Online Measurement Method for Characteristics of Foreign Objects Invading Bearings Driven by Model and Data Jointly

Abstract

The present invention discloses an online measurement method for the characteristics of foreign object intrusion in bearings driven jointly by models and data; First, a bearing foreign object test is carried out, and the outer ring temperature, inner ring rotation speed, cage rotation speed and vibration acceleration are measured at the macroscopic level, and the scratch morphology structure on the bearing surface is measured at the microscopic level to obtain test data; Secondly, the software ABAQUS, ADAMS and CFD are respectively used to carry out bearing foreign object dynamics simulation and fluid simulation to obtain simulation data such as the scratch morphology structure, cage rotation speed, vibration acceleration and outer ring temperature of the bearing under different working conditions; Then, the differences in outer ring temperature, cage rotation speed, vibration acceleration and scratch morphology structure are calculated respectively based on the test data and the simulation data, and the TOPSIS algorithm is used to determine the weights of the differences to obtain a comprehensive objective function; Finally, the multi-objective variable-speed grey wolf optimization algorithm is used to optimize the size, density, elastic modulus and Poisson's ratio of the bearing foreign object simulation model to obtain the true structure and material parameters of the foreign object.

Description

Technical Field

[0001] The present invention relates to the field of bearing foreign object monitoring, and particularly to an online measurement method for the characteristics of bearing intrusion foreign objects driven jointly by a model and data. Background Art

[0002] The entry of external foreign objects into the bearing interior is a common phenomenon in engineering practice. The wear of the bearing caused by the entry of external contaminants into the bearing interior and the lubrication system is one of the main reasons for bearing failure. Especially for high-speed aviation bearings, due to the particularity of their working environment, the cost of bearing replacement and maintenance is extremely high. Therefore, it is very necessary to explore an online measurement method for the characteristics of bearing intrusion foreign objects under variable working conditions.

[0003] Practice shows that although there are currently many studies on preventing foreign object intrusion into bearings, including ultrasonic cleaning of bearing components and lubricating oil to improve the cleanliness of the bearing working environment as much as possible. However, limited by the working environment and the inherent characteristics of mechanical equipment, foreign object intrusion always exists more or less. Analyzing the characteristics of foreign objects is the most fundamental measure to study its generation mechanism and reduce foreign object intrusion.

[0004] Currently, most of the research on the measurement method of bearing foreign objects is based on the observation and analysis after bearing disassembly. However, the disassembly cost of an assembled bearing is high and the technical difficulty is high, and foreign objects will be further damaged or even lost during the disassembly process, which is not conducive to the comprehensive and accurate analysis of foreign objects. Currently, there is no research on the online measurement technology of foreign objects at home and abroad. At the same time, with the continuous development of multidisciplinary simulation, deep learning, and data monitoring technologies, it has become possible to measure the characteristics of online foreign objects based on the multi-source information of the bearing measured online. Summary of the Invention

[0005] Object of the Invention: Based on the problems existing in the above background art, the present invention provides an online measurement method for the characteristics of bearing intrusion foreign objects driven jointly by a model and data, which can accurately measure the structural and material parameters such as the size, density, elastic modulus, and Poisson's ratio of the bearing intrusion foreign objects under variable working conditions.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] An online measurement method for the characteristics of bearing intrusion foreign objects driven jointly by a model and data, the method comprising:

[0008] Step S1: Conduct bearing foreign object tests to obtain variable working condition test data. The variable working condition test data is: adding foreign objects with different structures and materials to the bearing under different loads, speeds, and lubrication conditions, respectively measuring the test data of the outer ring temperature, cage speed, and vibration acceleration of the bearing, and obtaining the test data of the morphological structure of the bearing surface scratches;

[0009] Step S2: Conduct dynamic simulation of bearing foreign objects and fluid simulation of bearing foreign objects corresponding to the working conditions of Step S1 to obtain simulation data. The simulation data are as follows: obtain outer ring temperature simulation data based on CFD finite element simulation, obtain simulation data of cage rotation speed and vibration acceleration based on ADAMS multibody simulation, and obtain simulation data of bearing scratch morphology structure based on ABAQUS finite element simulation;

[0010] Step S3: Calculate the differences in outer ring temperature, cage rotation speed, vibration acceleration, and scratch morphology structure respectively based on the variable working condition test data and simulation data obtained in Steps S1 and S2, and use the TOPSIS algorithm to determine the weights of each difference, and then obtain the comprehensive objective function;

[0011] Step S4: Optimize the size, density, elastic modulus, and Poisson's ratio of the bearing foreign object simulation model using the multi-objective variable-speed grey wolf optimization algorithm based on the comprehensive objective function obtained in Step S3, and then obtain the real structure and material parameters of the foreign object.

[0012] Furthermore, Step S1 includes:

[0013] Step S101: Conduct bearing foreign object tests and prepare foreign objects with different structures and materials;

[0014] Step S102: Add foreign objects with different structures and materials to the bearing under different loads, speeds, and lubrication working conditions, and simultaneously measure the outer ring temperature T e , inner ring rotation speed ω ei , cage rotation speed ω ec and vibration acceleration test data a e ;

[0015] Step S103: Observe the morphology structure S of the bearing surface scratches e .

[0016] Furthermore, Step S2 includes:

[0017] Step S201: Build a bearing multibody simulation model based on ADAMS, conduct dynamic simulation of bearing foreign objects corresponding to the working conditions of Step S102, and measure the cage rotation speed ω sc and vibration acceleration data a s ;

[0018] Step S202: Build a bearing fluid simulation model based on CFD, conduct fluid simulation of bearing foreign objects corresponding to the working conditions of Step S102, and measure the outer ring temperature T s ;

[0019] Step S203: Based on ABAQUS, build a finite element simulation model of the bearing stress and strain, and conduct a bearing stress and strain simulation corresponding to the working conditions in Step S102 to measure the morphological structure S of the scratch on the bearing surface s .

[0020] Further, the said Step S3 includes:

[0021] Step S301: Calculate the bearing outer ring temperature difference T r , the cage rotational speed difference ω rc , the vibration acceleration difference a rc and the scratch morphological structure difference S r :

[0022] T r = |T s - T e | (1)

[0023] ω rc = |ω sc - ω ec | (2)

[0024] a rc = |a s - a e | (3)

[0025] S r = |S s - S e | = |S sl - S el | + |S sw - S ew | + |S sd - S ed | (4) In the formula: S sl is the length of the scratch on the bearing surface measured by the simulation model; S el is the length of the scratch on the bearing surface measured by the test; S sw is the width of the scratch on the bearing surface measured by the simulation model; S ew is the width of the scratch on the bearing surface measured by the test; S sd is the depth of the scratch on the bearing surface measured by the simulation model; S ed is the depth of the scratch on the bearing surface measured by the test; |·| represents taking the absolute value of the variable;

[0026] Step S302: Use the TOPSIS algorithm to determine the bearing outer ring temperature difference T r , the cage rotational speed difference ω rc , the vibration acceleration difference a rcand the difference S in scratch morphology structure r The weight coefficients are obtained to obtain the comprehensive objective function.

[0027] Furthermore, the comprehensive objective function is obtained in the following manner:

[0028] Step S3021, Homogenization of evaluation indicators:

[0029] The bearing outer ring temperature difference T obtained in step S301 r , the difference ω in cage rotation speed rc , the difference a in vibration acceleration rc and the difference S in scratch morphology structure r are standardized, and a standardized matrix χ = {X1, X2,... X α} and a weight vector ω τ = {W1, W2, W3, W4} are obtained, where α is the dimension of the data to be evaluated;

[0030] Step S3022, Calculation of the normalized matrix:

[0031] The matrix E is obtained by homogenizing the index attributes in the original dataset, and the weighted normalized matrix Z is constructed:

[0032]

[0033] In the formula: is the weight of the j ω th attribute, and the index weights are determined by the entropy weight method;

[0034] Step S3023, Entropy weight method calculation process:

[0035] Calculate each element in the probability matrix P as follows:

[0036]

[0037] Calculate the entropy of each index respectively as follows:

[0038]

[0039] The weight coefficients are calculated as follows:

[0040]

[0041] Step S3024, Calculation of the distances of each sample index from the positive and negative ideal solutions:

[0042] The formula for the positive ideal solution is as follows:

[0043]

[0044] In the formula: max{·} represents taking the maximum value of the variable; Z + represents the positive ideal solution;

[0045] The calculation formula for the negative ideal solution is as follows:

[0046]

[0047] In the formula: min{·} represents taking the minimum value of the variable; Z - represents the negative ideal solution;

[0048] The calculation formula for the distance of each sample index from the positive ideal solution is as follows:

[0049]

[0050] The calculation formula for the distance of each sample index from the negative ideal solution is as follows:

[0051]

[0052] Step S3025, Calculation of evaluation results:

[0053] Calculate the closeness degree of each evaluation object to the optimal solution and according to sort by size and give the evaluation result:

[0054]

[0055] In the formula: The closer to 1, the better the sample score;

[0056] Step S3026, Calculation of comprehensive objective function:

[0057] Design the comprehensive objective function to minimize formula (14) and formula (15):

[0058]

[0059]

[0060] In the formula: t is time; W1, W2, W3, and W4 are the weight coefficients corresponding to the temperature difference T of the bearing outer ring r , the rotational speed difference ω of the cage rc , the vibration acceleration difference a rc and the scratch morphology structure difference S r respectively; J1 is the optimization function of the dynamic characteristic error target; J2 is the optimization function of the scratch shape error target.

[0061] Furthermore, the said step S4 includes:

[0062] Step S401: Set the optimization ranges of the size φ, density ρ, elastic modulus G, and Poisson's ratio τ of the bearing foreign object as follows:

[0063] φ1 < φ < φ2 (16)

[0064] ρ1 < ρ < ρ2 (17)

[0065] G1 < G < G2 (18)

[0066] τ1 < τ < τ2 (19)

[0067] Where: φ1 is the lower limit of the size optimization range; φ2 is the upper limit of the size optimization range; ρ1 is the lower limit of the density optimization range; ρ2 is the upper limit of the density optimization range; G1 is the lower limit of the elastic modulus optimization range; G2 is the upper limit of the elastic modulus optimization range; τ1 is the lower limit of the Poisson's ratio optimization range; τ2 is the upper limit of the Poisson's ratio optimization range;

[0068] Step S402: Based on the comprehensive objective function obtained in Step S302, use the multi-objective variable-speed grey wolf optimization algorithm to optimize the size, density, elastic modulus, and Poisson's ratio of the foreign object in the bearing foreign object simulation. The calculation process of the optimization algorithm is as follows:

[0069]

[0070]

[0071] Where: and are the positions of the grey wolf and the prey, respectively; and are the coefficient vectors, respectively;

[0072] is an intermediate auxiliary variable; and are random vectors between 0 and 1; is a vector that linearly decreases during the iteration process;

[0073]

[0074]

[0075] Where: and are intermediate auxiliary variables, respectively; and are coefficient vectors, respectively; and are α GWO 、β GWO and δ GWOPosition of the gray wolf;

[0076] The update formulas for speed and position components are as follows:

[0077]

[0078] In the formula: is the moving speed of the gray wolf population; is the inertia factor; c VSGWO1 , c VSGWO2 and c VSGWO3 are the learning factors respectively; r VSGWO1 = random(0, 1), r VSGWO2 = random(0, 1) and r VSGWO3 = random(0, 1) are random numbers respectively.

[0079] The beneficial effects of the present invention are:

[0080] It can realize the online measurement of structural and material parameters such as the size, density, elastic modulus, and Poisson's ratio of foreign objects invading bearings under variable working conditions, and further provide a basis for reducing the damage caused by foreign object invasion to bearings and even affecting the working state of bearings. At the same time, it can also trace the origin of foreign objects and analyze the generation mechanism of foreign objects according to the structural and material parameters of foreign objects. Description of the Drawings

[0081] Figure 1 is a schematic flow diagram of the online measurement method for the characteristics of foreign objects invading bearings driven jointly by a model and data provided in Embodiment 1;

[0082] Figure 2 is the cylindrical roller bearing used in Embodiment 1, where a is the bearing geometric structure and b is the bearing mesh model;

[0083] Figure 3 is the bearing surface scratch diagram under Working Condition 1 in Table 2 provided in Embodiment 1, where a is the macroscopic morphology diagram of the bearing scratch and b is the microscopic morphology diagram of the bearing scratch;

[0084] Figure 4 is the bearing surface scratch diagram under Working Condition 2 in Table 2 provided in Embodiment 1, where a is the macroscopic morphology diagram of the bearing scratch and b is the microscopic morphology diagram of the bearing scratch;

[0085] Figure 5 is the bearing multi-body simulation model based on ADAMS provided in Embodiment 1;

[0086] Figure 6 is the bearing fluid simulation model based on CFD provided in Embodiment 1;

[0087] Figure 7The finite element simulation diagram of bearing foreign object intrusion based on ABAQUS provided in Example 1, where a is the global diagram and b is the partial enlarged diagram;

[0088] Figure 8 The foreign object finite element model provided in Example 1, where a is a foreign object with dimensions of 0.1×0.5×1mm 3 and b is a foreign object with dimensions of 0.05×0.5×0.2mm 3 ;

[0089] Figure 9 The simulation morphology structure diagram of bearing surface scratches provided in Example 1, where a is the global diagram and b is the partial enlarged diagram;

[0090] Figure 10 The finite element simulation model of bearing foreign object (displaying foreign object) with a bearing speed of 20000r / min and foreign object size of 0.1×0.5×1mm provided in Example 1 3 , where a is the diagram of the maximum contact initial stress (250step), b is the diagram of the maximum contact stress (700step), c is the diagram of the maximum contact stress after contact (982step), and d is the diagram of the maximum contact stress during contact (792step);

[0091] Figure 11 The finite element simulation model of bearing foreign object (not displaying foreign object) with a bearing speed of 20000r / min and foreign object size of 0.1×0.5×1mm provided in Example 1 3 , where a is the diagram of the maximum contact initial stress (264step), b is the diagram of the maximum contact stress (700step), c is the diagram of the maximum contact stress after contact (972step), and d is the diagram of the maximum contact stress during contact (792step);

[0092] Figure 12 The finite element simulation model of bearing foreign object (displaying foreign object) with a bearing speed of 10000r / min and foreign object size of 0.1×0.5×1mm provided in Example 1 3 , where a is the diagram of the maximum contact initial stress (214step), b is the diagram of the maximum contact stress (332step), c is the diagram of the maximum contact stress after contact (661step), and d is the diagram of the maximum contact stress during contact (541step);

[0093] Figure 13 The finite element simulation model of bearing foreign object (displaying foreign object) with a bearing speed of 10000r / min and foreign object size of 0.1×0.5×1mm provided in Example 1 3The finite element simulation model of bearing foreign objects (foreign objects not shown), where a is the maximum initial contact stress (214step) diagram, b is the maximum contact stress (420step) diagram, c is the maximum post-contact stress (660step) diagram, and d is the maximum contact stress (444step) diagram;

[0094] Figure 14 It is the finite element simulation model of bearing foreign objects with a bearing speed of 5000 r / min and a foreign object size of 0.1×0.5×1 mm provided in Example 1 3 The finite element simulation model of bearing foreign objects (foreign objects shown), where a is the maximum initial contact stress (314step) diagram, b is the maximum contact stress (697step) diagram, c is the maximum post-contact stress (1246step) diagram, and d is the maximum contact stress (723step) diagram;

[0095] Figure 15 It is the finite element simulation model of bearing foreign objects with a bearing speed of 5000 r / min and a foreign object size of 0.1×0.5×1 mm provided in Example 1 3 The finite element simulation model of bearing foreign objects (foreign objects not shown), where a is the maximum initial contact stress (314step) diagram, b is the maximum contact stress (697step) diagram, c is the maximum post-contact stress (1245step) diagram, and d is the maximum contact stress (788step) diagram;

[0096] Figure 16 It is the finite element simulation model of bearing foreign objects with a bearing speed of 20000 r / min and a foreign object size of 0.05×0.5×0.2 mm provided in Example 1 3 The finite element simulation model of bearing foreign objects (foreign objects shown), where a is the maximum initial contact stress (323step) diagram, b is the maximum contact stress (686step) diagram, c is the maximum post-contact stress (985step) diagram, and d is the maximum contact stress (641step) diagram;

[0097] Figure 17 It is the finite element simulation model of bearing foreign objects with a bearing speed of 20000 r / min and a foreign object size of 0.05×0.5×0.2 mm provided in Example 1 3 The finite element simulation model of bearing foreign objects (foreign objects not shown), where a is the maximum initial contact stress (316step) diagram, b is the maximum contact stress (520step) diagram, c is the maximum post-contact stress (972step) diagram, and d is the maximum contact stress (683step) diagram;

[0098] Figure 18 It is the bearing vibration acceleration test data provided in Example 1, where a is the vibration data in the X direction, b is the vibration data in the Y direction, and c is the vibration data in the Z direction;

[0099] Figure 19 The bearing vibration acceleration simulation data provided in Example 1, where a is the vibration data in the X direction, b is the vibration data in the Y direction, and c is the vibration data in the Z direction;

[0100] Figure 20 The bearing cage rotation speed data provided in Example 1, where a is the bearing cage rotation speed test data, b is the bearing cage rotation speed simulation data, and c is the difference between the simulation data and the test data;

[0101] Figure 21 The optimization results of the multi-objective variable-speed grey wolf optimization algorithm provided in Example 1. Detailed implementation manners

[0102] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0103] Example 1

[0104] Refer to Figures 1 - 21 , this embodiment provides an online measurement method for the characteristics of foreign objects invading bearings driven jointly by models and data. The flow of this method is as Figure 1 shown. In this embodiment, the cylindrical roller bearing N208 is taken as an example to verify the algorithm performance. The bearing geometric structure is as Figure 2 shown in a, and the bearing mesh model is as Figure 2 shown in b. The specific strategy includes the following steps:

[0105] Step S1: Conduct bearing foreign object tests. Add foreign objects with different structures and materials to the bearing under different loads, rotation speeds, and lubrication conditions. Measure the test data of the outer ring temperature, inner ring rotation speed, cage rotation speed, and vibration acceleration simultaneously, and use a microscope to observe the morphological structure test data of the scratches on the bearing surface;

[0106] Specifically, in this embodiment, this step S1 includes:

[0107] Step S101: Conduct bearing foreign object tests. Prepare foreign objects with different structures and materials. The types and sizes of the foreign objects are shown in Table 1:

[0108] Table 1 Foreign objects of different types and sizes

[0109]

[0110]

[0111] Step S102: Add foreign objects with different structures and materials to the bearing under different loads, speeds, and lubrication conditions, and simultaneously measure the outer ring temperature T of the bearing e of the inner ring speed ω ei of the cage speed ω ec and the vibration acceleration test data a e ; The working condition parameters are shown in Table 2:

[0112] Table 2 Working condition parameters

[0113] Operating condition Inner ring rotational speed / r / min Lubricating oil flow rate / L / min Load / kN 1 20000 2.5 1 2 10000 2.5 1 3 5000 2.5 1

[0114] Step S103: Observe the morphological structure S of the scratches on the bearing surface using a microscope e , Figure 3 is the microscopic image of the scratch under working condition 1, and the scratch size is 2 mm in length × 0.3 mm in width × 0.059 mm in depth; Figure 4 is the microscopic image of the scratch under working condition 2, and the scratch size is 0.9 mm in length × 0.1 mm in width × 0.01 mm in depth. Further physical and chemical analysis results of the bearing after foreign object intrusion are shown in Table 3:

[0115] Table 3 Quantitative test analysis results of the chemical compositions of the inner race, outer race, and rollers (wt%)

[0116]

[0117] Step S2: Conduct bearing foreign object dynamics simulation and bearing foreign object fluid simulation corresponding to the working conditions of Step S1. Obtain the outer ring temperature simulation data based on CFD finite element simulation, obtain the simulation data such as cage speed and vibration acceleration based on ADAMS multibody simulation, and obtain the bearing scratch morphological structure simulation data based on ABAQUS finite element simulation;

[0118] Specifically, in this embodiment, this Step S2 includes:

[0119] Step S201: Build a bearing multibody simulation model based on ADAMS. As Figure 5 shown, the material parameters of different bearing components are shown in Table 4. Conduct bearing foreign object dynamics simulation corresponding to the working conditions of Step S102, and simultaneously measure the inner ring speed ω si of the cage speed ω sc and the vibration acceleration data a s ;

[0120] Table 4 Material parameters of different bearing components

[0121] <![CDATA[Density / kg / mm 3 > Elastic modulus / MPa Poisson's ratio Bearing roller <![CDATA[7.85×10 -9 > 210000 0.27 Bearing inner and outer rings <![CDATA[7.85×10 -9 > 210000 0.27

[0122] Step S202: Based on CFD, establish a bearing fluid simulation model, perform bearing foreign object fluid simulation corresponding to the working conditions in Step S201, and measure the temperature T of the outer ring of the bearing at the same time s ;

[0123] Use ANSYS ICEM software to perform mesh division on the bearing. The mesh model includes a nozzle domain, a fluid domain inside the bearing cavity, and two side return domains. Structured meshes are used for all fluid domains, and the meshes in the contact areas between the rolling elements and the inner and outer raceways are refined. When the bearing is running at high speed, the inside of the bearing cavity is in a complex oil-gas two-phase flow state. The VOF model is used to track the oil-gas two-phase interface and the flow process of the fluid, and the RNG k-ε turbulence model is selected. The bearing fluid simulation model is as Figure 6 shown

[0124] Step S203: Based on ABAQUS, establish a bearing stress-strain finite element simulation model (the bearing finite element simulation model considering foreign objects is as Figure 7 shown, and the foreign object finite element model is as Figure 8 shown), perform bearing stress-strain simulation corresponding to the working conditions in Step S102, and measure the morphological structure S of the scratch on the bearing surface at the same time s , the morphological structure of the bearing scratch under Condition 1 is as Figure 9 shown, and the scratch size is 1.75 mm in length × 0.32 mm in width × 0.065 mm in depth; further, the element results of the bearing simulation model after foreign object intrusion are shown in Table 5 as follows

[0125] Table 5 Element simulation analysis results of bearing foreign object intrusion (wt%)

[0126]

[0127] When there is a foreign object, a combined simulation model is built using Hypermesh and ABAQUS: Mesh generation is performed on the bearing roller and the inner ring of the bearing using Hypermesh. Considering the small size of the foreign object and taking into account the impact of the number of meshes on the simulation calculation time, both the roller and the inner ring of the bearing are internally tempered, while the outer ring of the bearing maintains its physical properties. At the same time, only a thin slice structure of "roller - foreign object - inner ring" with a thickness of 0.5 mm is cut for finite element analysis. The mesh size of the roller is 0.1 mm and the number of meshes is 11,840; the mesh size of the inner ring is 0.4 mm and the number of meshes is 10,800. Due to the small size of the foreign object, the stress at the contact mainly concentrates near the edges of the roller and the inner ring of the bearing. Therefore, both the inside of the bearing roller and the inner ring are connected as rigid units through the COUP_KIN element type for tempering treatment. After completing the mesh generation and material property settings in Hypermesh, the model is imported into Abaqus, and then relevant settings are made for its contact, simulation steps, bearing speed, result output, etc.: On the basis of coupling the bearing roller and the inner ring, the contact between the inner ring of the bearing roller and the foreign object is set as a surface - to - surface contact with a tangential friction coefficient of 0.2 and a normal "hard contact". The rotation directions of the bearing roller and the inner ring are opposite, and the bearing speeds are 20,000 r / min, 10,000 r / min, and 5,000 r / min respectively. And the implicit dynamics method is used for simulation analysis.

[0128] (1) Influence of bearing speed on bearing stress when there is a foreign object (bearing speed is 20,000 r / min, foreign object size is 0.1×0.5×1 mm 3 )

[0129] Due to the large difference in size between the foreign object and the bearing roller and the inner ring of the bearing, and not being concerned about the force on the foreign object during the contact process, the foreign object is meshed into 3×1 mesh elements. Set the bearing speed to 20,000 r / min. During its rolling process, due to the presence of the foreign object and the too high bearing speed, the analysis step size is 0.0001 seconds, and the minimum increment step is 1×10 -8 seconds, and a large - deformation nonlinear analysis of 1000 steps is performed.

[0130] Figure 10 a. When the foreign object size is 0.1×0.5×1 mm at a rotational speed of 20,000 r / min 3 , the maximum initial contact stress (250 steps) is 4000 MPa; Figure 10 b. When the foreign object size is 0.1×0.5×1 mm at a rotational speed of 20,000 r / min 3 , the maximum contact stress during contact (700 steps) is 21,690 MPa; Figure 10 c. When the foreign object size is 0.1×0.5×1 mm at a rotational speed of 20,000 r / min3 When it is [condition], the maximum stress after contact (982 steps) is 4047 MPa; Figure 10 d is at a rotational speed of 20000 r / min, and the foreign object size is 0.1×0.5×1 mm 3 When it is [condition], the maximum stress during contact (792 steps) is 33860 MPa. From the above simulation results, it can be seen that during the operation of the bearing, due to stress concentration and other reasons, the maximum stress usually occurs on the foreign object. Since the damage to the raceway is one of the main reasons for the premature failure of the bearing, therefore, the following four groups of pictures are the stress nephograms of the bearing during the simulation process, and at the same time, two groups of comparison nephograms with and without the display of foreign objects are also retained at other rotational speeds.

[0131] Figure 11 a is at a rotational speed of 20000 r / min, without displaying the foreign object, the maximum stress at the beginning of contact (264 steps) is 9051 MPa; Figure 11 b is at a rotational speed of 20000 r / min, without displaying the foreign object, the maximum stress during contact (700 steps) is 21690 MPa; Figure 11 c is at a rotational speed of 20000 r / min, without displaying the foreign object, the maximum stress after contact (972 steps) is 7222 MPa; Figure 11 d is at a rotational speed of 20000 r / min, without displaying the foreign object, the maximum stress during contact (792 steps) is 33860 MPa.

[0132] (2) Influence of bearing rotational speed on bearing stress when there is a foreign object (bearing rotational speed is 10000 r / min, foreign object size is 0.1×0.5×1 mm 3 ):

[0133] The size of the foreign object is 0.1×0.5×1 mm 3 , set the bearing rotational speed to 10000 r / min. During its rolling process, due to the existence of the foreign object and the relatively high bearing rotational speed, the analysis step size is 0.0001 s, the minimum increment step is 1×10 -8 s, and a large deformation non-linear analysis of 1000 steps is performed.

[0134] Figure 12 a is at a rotational speed of 10000 r / min, foreign object size 0.1×0.5×1 mm 3 When it is [condition], the maximum stress at the beginning of contact (214 steps) is 14930 MPa; Figure 12 b is at a rotational speed of 10000 r / min, foreign object size 0.1×0.5×1 mm 3 When it is [condition], the maximum stress during contact (332 steps) is 23980 MPa; Figure 12 c is at a rotational speed of 10000 r / min, foreign object size 0.1×0.5×1 mm 3When the time is [specific time], the maximum stress after contact (661 step) is 24910 MPa; Figure 12 When d is at a rotational speed of 10000 r / min, the foreign object size is 0.1×0.5×1 mm 3 When the time is [specific time], the maximum stress during contact (541 step) is 38720 MPa. From the above simulation results, it can be seen that during the operation of the bearing, due to stress concentration and other reasons, the maximum stress usually occurs on the foreign object. Since the damage of the raceway is one of the main reasons for the premature failure of the bearing, therefore, the following four groups of pictures are the stress nephograms of the bearing during the simulation process, and at the same time, two groups of comparison nephograms with and without the display of foreign objects are also retained at other rotational speeds.

[0135] Figure 13 When a is at a rotational speed of 10000 r / min and the foreign object is not displayed, the maximum stress at the beginning of contact (214 step) is 8442 MPa; Figure 13 When b is at a rotational speed of 10000 r / min and the foreign object is not displayed, the maximum stress during contact (420 step) is 19110 MPa; Figure 13 When c is at a rotational speed of 10000 r / min and the foreign object is not displayed, the maximum stress after contact (660 step) is 6023 MPa; Figure 13 When d is at a rotational speed of 10000 r / min and the foreign object is not displayed, the maximum stress during contact (444 step) is 20450 MPa.

[0136] (3) Influence of bearing rotational speed on bearing stress when there is a foreign object (bearing rotational speed is 5000 r / min, foreign object size is 0.1×0.5×1 mm 3 )

[0137] The size of the foreign object is 0.1×0.5×1 mm 3 , set the bearing rotational speed to 5000 r / min. During its rolling process, due to the presence of the foreign object and the relatively high bearing rotational speed, the analysis step size is 0.001 seconds, and the minimum increment step is 1×10 -8 seconds, and a large deformation non-linear analysis of 2000 steps is performed.

[0138] Figure 14 When a is at 5000 r / min and the foreign object size is 0.1×0.5×1 mm 3 , the maximum stress at the beginning of contact (314 step) is 9550 MPa; Figure 14 When b is at 5000 r / min and the foreign object size is 0.1×0.5×1 mm 3 , the maximum stress during contact (697 step) is 30690 MPa; Figure 14 When c is at 5000 r / min and the foreign object size is 0.1×0.5×1 mm 3When the time is [specific time], the maximum stress after contact (1246 steps) is 15990 MPa; Figure 14 d is at 5000 r / min, the foreign object size is 0.1×0.5×1 mm 3 When the time is [specific time], the maximum stress during contact (723 steps) is 30950 MPa. From the above simulation results, it can be seen that during the operation of the bearing, due to stress concentration and other reasons, the maximum stress usually occurs on the foreign object. Since the damage of the raceway is one of the main reasons for the premature failure of the bearing, therefore, the following four groups of pictures are the stress nephograms of the bearing during the simulation process, and at the same time, two groups of comparison nephograms with and without the display of the foreign object are also retained at other rotational speeds.

[0139] Figure 15 a is at 5000 r / min, when the foreign object is not displayed, the maximum stress at the beginning of contact (314 steps) is 1784 MPa; Figure 15 b is at 5000 r / min, when the foreign object is not displayed, the maximum stress during contact (697 steps) is 19010 MPa; Figure 15 c is at 5000 r / min, when the foreign object is not displayed, the maximum stress after contact (1245 steps) is 1846 MPa; Figure 15 d is at 5000 r / min, when the foreign object is not displayed, the maximum stress during contact (788 steps) is 20300 MPa.

[0140] (4) Influence of bearing rotational speed on bearing stress when there is a foreign object (bearing rotational speed is 20000 r / min, foreign object size is 0.05×0.5×1 mm 3 )

[0141] The size of the foreign object is 0.1×0.5×1 mm 3 , set the bearing rotational speed to 20000 r / min. During its rolling process, due to the presence of the foreign object and the relatively high bearing rotational speed, the analysis step size is 0.001 seconds, and the minimum increment step is 1×10 -8 seconds, and a large deformation non-linear analysis of 2000 steps is performed. For the foreign object with a size of 0.05×0.5×0.2 mm 3 when, the foreign object mesh is divided into 4×2 mesh elements. Due to the too high bearing rotational speed, the analysis step size is 0.0001 seconds, and the minimum increment step is 1×10 -8 seconds, and a large deformation non-linear analysis of 2000 steps is performed.

[0142] Figure 16 a is at a rotational speed of 20000 r / min, the foreign object size is 0.05×0.5×0.2 mm 3 when, the maximum stress at the beginning of contact (323 steps) is 4671 MPa; Figure 16When b is at a rotational speed of 20000 r / min and the foreign object size is 0.05×0.5×0.2 mm 3 the maximum stress during contact (686 steps) is 13890 MPa; Figure 16 When c is at a rotational speed of 20000 r / min and the foreign object size is 0.05×0.5×0.2 mm 3 the maximum stress after contact (985 steps) is 3931 MPa; Figure 16 When d is at a rotational speed of 20000 r / min and the foreign object size is 0.05×0.5×0.2 mm 3 the maximum stress during contact (641 steps) is 19470 MPa. From the above simulation results, it can be seen that during the operation of the bearing, due to stress concentration and other reasons, the maximum stress usually occurs on the foreign object. Since the damage of the raceway is one of the main reasons for the premature failure of the bearing, therefore, the following four groups of pictures are the stress nephograms of the bearing during the simulation process, and at the same time, two groups of comparative nephograms with and without the display of foreign objects are also retained at other rotational speeds.

[0143] Figure 17 When a is at a rotational speed of 20000 and without displaying the foreign object, the maximum stress at the beginning of contact (316 steps) is 1481 MPa; Figure 17 When b is at a rotational speed of 20000 r / min and without displaying the foreign object, the maximum stress during contact (520 steps) is 6531 MPa; Figure 17 When c is at a rotational speed of 20000 r / min and without displaying the foreign object, the maximum stress after contact (972 steps) is 1688 MPa; Figure 17 When d is at a rotational speed of 20000 r / min and without displaying the foreign object, the maximum stress during contact (683 steps) is 8888 MPa. When the bearing rotational speed is 20000 r / min and without displaying the foreign object, the comparison of the influence of the foreign object size on the bearing stress is shown in Table 6:

[0144] Table 6 Comparison table of the influence of foreign object size on bearing stress when the bearing rotational speed is 20000 r / min and without displaying the foreign object

[0145]

[0146] The influence of foreign object material on bearing stress during simulation: For bearing foreign objects, sometimes they are made of alumina and some are made of silica. Based on the different foreign object materials, through finite element simulation analysis, the influence of different materials on bearing stress is roughly explored. It is found during the simulation that due to the smaller elastic modulus of silica, the phenomenon of foreign object shearing will occur.

[0147] Step S3: Calculate the differences in outer ring temperature, cage rotation speed, vibration acceleration, and scratch morphology structure based on the variable condition test data and simulation data obtained in Steps S1 and S2 respectively, and use the TOPSIS method to determine the weight of each difference to obtain a comprehensive objective function;

[0148] Specifically, in this embodiment, Step S3 includes:

[0149] Step S301: Calculate the bearing outer ring temperature difference T r , vibration acceleration difference a rc (vibration acceleration test data a e as shown in Figure 18 , vibration acceleration simulation data a s as shown in Figure 19 ), cage rotation speed difference ω rc (cage test rotation speed data ω ec as shown in Figure 20 a, cage simulation rotation speed data ω sc as shown in Figure 20 b, cage test and simulation rotation speed data difference as shown in Figure 20 c) and scratch morphology structure difference S r :

[0150] T r =|T s -T e | (1)

[0151] ω rc =ω sc -ω ec | (2)

[0152] a rc =|a s -a e | (3)

[0153] S r =|S s -S e |=|S sl -S el |+S sw -S ew |+S sd -S ed | (4)

[0154] In the formula: S sl is the measured scratch length on the bearing surface by the simulation model; S el is the measured scratch length on the bearing surface by the test; S swThe width of the scratch on the bearing surface, which is a quantity measured by the simulation model; S ew The width of the scratch on the bearing surface, which is a quantity measured by the test; S sd The depth of the scratch on the bearing surface, which is a quantity measured by the simulation model; S ed The depth of the scratch on the bearing surface, which is a quantity measured by the test; |·| represents the absolute value of the variable;

[0155] Step S302: Use the TOPSIS method to determine the difference weights of each item to obtain the comprehensive objective function:

[0156] (1) Homogenization of evaluation indicators:

[0157] Standardize the temperature difference T of the bearing outer ring, r the rotational speed difference ω of the cage, rc the vibration acceleration difference a, rc and the scratch morphology structure difference S r obtained in step S301, and obtain the standardized matrix χ = {X1, X2, …, X α} and the weight vector ω τ = {W1, W2, W3, W4}, where α is the dimension of the data to be evaluated;

[0158] (2) Calculation of the normalized matrix:

[0159] Homogenize the index attributes in the original dataset to obtain the matrix E, and construct the weighted normalized matrix Z, which is obtained by multiplying E by ω τ :

[0160]

[0161] In the formula: is the weight of the j ω th attribute, and the index weight is determined by the entropy weight method;

[0162] (3) Calculation process of the entropy weight method:

[0163] Calculate each element in the probability matrix P

[0164]

[0165] as follows:

[0166]

[0167] Calculate the entropy of each index as follows:

[0168]

[0169] (4) Calculate the distances of each sample index from the positive and negative ideal solutions:

[0170] The calculation formula for the positive ideal solution is as follows:

[0171]

[0172] The calculation formula for the negative ideal solution is as follows:

[0173]

[0174] The calculation formula for the distance of each sample index from the positive ideal solution is as follows:

[0175]

[0176] The calculation formula for the distance of each sample index from the negative ideal solution is as follows:

[0177]

[0178] (5) Calculation of evaluation results:

[0179] Calculate the closeness degree of each evaluation object to the optimal solution And according to Sort by size and give the evaluation results:

[0180]

[0181] In the formula: The closer to 1, the better the sample score;

[0182] (6) Calculation of the comprehensive objective function:

[0183] Design the comprehensive objective function to minimize the following formula:

[0184]

[0185]

[0186] In the formula: t is time; W1, W2, W3, and W4 are the weight coefficients corresponding to the temperature difference T of the bearing outer ring r , the rotational speed difference ω of the cage rc , the vibration acceleration difference a rc and the scratch morphology structure difference S r ; J1 is the dynamic characteristic error target optimization function; J2 is the scratch shape error target optimization function;

[0187] Step S4: Based on the comprehensive objective function obtained in step S3, use the multi-objective variable-speed grey wolf optimization algorithm to optimize the size, density, elastic modulus, and Poisson's ratio of the bearing foreign object simulation, and then obtain the real structure and material parameters of the foreign object;

[0188] Specifically, in this embodiment, step S4 includes:

[0189] Step S401: Set the optimization ranges of the size φ, density ρ, elastic modulus G, and Poisson's ratio τ of the bearing foreign matter as follows:

[0190] φ1 < φ < φ2 (16)

[0191] ρ1 < ρ < ρ2 (17)

[0192] G1 < G < G2 (18)

[0193] τ1 < τ < τ2 (19)

[0194] Where: φ1 is the lower limit of the size optimization range, with a value of 0 mm; φ2 is the upper limit of the size optimization range, with a value of 1 mm; ρ1 is the lower limit of the density optimization range, with a value of 2×10 -9 kg / mm 3 ; ρ2 is the upper limit of the density optimization range, with a value of 4×10 -9 kg / mm; G1 is the lower limit of the elastic modulus optimization range, with a value of 50000 MPa; G2 is the upper limit of the elastic modulus optimization range, with a value of 500000 MPa; τ1 is the lower limit of the Poisson's ratio optimization range, with a value of 0.1; τ2 is the upper limit of the Poisson's ratio optimization range, with a value of 0.33.

[0195] Step S402: Based on the comprehensive objective function obtained in step S3, use the multi-objective variable-speed grey wolf optimization algorithm to optimize the foreign matter size, density, elastic modulus, and Poisson's ratio of the bearing foreign matter simulation. The multi-objective optimization results are as Figure 21 shown. The calculation process of the optimization algorithm is as follows:

[0196]

[0197]

[0198] Where: and are the positions of the grey wolf and the prey, respectively; and

[0199] are coefficient vectors, respectively; is an intermediate auxiliary variable; r1 and r2 are random vectors between 0 and 1; is a vector that linearly decreases during the iteration process;

[0200]

[0201]

[0202] Wherein: and are intermediate auxiliary variables respectively; and are coefficient vectors respectively; and are the positions of α GWO , β GWO and δ GWO of the gray wolf respectively;

[0203] The update formulas for the velocity and position components are as follows:

[0204]

[0205] Wherein: is the moving speed of the gray wolf population; is the inertia factor; c VSGWO1 , c VSGWO2 and c VSGWO3 are learning factors respectively; r VSGWO1 = random(0, 1), r VSGWO2 = random(0, 1) and r VSGWO3 = random(0, 1) are random numbers respectively. The optimization results are shown in Table 7:

[0206] Table 7 Data Analysis of the Optimization Results of the Foreign Object Size

[0207]

[0208] As can be seen from Table 7, parameters such as the density, elastic modulus, Poisson's ratio, and size of the foreign object have relatively small optimization errors and can achieve the expected effects.

[0209] Details not described in the present invention are all well-known techniques to those skilled in the art.

[0210] The preferred specific embodiments of the present invention have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations according to the concept of the present invention without creative labor. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field based on the concept of the present invention through logical analysis, reasoning, or limited experiments on the basis of the prior art should be within the protection scope determined by the claims.

Claims

1. An online measurement method for the characteristics of foreign objects invading bearings driven jointly by a model and data, characterized in that, The method includes: Step S1: Conduct a bearing foreign object test to obtain variable working condition test data. The variable working condition test data is as follows: Different foreign objects with different structures and materials are added to the bearing under different loads, speeds, and lubrication conditions, and the outer ring temperature, cage rotation speed, and vibration acceleration test data of the bearing are measured respectively, and the morphological structure test data of the scratches on the bearing surface is obtained. Step S1 includes: Step S101: Conduct a bearing foreign object test to prepare different foreign objects with different structures and materials. Step S102: Add foreign objects with different structures and materials into the bearing under different loads, speeds, and lubrication conditions, and simultaneously measure the temperature T of the outer ring of the bearing e , the rotational speed ω of the inner ring ei , the rotational speed ω of the cage ec and the vibration acceleration test data a e ; Step S103, observe the morphological structure S of the scratch on the bearing surface e ; Step S2: Conduct a bearing foreign object dynamics simulation and a bearing foreign object fluid simulation corresponding to the working conditions of Step S1 to obtain simulation data. The simulation data is as follows: The outer ring temperature simulation data is obtained based on CFD finite element simulation, the cage rotation speed and vibration acceleration simulation data are obtained based on ADAMS multibody simulation, and the bearing scratch morphological structure simulation data is obtained based on ABAQUS finite element simulation. Step S2 includes: Step S201: Build a multi-body simulation model of the bearing based on ADAMS, conduct a dynamic simulation of bearing foreign objects corresponding to the working conditions in Step S102, and measure the rotational speed ω of the bearing cage sc and the vibration acceleration data a s ; Step S202: Build a bearing fluid simulation model based on CFD, conduct bearing foreign object fluid simulation corresponding to the working conditions in Step S201, and measure the temperature T of the outer ring of the bearing s ; Step S203: Build a finite element simulation model of bearing stress and strain based on ABAQUS, perform bearing stress and strain simulation corresponding to the working conditions in Step S201, and measure the morphological structure S of the scratch on the bearing surface s ; Step S3: Calculate the differences in outer ring temperature, cage rotation speed, vibration acceleration, and scratch morphological structure based on the variable working condition test data and simulation data obtained in Steps S1 and S2 respectively, and use the TOPSIS algorithm to determine the weights of each difference, and then obtain a comprehensive objective function. Step S3 includes: Step S301: Calculate the bearing outer ring temperature difference T, the cage rotational speed difference ω, the vibration acceleration difference a, and the scratch morphology structure difference S respectively based on the off-design test data and simulation data obtained in Steps S1 and S2. r The cage rotational speed difference ω rc The vibration acceleration difference a rc And the scratch morphology structure difference S r : T r = |T s -T e | (1) ω rc = |ω sc - ω ec | (2) a rc = |a s -a e | (3) S r = |S s -S e | = |S sl -S el | + |S sw -S ew | + |S sd -S ed | (4) Where: S sl is the length of the bearing surface scratch measured by the simulation model; S el is the length of the bearing surface scratch measured by the experiment; S sw is the width of the bearing surface scratch measured by the simulation model; S ew is the width of the bearing surface scratch measured by the experiment; S sd is the depth of the bearing surface scratch measured by the simulation model; S ed is the depth of the bearing surface scratch measured by the experiment; |·| represents taking the absolute value of the variable; Step S302: Use the TOPSIS algorithm to determine the temperature difference T of the bearing outer ring r , the rotational speed difference ω of the cage rc , the vibration acceleration difference a rc and the scratch morphology structure difference S r of the weight coefficients to obtain the comprehensive objective function; The comprehensive objective function is to minimize equations (14) and (15): where: t is time; W1, W2, W3, and W4 are the weight coefficients corresponding to the temperature difference T of the bearing outer ring r , the rotational speed difference ω of the cage rc , the vibration acceleration difference a rc , and the scratch morphology structure difference S r ; J1 is the dynamic characteristic error target optimization function; J2 is the scratch shape error target optimization function; Step S4: Optimize the size, density, elastic modulus, and Poisson's ratio of the bearing foreign object simulation model using the multi-objective variable-speed grey wolf optimization algorithm based on the comprehensive objective function obtained in Step S3, and then obtain the real structure and material parameters of the foreign object. Step S4 includes: Step S401: Set the optimization ranges of the size φ, density ρ, elastic modulus G, and Poisson's ratio τ of the bearing foreign object as follows: φ1 < φ < φ2 (16) ρ1 < ρ < ρ2 (17) G1 < G < G2 (18) τ1 < τ < τ2 (19) Where: φ1 is the lower limit of the size optimization range; φ2 is the upper limit of the size optimization range; ρ1 is the lower limit of the density optimization range; ρ2 is the upper limit of the density optimization range; G1 is the lower limit of the elastic modulus optimization range; G2 is the upper limit of the elastic modulus optimization range; τ1 is the lower limit of the Poisson's ratio optimization range; τ2 is the upper limit of the Poisson's ratio optimization range. Step S402: Optimize the foreign object size, density, elastic modulus, and Poisson's ratio of the bearing foreign object simulation using the multi-objective variable-speed grey wolf optimization algorithm based on the comprehensive objective function obtained in Step S302. The calculation process of the optimization algorithm is as follows: Where: and are the positions of the grey wolf and the prey, respectively; and are coefficient vectors, respectively; is an intermediate auxiliary variable; and are random vectors between 0 and 1; is a vector that linearly decreases during the iteration process; In the formula: and are intermediate auxiliary variables respectively; and are coefficient vectors respectively; and are the positions of α GWO , β GWO and δ GWO of the grey wolf respectively; The update formulas for the velocity and position components are as follows: In the formula: is the moving speed of the gray wolf population; is the inertia factor; c VSGWO1 , c VSGWO2 and c VSGWO3 are the learning factors respectively; r VSGWO1 = random(0, 1), r VSGWO2 = random(0, 1) and r VSGWO3 = random(0, 1) are random numbers respectively.

2. The on-line measurement method for the characteristics of foreign object intrusion in a bearing jointly driven by a model and data according to claim 1, characterized in that The comprehensive objective function is obtained through the following method: Step S3021: Homogenize the evaluation indicators. The bearing outer ring temperature difference T obtained in step S301 r , the cage rotation speed difference ω rc , the vibration acceleration difference a rc and the scratch morphology structure difference S r are standardized, and a standardized matrix χ = {X1, X2,..., X α} and a weight vector ω τ = {W1, W2, W3, W4} are obtained, where α is the dimension of the data to be evaluated; Step S3022: Calculate the normalized matrix. Homogenize the index attributes in the original dataset to obtain matrix E, and construct the weighted normalized matrix Z: Where: is the weight of the j ω th attribute, and the index weight is determined by the entropy weight method; Step S3023: Entropy weight method calculation process: Calculate each element in the probability matrix P as follows: Calculate the entropy of each index respectively as follows: The calculation formulas for each weight coefficient are as follows: Step S3024: Calculate the distances of each sample index from the positive and negative ideal solutions: The calculation formula for the positive ideal solution is as follows: where: max{·} represents the maximum value of a variable; Z + represents the positive ideal solution; The calculation formula for the negative ideal solution is as follows: where: min{·} represents taking the minimum value of a variable; Z - represents the negative ideal solution; The calculation formula for the distance of each sample index from the positive ideal solution is as follows: The distance calculation formula for each sample index from the negative ideal solution is as follows: Step S3025, evaluation result calculation: Calculate the closeness degree of each evaluation object to the optimal solution And based on the magnitude for sorting and give the evaluation result: In the formula: The closer to 1, the better the sample score; Step S3026, comprehensive objective function calculation.

Images