery ring rotation speed n range, nozzle angle θ range, and nozzle distance L range;

[0062] S3: After selecting the target parameters of oil supply pressure and oil temperature, start oil supply and start timing, and collect actual parameter data in real time until the oil supply ends;

[0063] S4: Calculate the oil recovery efficiency based on the collected data, compare it with the oil recovery efficiency threshold to determine whether it meets the standard. If it meets the standard, switch the target parameters. If it does not meet the standard, use the target parameters of oil supply pressure and oil temperature as input, and calculate the radial oil recovery ring speed, nozzle angle and nozzle distance when the oil recovery efficiency is optimal through the Gaussian process proxy model optimized by the differential evolution algorithm.

[0064] S5: Using the radial oil-collecting ring speed, nozzle angle and nozzle distance calculated in step S4, repeat step S3. Repeat steps S3 and S4 to obtain the optimal oil-collecting efficiency within the target parameter range and construct a multi-objective dataset.

[0065] S6: Input the multi-objective dataset into the machine learning-based coupled model, with the maximum oil recovery efficiency as the core objective and the minimum nozzle distance as the auxiliary objective, and calculate the equilibrium solution [T, n, P, θ, L];

[0066] S7: Repeat step S3 using the equilibrium solution to calculate the oil recovery efficiency and output the oil recovery efficiency corresponding to the equilibrium solution.

[0067] S8: Output the multi-objective dataset and the oil recovery efficiency corresponding to the equilibrium solution to obtain the oil recovery efficiency test dataset of the radial oil recovery ring.

[0068] This invention first addresses the single-objective problem of "performance not meeting standards" by remedial optimization using the PSO algorithm, ensuring the feasibility of other parameters under fixed target parameters of oil supply pressure and oil temperature. After obtaining performance-compliant data within the target parameter range, NSGA-II is used to find Pareto front solutions under multi-objective conditions, providing multiple high-quality candidates. Finally, TOPSIS is used to select the equilibrium solution [T, n, P, θ, L] from the candidate solutions. The method of this invention is used for global optimization of the basic data, ultimately obtaining the corresponding basic parameters for the radial oil recovery ring under the target oil recovery efficiency. This provides data reference for the performance and applicable operating conditions of the radial oil recovery ring, and the experiment is conducted using an automated testing method to improve experimental efficiency.

[0069] As an example, in step S4 of this invention, the method for calculating the radial oil collection ring speed, nozzle angle, and nozzle distance when the oil collection efficiency is optimal using the target parameters of oil supply pressure and oil temperature as inputs and the Gaussian process surrogate model optimized by the differential evolution algorithm is as follows:

[0070] The oil supply pressure and oil temperature are used as fixed input parameters, and the range of radial oil receiving ring speed n, nozzle angle θ, and nozzle distance L are used as the value constraints, which serve as the optimization boundaries for the differential evolution (DE) algorithm.

[0071] Step 1: Using the DE algorithm, within the set value constraints, generate an initial candidate solution population or a new generation of candidate solutions population during the iteration process. Each candidate solution in the candidate solution population is a combination of three variables: radial oil receiving ring speed, nozzle angle, and nozzle distance.

[0072] Step 2: Combine the ternary variables corresponding to each candidate solution with the binary fixed parameters of oil supply pressure and oil temperature in a preset order to form a complete input vector. Then, standardize the complete input vector to eliminate dimensional differences. Input the standardized complete input vector into the GP surrogate model. Through the nonlinear mapping calculation of the GP surrogate model, the corresponding oil recovery efficiency prediction value is output.

[0073] Step 3: The DE algorithm compares the predicted oil recovery efficiency of all candidate solutions, and completes the population iteration update according to the logic of mutation, crossover and selection. It retains the candidate solutions with better predicted oil recovery efficiency and eliminates the candidate solutions with poor predicted oil recovery efficiency.

[0074] Step 4: Repeat steps 2 to 3 until the preset iteration termination condition is met; output the optimal candidate solution. The radial oil collection ring speed, nozzle angle, and nozzle distance corresponding to the best candidate solution are the target parameters when the oil collection efficiency is optimal.

[0075] The Gaussian process surrogate model optimized by the differential evolution algorithm in step S4 of this invention is a trained model. The actual oil collection efficiency obtained in step S3, based on the radial oil collection ring speed, nozzle angle, and nozzle distance calculated in step S4, is compared with the predicted oil collection efficiency. If the deviation between the actual oil collection efficiency and the oil collection efficiency predicted by the GP surrogate model exceeds a preset range, real oil collection data is added to retrain the GP surrogate model, and steps 2 to 3 are repeated to complete the optimization and correction.

[0076] In step S6 of this embodiment of the invention, the step of calculating the equilibrium solution based on the coupled model of machine learning includes:

[0077] S61: The input features of the features in the multi-target dataset are x=[T, n, P, θ, L]. The input features are standardized and the target oil recovery efficiency η is normalized to [0, 1].

[0078] S62: Using Gaussian Regression (GPR) and Gradient Boosting Tree (CatBoost) models as base learners, the oil recovery efficiency predictions are output respectively and combined into a meta-feature vector [η1, η2].

[0079] S63: Random Forest (RF) is used as the meta-learner. The meta-feature vector is input into the meta-learner to obtain the final predicted oil recovery efficiency value η. final ;

[0080] S64: Define the core objective as maximizing the oil recovery efficiency f1=η final The auxiliary objective is to minimize the radial distance of the nozzle, f2=l L By iteratively searching using the NSGA-II algorithm, the Pareto front solution is used to select the equilibrium solution set between oil recovery efficiency and nozzle distance.

[0081] S65: Use TOPSIS to balance and filter the equilibrium solution set, calculate the relative proximity and sort them, and output the optimal parameter equilibrium solution [η]. final [n, θ, L], output result [n, θ, L].

[0082] The machine learning-based coupled model in this scheme is the Stacking surrogate model, which forms a closed loop through NSGA-II multi-objective optimization and TOPSIS decision screening. It ensures prediction accuracy through complementary ensemble model and achieves multi-objective balance through efficient optimization algorithm and decision method.

[0083] The training method for the machine learning-based coupled model includes:

[0084] Collect experimental data, construct a training dataset, standardize the input features x=[T,n,P,θ,L], normalize the target oil recovery efficiency η to [0,1], and divide the training dataset into a training set and a validation set;

[0085] The input training set is used to train two base learners, Gaussian regression model and gradient boosting tree model, respectively, using five-fold cross-validation. The predicted feature matrix of the dual output is used as the input feature of the meta-learner to train the meta-learner.

[0086] Two base learners are input with the validation set, and the output meta-learning vectors are input with the meta-learners to output the predicted oil recovery efficiency. Then, the Pareto front solution is output using the NSGA-II algorithm, and the equilibrium solution is output using the TOPSIS decision. The performance of the coupled machine learning-based model after training is evaluated.

[0087] Understandably, the NSGA-II algorithm and TOPSIS decision-making are based on the data processing results of two trained base learners and one meta learner, and perform optimization and decision-making accordingly.

[0088] As an example, the training method for a machine learning-based coupled model is as follows:

[0089] ① Preprocess the data in the training dataset to construct the input features x=[T, n, P, θ, L] (a total of 5 feature dimensions), corresponding to the target oil recovery efficiency Y=[η] (single target). Perform Z-score standardization on the input features and normalize the target oil recovery efficiency to [0, 1], and output the standardized training set;

[0090] ② The standardized training set is randomly divided into 5 mutually exclusive subsets. One subset is selected as the validation fold and the remaining 4 subsets are used as the training fold to train the GPR model. The prediction results of all validation sets are concatenated using RBF as the kernel function to obtain the GPR single output prediction vector.

[0091] ③ Referring to step ②, select one subset as the validation fold and the remaining four subsets as the training fold to train the CatBoost model. The core hyperparameters are learning rate α=0.05, tree depth=6, number of iterations T=200, and loss function is MAE. Concatenate the prediction results of all validation sets to obtain the CatBoost single output prediction vector.

[0092] ④ Construct the meta-learner input matrix using the single-output predicted value vectors of GPR and CatBoost; train the Random Forest (RF) meta-learner using the standardized target oil recovery efficiency η as the label; the number of decision trees is 100, the tree depth is None, the minimum number of samples for node splitting is 2, and bootstrap sampling is True. RF learns the optimal weighted fusion rule of GPR and CatBoost predicted values ​​by integrating multiple decision trees, and finally obtains the trained Stacking ensemble model (GPR+CatBoost+RF);

[0093] ⑤ Input the preprocessed test set into the trained GPR and CatBoost to obtain the dual-output prediction feature matrix, and then input it into the Random Forest (RF) meta-learner to obtain the final efficiency prediction value. Validate the prediction accuracy of the entire Stacking model through R² and RMSE metrics, satisfying R²≥0.8.

[0094] As an example, in step S2, the oil temperature is 80~100℃, the oil supply pressure is 0.2MPa~0.5MPa, the radial oil receiving ring speed is 15000rpm~20000rpm, the nozzle angle is 10°~15°, and the nozzle distance is 10mm~20mm.

[0095] The nozzle angle refers to the angle between the centerline of the nozzle and the axis of the radial oil-collecting ring, and the nozzle distance refers to the horizontal distance between the nozzle tip and the radial oil-collecting ring.

[0096] When testing the oil collection efficiency of the radial oil collection ring in this invention, the oil temperature range and oil supply pressure range are preset. The resulting multi-objective dataset contains the radial oil collection ring rotation speed n, nozzle angle θ, and nozzle distance L at each temperature and pressure within the oil temperature range and oil supply pressure range, which provide a reference for the operating conditions of the radial oil collection ring. The equilibrium solution [T, n, P, θ, L] under the optimal oil collection efficiency provides the best operating conditions for the radial oil collection ring.

[0097] Accordingly, the present invention also provides a test system for the oil supply and lubrication efficiency of radial ring under high-speed aviation bearings, used to implement the above-mentioned test method for the oil supply and lubrication efficiency of radial ring under high-speed aviation bearings.

[0098] The system includes a testing system, an experimental platform 2, and a lubricating oil system 1. The experimental platform 2 includes a radial oil-collecting ring drive device and a nozzle drive device.

[0099] The test control system is used to control the start and stop of the experimental platform 2 and the lubricating oil system 1, as well as to adjust the rotation speed of the radial oil receiving ring 21, the angle of the nozzle 25, the distance of the nozzle 25, the oil supply pressure and the oil temperature.

[0100] The test and control system is also used to set the oil recovery efficiency threshold and set basic parameters, and to collect actual parameter data in real time; calculate the oil recovery efficiency based on the collected data; if the oil recovery efficiency meets the standard, control the lubricating oil system 1 to adjust the oil supply pressure and oil temperature; if it does not meet the standard, calculate the oil recovery ring speed, nozzle 25 angle and nozzle 25 distance obtained by the Gaussian process surrogate model optimized by the differential evolution algorithm, and collect actual parameter data in real time again; construct a multi-objective dataset with the optimal oil recovery efficiency under each target parameter; process the multi-objective dataset with a machine learning-based coupled model to output the equilibrium solution; determine whether the oil recovery efficiency corresponding to the equilibrium solution meets the standard; repeat the experimental process of calculating the oil recovery efficiency and adjusting the oil supply pressure and oil temperature, and finally obtain the oil recovery efficiency test dataset of the radial oil recovery ring 21.

[0101] The system of this invention enables efficient and high-precision testing of the oil supply and lubrication efficiency of radial ring under-circuit lubrication in high-speed aerospace bearings.

[0102] Preferably, the radial oil collection ring drive device includes an oil collection seat 23, an oil collection simulation shaft 22, and a high-speed motor 24. The oil collection seat 23 is provided with an oil collection cavity 231, and the oil collection simulation shaft is used to support the radial oil collection ring 21.

[0103] The oil collection simulation shaft 22 is rotatably mounted on the oil collection seat 23. One end of the oil collection channel 222 of the oil collection simulation shaft 22 is connected to the oil collection groove of the radial oil collection ring 21, and the other end is connected to the oil collection cavity 231.

[0104] The high-speed motor 24 is driven by the oil recovery simulation shaft 22.

[0105] This design incorporates an axial oblique hole (oil collection channel 222) inside the simulated oil collection shaft 22. This simulates the trajectory of oil flowing into the bearing under centrifugal force from the radial oil collection ring 21, resulting in a more accurate measurement of the collected oil volume compared to actual operating conditions. The lubricating oil travels from the radial oil collection ring 21 to the oil collection channel 222 of the simulated oil collection shaft 22, and then to the oil collection cavity within the oil collection seat 23. This closed and concise oil collection path allows the oil to flow within the equipment, reducing the impact of oil splashing and improving the accuracy of the collected oil volume measurement. As an example, the oil collection seat 23 is a hollow seat, with its internal cavity forming the oil collection cavity 231.

[0106] As an example, the high-speed motor 24 is positioned on the experimental platform 29 via the motor mounting bracket 241, and the nozzle 25 drive device is positioned on the experimental platform 29 via the motion mounting bracket 283. The motor mounting bracket 241 and the motion mounting bracket 283 are perpendicular, meaning that the oil-receiving simulation shaft 22 is perpendicular to the displacement direction of the nozzle 25. The speed of the high-speed motor 24 can be controlled by a frequency converter, and the speed of the high-speed motor 24 meets the requirements of 20,000 rpm to 32,000 rpm.

[0107] Specifically, the oil collection simulation shaft 22 has a support platform 223 for the radial oil collection ring 21, the oil inlets of the plurality of oil collection channels 222 are located on the support platform 223, and the oil outlets of the oil collection channels 231 are located on the end face of the oil collection simulation shaft 22.

[0108] The oil collecting channel 222 is inclined toward the outer peripheral wall of the oil collection simulation shaft 22 along the oil flow direction.

[0109] As an example, the oil collection simulation shaft 22 has six oil collection channels 222 evenly distributed around its circumference on the abutment platform. Each oil collection channel 222 has a diameter of 2mm, and the channels form a 10° angle with the axial direction of the oil collection simulation shaft 22. These six channels collect lubricating oil into the oil collection cavity 231 of the oil collection seat 23. The inner diameter of the installed radial oil collection ring 21 fits onto the oil collection simulation shaft 22, with the oil collection end face of the radial oil collection ring 21 tightly against the abutment platform.

[0110] The radial oil collecting ring 21 is installed in close contact with one end of the inclined oil collecting channel 222. Utilizing the centrifugal force generated when the radial oil collecting ring 21 rotates along with the simulated oil collecting shaft 22, the oil collected by the radial oil collecting ring 21 flows into the oil collecting seat 23 through the axial inclined hole, thereby enabling the measurement of the collected oil volume and the calculation of the oil collecting efficiency of the radial oil collecting ring 21. This provides a data foundation for the analysis of the oil collecting efficiency of the radial ring under-lubrication system in high-speed aerospace bearings and for structural optimization.

[0111] Preferably, the oil collection simulation shaft 22 is fitted with a locking ring 221, which positions the radial oil collection ring 21 against the abutment 223;

[0112] The end of the oil-collecting simulation shaft 22 is detachably connected to the output end of the high-speed motor 24.

[0113] The locking ring 221 is bolted to the oil-collecting simulation shaft 22, allowing for quick-release replacement of the oil-collecting simulation shaft 22 and the radial oil-collecting ring 21. Based on the detachable connection between the high-speed motor 24 and the oil-collecting simulation shaft 22, the oil-collecting simulation shaft 22 can be quickly assembled and disassembled, enabling different simulation shaft diameters to accommodate radial oil-collecting rings 21 with different inner diameters. As an example, the output shaft of the high-speed motor 24 is threadedly connected to a coupling 242, the other end of which is bolted to the oil-collecting simulation shaft 22.

[0114] In order to achieve the adjustment of the nozzle 25 angle and distance, the nozzle 25 driving device in this embodiment of the invention includes a worm gear mechanism 27 and a lead screw guide mechanism 28. The nozzle 25 is fixed to the worm gear 271 of the worm gear mechanism 27 by a mounting base 26. The worm gear mechanism is mounted on the slide plate 281 of the lead screw guide mechanism 28.

[0115] The worm gear mechanism is used to adjust the angle of the nozzle 25, and the lead screw guide mechanism 28 is used to adjust the distance of the nozzle 25.

[0116] As an example, the worm 272 of the worm gear mechanism can be rotated manually or by a motor, preferably by a motor; the lead screw of the lead screw guide mechanism 28 can be rotated manually or by a motor, preferably by a motor. Through the motor drive of the worm 272 and the lead screw, the testing system automatically adjusts the distance and angle of the nozzle 25. The guide rail 282 of the lead screw guide mechanism 28 is mounted on the motion fixed base 283, and the slide plate 281 slides with the guide rail and is connected to the lead screw nut. When the lead screw rotates, the lead screw nut drives the slide plate 281 to move along the guide rail, thereby adjusting the position of the nozzle 25.

[0117] As an example, an angle sensor is provided at the connection between the nozzle 25 and the mounting base 26, and a displacement sensor is provided on the lead screw guide mechanism 28. The angle sensor is used to measure the angle of the nozzle 25, and the displacement sensor is used to detect the distance the nozzle 25 moves. The angle data and displacement data are sent to the test system for data processing and control of the distance and angle of the nozzle 25.

[0118] The lubricating oil system 1 in this embodiment of the invention includes an oil tank 11, an oil supply line 12 and an oil return line 13, wherein the oil supply line 12 delivers the oil from the oil tank to the nozzle 25.

[0119] The oil outlet of the oil collecting chamber has an oil collection tank 14. The radial oil collecting ring 21 driving device and the nozzle 25 driving device are located on the experimental platform 29. The experimental platform 29 has an oil return port 291, and the oil return port 291 corresponds to an oil return tank 15. Both the oil collection tank 14 and the oil return tank 15 are connected to the oil return passage 13. Correspondingly, the experimental platform 29 has a transparent cover 292, and the radial oil collecting ring 21 driving device and the nozzle 25 driving device are located inside the cover.

[0120] The oil collection tank 14 is used to hold the oil transported by the radial oil collection ring 21, and the return oil tank 15 is used to hold the oil splashed out on the experimental platform 29. The two types of oil are strictly isolated to improve the accuracy of oil collection efficiency calculation. Specifically, there is an oil delivery pipe 293 between the bottom of the oil collection chamber and the return oil collection tank.

[0121] As an example, the amount of oil in the oil collection tank 14 is automatically acquired and sent to the testing system by taking a picture or weighing, preferably by taking a picture to read the number of milliliters of oil.

[0122] As an example, the oil tank is equipped with a level gauge 111 and an oil tank temperature sensor 114 to obtain the oil volume and temperature in real time. A circulation pump 112 and a heater 113 are connected to the side of the oil tank to heat the lubricating oil, maintaining its temperature between room temperature and 130°C. The testing system controls the start / stop of the circulation pump 112 and the heater 113, as well as their heating power, based on the oil temperature. The oil supply circuit 12 is equipped with an oil supply pump assembly 121, an oil supply filter 122, a throttle valve 123, a flow meter 124, and a pressure sensor 125. The testing system combines the numerical monitoring from the flow meter 124 and the pressure sensor 125 to regulate the oil supply pressure. The oil return circuit 13 is equipped with a return oil temperature sensor 131, a return oil pump assembly 132, an air-cooled radiator 133, and a return oil filter 134.

[0123] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0124] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0125] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory / random access memory, magnetic disk, optical disk) and includes several instructions to cause a multimedia terminal device (which may be a mobile phone, computer, television receiver, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0126] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for testing the oil supply and recovery efficiency of radial ring under-circuit lubrication in high-speed aerospace bearings, characterized in that... Includes the following steps: S1: Adjust the nozzle angle θ and nozzle distance L to their initial values ​​respectively; S2: Set the oil recovery efficiency threshold and basic parameters, including oil temperature T, oil supply pressure P, radial oil recovery ring rotation speed n range, nozzle angle θ range, and nozzle distance L range; S3: After selecting the target parameters of oil supply pressure and oil temperature, start oil supply and start timing, and collect actual parameter data in real time until the oil supply ends; S4: Calculate the oil recovery efficiency based on the collected data, compare it with the oil recovery efficiency threshold to determine whether it meets the standard. If it meets the standard, switch the target parameters. If it does not meet the standard, use the target parameters of oil supply pressure and oil temperature as input, and calculate the radial oil recovery ring speed, nozzle angle and nozzle distance when the oil recovery efficiency is optimal through the Gaussian process proxy model optimized by the differential evolution algorithm. S5: Using the radial oil-collecting ring speed, nozzle angle and nozzle distance calculated in step S4, repeat step S3. Repeat steps S3 and S4 to obtain the optimal oil-collecting efficiency within the target parameter range and construct a multi-objective dataset. S6: Input the multi-objective dataset into the machine learning-based coupled model, with the maximum oil recovery efficiency as the core objective and the minimum nozzle distance as the auxiliary objective, and calculate the equilibrium solution [T, n, P, θ, L]; S7: Repeat step S3 using the equilibrium solution to calculate the oil recovery efficiency and output the oil recovery efficiency corresponding to the equilibrium solution. S8: Output the oil recovery efficiency corresponding to the multi-objective dataset and equilibrium solution to obtain the oil recovery efficiency test dataset of the radial oil recovery ring.

2. The method according to claim 1, characterized in that, In step S6, the step of calculating the equilibrium solution based on the coupled machine learning model includes: S61: The input features of the features in the multi-target dataset are x=[T, n, P, θ, L]. The input features are standardized and the target oil recovery efficiency η is normalized to [0, 1]. S62: Using Gaussian regression model and gradient boosting tree model as base learners, output the oil recovery efficiency prediction values ​​respectively, which are combined into a meta-feature vector [η1, η2]; S63: Random forest is used as the meta-learner. The meta-feature vector is input into the meta-learner to obtain the final predicted oil recovery efficiency value η. final ; S64: Define the core objective as maximizing the oil recovery efficiency f1=η final The auxiliary objective is to minimize the radial distance of the nozzle, f2=l L By iteratively searching using the NSGA-II algorithm, the Pareto front solution is used to select the equilibrium solution set between oil recovery efficiency and nozzle distance. S65: Use TOPSIS to balance and filter the equilibrium solution set, calculate the relative proximity and sort them, and output the optimal parameter equilibrium solution [η]. final [n, θ, L], output result [n, θ, L].

3. The method according to claim 2, characterized in that, The training method for the machine learning-based coupled model includes: Collect experimental data, construct a training dataset, standardize the input features x=[T,n,P,θ,L], normalize the target oil recovery efficiency η to [0,1], and divide the training dataset into a training set and a validation set; The input training set is used to train two base learners, Gaussian regression model and gradient boosting tree model, respectively, using five-fold cross-validation. The predicted feature matrix of the dual output is used as the input feature of the meta-learner to train the meta-learner. Two base learners are input with the validation set, and the output meta-learning vectors are input with the meta-learners to output the predicted oil recovery efficiency. Then, the Pareto front solution is output using the NSGA-II algorithm, and the equilibrium solution is output using the TOPSIS decision. The performance of the coupled machine learning-based model after training is evaluated.

4. The method according to claim 1, characterized in that, In step S2, the oil temperature is 80~100℃, the oil supply pressure is 0.2MPa~0.5MPa, the radial oil receiving ring speed is 15000rpm~20000rpm, the nozzle angle is 10°~15°, and the nozzle distance is 10mm~20mm. The nozzle angle refers to the angle between the centerline of the nozzle and the axis of the radial oil-collecting ring, and the nozzle distance refers to the horizontal distance between the nozzle tip and the radial oil-collecting ring.

5. A system for testing the efficiency of radial ring-type oil supply and lubrication in high-speed aerospace bearings, characterized in that... This method is used to implement the radial ring under-lubrication oil recovery efficiency test method for aviation high-speed bearings as described in any one of claims 1 to 4. The system includes a testing system, an experimental platform, and a lubricating oil system. The experimental platform includes a radial oil-receiving ring drive device and a nozzle drive device. The test control system is used to control the start and stop of the experimental platform and lubricating oil system, as well as adjust the radial oil receiving ring speed, nozzle angle, nozzle distance, oil supply pressure and oil temperature; The test and control system is also used to set the oil recovery efficiency threshold and basic parameters, and to collect actual parameter data in real time. Based on the collected data, the oil recovery efficiency is calculated. If the oil recovery efficiency meets the standard, the lubricating oil system is controlled to adjust the oil supply pressure and oil temperature. If it does not meet the standard, the actual parameter data is collected again in real time using the oil recovery ring speed, nozzle angle, and nozzle distance calculated by the Gaussian process surrogate model optimized by the differential evolution algorithm. A multi-objective dataset is constructed using the optimal oil recovery efficiency under each target parameter. A machine learning-based coupled model processes the multi-objective dataset to output an equilibrium solution, and it is determined whether the oil recovery efficiency corresponding to the equilibrium solution meets the standard. The experimental process of calculating the oil recovery efficiency and adjusting the oil supply pressure and oil temperature is repeated to finally obtain the oil recovery efficiency test dataset of the radial oil recovery ring.

6. The system according to claim 5, characterized in that, The radial oil collection ring drive device includes an oil collection seat, an oil collection simulation shaft and a high-speed motor. The oil collection seat is provided with an oil collection cavity, and the oil collection simulation shaft is used to support the radial oil collection ring. The oil collection simulation shaft is rotatably mounted on the oil collection seat. One end of the oil collection channel of the oil collection simulation shaft is connected to the oil collection groove of the radial oil collection ring, and the other end is connected to the oil collection cavity. The high-speed motor is driven by the simulated oil recovery shaft.

7. The system according to claim 6, characterized in that, The oil collection simulation shaft has a support platform for a radial oil collection ring, the oil inlets of the plurality of oil collection channels are located on the support platform, and the oil outlets of the oil collection channels are located on the end face of the oil collection simulation shaft. The oil collecting channel is inclined toward the outer peripheral wall of the oil collection simulation shaft along the direction of oil flow.

8. The system according to claim 7, characterized in that, The simulated oil recovery bushing is equipped with a locking ring, which positions the radial oil recovery ring against the abutment. The end of the oil recovery simulation shaft is detachably connected to the output end of the high-speed motor.

9. The system according to claim 5, characterized in that, The nozzle driving device includes a worm gear mechanism and a lead screw guide mechanism. The nozzle is fixed to the worm gear of the worm gear mechanism by a mounting base, and the worm gear mechanism is mounted on the slide plate of the lead screw guide mechanism. The worm gear mechanism is used to adjust the angle of the nozzle, and the lead screw guide mechanism is used to adjust the distance of the nozzle.

10. The system according to claim 6, characterized in that, The lubricating oil system includes an oil tank, an oil supply line, and an oil return line. The oil supply line delivers the oil from the oil tank to the nozzle. The oil outlet of the oil collecting chamber has an oil collection tank. The radial oil collection ring drive device and the nozzle drive device are located on the experimental platform. The experimental platform has an oil return port, and the oil return port has a corresponding oil return tank. The oil collection tank and the oil return tank are both connected to the oil return circuit.

Images