Bearing anti-gluing design method based on internal structure parameter optimization

Abstract

The invention discloses a bearing anti-gluing design method based on internal structure parameter optimization, and relates to the technical field of bearing design. The problems that an existing bearing anti-gluing design mainly depends on experience, collaborative optimization of internal structure parameters is lacked, and the gluing risk is difficult to quantitatively predict are solved. The method comprises the following steps: firstly, constructing a basic database containing key internal structure parameters and typical service conditions of a bearing, obtaining gluing failure characteristic data through test or numerical simulation, and calculating a quantitative index-gluing risk factor for comprehensively representing a gluing risk; on this basis, an optimization model with the goal of minimizing a gluing risk factor and maximizing a bearing rated dynamic load and an operation stability index is established, a multi-objective optimization algorithm is adopted to carry out collaborative optimization on key structure parameters, and finally an optimal parameter combination is selected from a Pareto optimal solution set. According to the method, the bearing gluing risk can be quantitatively predicted and effectively reduced in the design stage, and the operation reliability of the bearing under harsh working conditions is remarkably improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of bearing design, in particular to the working condition adaptability design of bearings under harsh conditions such as high speed and heavy load, variable working conditions, and aims to improve the anti-gluing ability of bearing product design, and specifically relates to a bearing anti-gluing design method based on internal structure parameter optimization. BACKGROUND

[0002] Bearing is the core basic part of rotating machinery, and its performance directly affects the reliability, accuracy and service life of the equipment. Gluing (Scuffing or Scoring) is a serious failure form of bearings, especially rolling bearings, under conditions of high speed, heavy load, poor lubrication or severe variable working conditions. It is characterized by local metal tearing of the bearing working surface (raceway, rolling element) after fusion due to instantaneous high temperature, resulting in surface material transfer and serious damage, often accompanied by rapid increase of torque, vibration and temperature rise out of control, and finally leading to bearing seizure failure. Gluing failure is sudden, develops rapidly and is extremely harmful.

[0003] Traditional bearing anti-gluing design mainly relies on empirical formula, simplified calculation and a large number of bench simulation tests. Common methods include: selecting materials with good anti-gluing performance (such as high-temperature bearing steel, surface modification coating, special lubricant), optimizing heat treatment process, improving lubrication (such as selecting extreme pressure additive lubricating oil, increasing oil supply), controlling working temperature and load, etc. However, these methods often focus on the adjustment of external factors or a single type of element, and lack systematic and quantitative research on the complex interaction between internal structure parameters of bearings (such as raceway geometry design, rolling element size and number matching, cage detail design, surface micro-topography control, etc.) and their influence on gluing failure.

[0004] Currently, with the development of computer technology and numerical simulation methods, bearing performance simulation based on finite element analysis (FEA) and elastohydrodynamic lubrication (EHL) theory has been widely applied in the design stage. However, existing simulation optimization focuses on static load capacity, fatigue life or stiffness calculation, and the prediction model and optimization design method specifically for the complex physical and chemical process of gluing failure are still immature. Gluing failure involves multiple disciplines such as tribology, thermodynamics and materials science, and has many influencing factors and high nonlinearity, which is difficult to accurately predict and optimize through a single parameter or simple model.

[0005] In summary, the existing bearing anti-gluing design mainly relies on experience, lacks coordinated optimization of internal structure parameters, and is difficult to quantitatively predict gluing risk. SUMMARY

[0006] The application aims to solve the problems of major reliance on experience, lack of internal structure parameter optimization, and difficulty in quantitatively predicting the risk of gluing in the existing bearing anti-gluing design.

[0007] The technical solution of the application is:

[0008] A bearing anti-gluing design method based on internal structure parameter optimization, comprising the following steps

[0009] Step one: build a basic database

[0010] Step one: select the target bearing type, determine the key internal structure parameter set S = {S1, S2, …, Sm} that affects gluing failure; wherein Si represents the i-th structure parameter, and m is the total number of parameters;

[0011] Step two: set a plurality of typical service condition sets W = {W1, W2, …, Wn}; wherein Wj represents the j-th condition, and n is the number of conditions;

[0012] Step three: for each bearing structure parameter and service condition, obtain the gluing failure characteristic data of the bearing through physical bench test or numerical simulation;

[0013] Step two: calculate the gluing risk factor

[0014] According to the gluing failure characteristic data obtained in step one, the gluing risk factor γ under the corresponding parameter combination and condition is calculated; wherein the gluing risk factor γ is defined as the tendency or risk level of the bearing to occur gluing failure under the specific condition;

[0015] Step three: establish an association database

[0016] The bearing internal structure parameter combination S, service condition W in step one and the gluing risk factor γ calculated in step two are associated to establish a structure parameter-service condition-gluing risk factor basic database DB_base;

[0017] Step four: parameter optimization design

[0018] Step four: set the optimization goal: define a multi-objective optimization problem

[0019] Step four: define the design variables

[0020] Select a subset from the key internal structure parameter set S that significantly affects gluing and can be optimized and adjusted as design variables:

[0021] Step four: set the constraint conditions

[0022] S1: Design variable range constraint: each design variable Xi needs to satisfy its engineering feasible range Xi_min ≤ Xi ≤ Xi_max;

[0023] S2: Bearing basic performance constraint: static safety factor S0 ≥ S0_min, basic rated life L10 ≥ L10_req;

[0024] S3: Manufacturing process constraint: machining precision accessibility, cost limit;

[0025] Step four: apply multi-objective optimization algorithm to solve:

[0026] Under the constraint condition, the design variable X is optimized and solved; a set of non-dominated solutions, i.e. the Pareto optimal solution set Pareto_Set, is obtained, each solution in the set represents a design scheme that achieves a good compromise among multiple objectives;

[0027] Step five: scheme selection and verification:

[0028] Step five one: according to the specific needs of actual application, select a set of optimal solutions that best meet the needs from the Pareto optimal solution set Pareto_Set; determine the corresponding optimal internal structure parameter combination X_opt according to the optimal solution;

[0029] Step five two: design and manufacture bearing samples according to the parameter values of X_opt; verify whether the actual anti-gluing performance of the optimized bearing meets the expected target through strict bench test.

[0030] Further, the target bearing type in step one is cylindrical roller bearing or angular contact ball bearing.

[0031] Further, the key internal structure parameters that affect gluing failure in step one are: inner ring raceway curvature radius coefficient fi, outer ring raceway curvature radius coefficient fe, rolling element diameter Dw, rolling element number Z, cage pocket gap δc, cage guide gap δg, roller modification parameters such as logarithmic curve modification amount Cp and modification length Lp, surface roughness Ra, coating type and thickness Tc.

[0032] Further, in step two, multiple typical service conditions are set: constant load F_const, amplitude F_amp and frequency f_load in variable load, constant speed N_const, acceleration a_acc or deceleration a_dec in acceleration and deceleration, lubricating oil viscosity η, oil supply rate Q, pollution level and environmental temperature T_env.

[0033] Further, the numerical simulation in step 131 is: combined with elastohydrodynamic lubrication theory, thermal analysis or micro contact model, and the simulation needs to establish a bearing model containing detailed internal structure, and consider the thermal elastohydrodynamic lubrication effect and surface interaction.

[0034] Further, the gluing failure feature data in step 211 includes:

[0035] The minimum load at which the bearing begins to gluing under this working condition, that is, the critical load of gluing F critical ;

[0036] The minimum speed at which the bearing begins to gluing under this working condition, that is, the critical speed of gluing N critical ;

[0037] The increment of torque at the moment of gluing relative to stable operation, the friction torque mutation ΔT when gluing occurs;

[0038] The proportion of the area of gluing damage to the total contact area is A ratio ;

[0039] The time from starting operation to gluing is t failure .

[0040] Further, the calculation formula of the gluing risk factor γ in step 2 is in the form of weighted sum:

[0041]

[0042] Wherein, F ref , N ref , ΔT ref , t ref are reference values set according to bearing type and industry standard;

[0043] k1, k2, k3, k4, k5 are weight coefficients reflecting the importance of each feature, and satisfy Σki=1; The larger the value of γ, the higher the risk of gluing under the combination of parameters and working conditions.

[0044] Further, the process of defining a multi-objective optimization problem in step 411 is:

[0045] Minimizing the gluing risk factor γ, maximizing the bearing rated dynamic load C, and maximizing the bearing operation stability index St are multi-objectives.

[0046] Further, the solving process of the Pareto optimal solution set Pareto_Set in step 444 is:

[0047] (1) Adaptive inertia weight w: adopt cosine decreasing strategy:

[0048] wherein, w max , w min are the maximum and minimum inertia weight respectively; t is the current iteration number; T is the maximum iteration number;

[0049] (2) Dynamic learning factor: the individual learning factor c1 and the population learning factor c2 are updated using an exponential decay strategy:

[0050]

[0051]

[0052] wherein, c 1max , c 2max are the maximum values of c1 and c2 respectively; c 1min , c 2min are the minimum values of c1 and c2 respectively;

[0053] (3) After the algorithm is run, a set of non-dominated solutions, i.e. the Pareto optimal solution set Pareto_Set, is obtained.

[0054] Compared with the prior art, the present application has the following effects:

[0055] 1. By constructing the comprehensive quantitative index of the "bonding risk factor γ", the present application integrates multiple key features of bonding failure into a single calculable value, so that the designer can quantitatively compare the bonding risk under different parameter combinations in the design stage, thereby breaking away from excessive dependence on experience, realizing the transition from qualitative judgment to quantitative decision, and significantly improving the scientificity and reliability of the design.

[0056] 2. The present application integrates key internal structural parameters such as raceway geometry, rolling element parameters, cage clearance, roller modification parameters and surface characteristics, and takes them as design variables for collaborative optimization, overcoming the limitations of isolated parameter adjustment in traditional methods. Based on the multi-objective optimization algorithm, the interaction between parameters is revealed, and the systematic optimization of the bearing internal structure is realized, fundamentally improving the comprehensiveness and effectiveness of the anti-bonding design based on the bearing internal structural parameters.

[0057] 3. In the optimization process, the present application takes the minimization of the bonding risk factor γ, the maximization of the rated dynamic load C and the operation stability index St as optimization objectives at the same time, and solves the Pareto optimal solution set by the improved multi-objective optimization algorithm. This method can automatically find the best balance point between multiple conflicting objectives, ensuring that the optimized bearing achieves comprehensive optimization in terms of anti-bonding performance, load capacity and running stability.

[0058] 4. By establishing a basic database of "structural parameters, service conditions, and bonding risk factors" and combining it with high-precision numerical simulation technology, this invention can accurately predict bonding behavior under different parameter combinations during the design phase. This method significantly reduces reliance on physical testing, enabling designers to quickly evaluate multiple design schemes in a virtual environment, significantly shortening the R&D cycle, reducing development costs, and improving the foresight and controllability of the design.

[0059] 5. The method framework proposed in this invention has high flexibility and scalability. By adjusting the set of key parameters and operating conditions, it can be applied to the anti-galling design of various rolling bearings under different operating conditions. This universality makes the method widely applicable in engineering and has great potential for widespread application.

[0060] 6. To address the shortcomings of standard multi-objective optimization algorithms, this invention employs two key improvements: cosine-decreasing adaptive inertia weights and exponentially decaying dynamic learning factors. These strategies enhance the algorithm's global exploration capabilities in the early stages of iteration, while focusing on refined local search in later stages, effectively balancing the exploration and development process and improving convergence speed and solution set quality.

[0061] In summary, this invention effectively solves the problems of reliance on old experience, isolated internal parameters, difficulty in quantifying risks, and excessively high testing costs in traditional bearing anti-galling design through a series of technological innovations, such as quantifying risk indicators, integrating system parameters, multi-objective collaborative optimization, combining simulation and database technologies, and improving algorithms. It provides a systematic and scientific design method for the reliable operation of bearings under harsh working conditions. Attached Figure Description

[0062] Figure 1 This is an overall flowchart of the bearing anti-adhesion design method based on internal structural parameter optimization of the present invention. Detailed Implementation

[0063] Specific implementation method one: Combining Figure 1 This embodiment describes the following steps.

[0064] Step 1: Build the basic database:

[0065] Step 11: Select the target bearing type and determine the set of key internal structural parameters S = {S1, S2, ..., Sm} that affect the adhesion failure; where Si represents the i-th structural parameter and m is the total number of parameters;

[0066] Steps 1 and 2: Define a set of multiple typical service conditions W = {W1, W2, ..., Wn}; where Wj represents the j-th service condition and n is the number of service conditions;

[0067] Step one three: for each bearing structure parameter and service condition, obtain the bonding failure characteristic data of the bearing through physical bench test or numerical simulation;

[0068] Step two: calculate the bonding risk factor:

[0069] According to the bonding failure characteristic data obtained in step one, the bonding risk factor γ under the corresponding parameter combination and working condition is calculated; wherein, the bonding risk factor γ is defined as the tendency or risk level of the bearing to occur bonding failure under the specific condition;

[0070] Step three: establish the association database:

[0071] The bearing internal structure parameter combination S in step one, the service condition W and the bonding risk factor γ calculated in step two are associated, and the structure parameter-service condition-bonding risk factor basic database DB_base is established;

[0072] Step four: parameter optimization design:

[0073] Step four one: set the optimization goal: define the multi-objective optimization problem;

[0074] Step four two: define the design variables:

[0075] Select a subset from the key internal structure parameter set S which has significant influence on bonding and can be optimized and adjusted As the design variable;

[0076] Step four three: set the constraint condition:

[0077] S1: design variable range constraint: each design variable Xi needs to meet its engineering feasible range Xi_min ≤ Xi≤ Xi_max;

[0078] S2: bearing basic performance constraint: static safety factor S0 ≥ S0_min, basic rated life L10 ≥ L10_req;

[0079] S3: manufacturing process constraint: machining precision accessibility, cost limit;

[0080] Step four four: apply multi-objective optimization algorithm to solve:

[0081] Under the condition of meeting the constraint condition, the design variable X is optimized and solved; a set of non-dominated solutions, i.e. the Pareto optimal solution set Pareto_Set, is obtained, and each solution in the set represents a design scheme which achieves a good compromise among multiple objectives;

[0082] Step five: scheme selection and verification:

[0083] Step five: according to the specific needs of practical application, select a set of optimal solutions that best meet the needs from the Pareto optimal solution set Pareto Set; determine the corresponding optimal internal structure parameter combination X_opt according to the optimal solution;

[0084] Step five two: according to the parameter value of X_opt, design and manufacture bearing samples; through strict bench test, verify whether the actual anti-bonding performance of the optimized bearing reaches the expected target.

[0085] The application proposes a system, quantitative, internal structure parameter collaborative optimization based bearing anti-bonding design method, which can effectively evaluate the bonding risk under different structure parameter combinations in the design stage, and guide the designer to select the optimal parameter configuration, thereby improving the anti-bonding ability of the bearing from the source. The internal parameter collaborative optimization method of the application is also applicable to the optimization research of other characteristic index capabilities, and has a wide application prospect.

[0086] Specific implementation method two: combined with Figure 1 In this embodiment, the target bearing type in step one-one is a cylindrical roller bearing or an angular contact ball bearing.

[0087] These two types of bearings are prone to bonding failure under high-speed heavy load working conditions. The unique line contact stress characteristics make it easy to produce stress concentration at the edge of the contact area when the lubrication condition deteriorates, thereby causing local high temperature and material welding. Specifically, cylindrical roller bearings mainly bear radial load, and the line contact form of the rollers and the raceways is prone to edge stress at the end during high-speed operation; while angular contact ball bearings bear radial and axial loads at the same time, and the contact stress distribution is more complex, and the bonding risk is more prominent under variable working conditions. By focusing on these two types of high-risk bearings, the established bonding risk model is more targeted, and the obtained optimization parameters are more in line with the actual engineering needs. This targeted research not only solves the most urgent bonding failure prevention problem in the current bearing industry, but also lays a technical foundation for subsequent promotion to other types of bearings by verifying the effectiveness of the method on these two types of typical bearings.

[0088] Specific implementation method three: combined with Figure 1 In this embodiment, the key internal structure parameters that affect bonding failure in step one-one are: inner ring raceway curvature radius coefficient fi, outer ring raceway curvature radius coefficient fe, rolling element diameter Dw, rolling element number Z, cage pocket gap δc, cage guide gap δg, roller modification parameters such as logarithmic curve modification amount Cp and modification length Lp, surface roughness Ra, coating type and thickness Tc.

[0089] The key internal structural parameters affecting the gluing failure are explicitly listed in Step One, which aims to systematically build a core factor system affecting the bearing's anti-gluing performance. These parameters cover three key dimensions of bearing contact geometry, motion relationship, and surface state: the inner and outer ring raceway curvature radius coefficients fi and fe directly affect the lubricating oil film formation conditions and contact stress distribution; the rolling element diameter Dw and the number Z jointly affect the load distribution and sliding friction trend; the cage pocket gap δc and the guide gap δg affect the rolling element motion trajectory and collision energy, leading to local temperature rise; the roller modification parameters Cp and Lp affect the edge stress concentration, leading to damage risk; the surface roughness Ra directly affects the oil film state of the micro contact area; the coating type and thickness Tc directly affect the change of boundary lubrication state. These parameters not only have independent effects on gluing failure, but also jointly determine the bearing's anti-gluing performance through complex coupling. By incorporating these parameters into a unified optimization framework, designers can match the design from a system level, thereby fundamentally improving the bearing's anti-gluing ability.

[0090] Specific implementation method four: combination Figure 1 In this embodiment, the typical service conditions set in Steps One and Two are: constant load F const, amplitude F amp and frequency f load in variable load, constant speed N const, acceleration a acc or deceleration a dec in acceleration and deceleration, lubricating oil viscosity η, oil supply rate Q, pollution level and environmental temperature T env.

[0091] These typical service conditions are set in Step One in order to fully simulate the complex working conditions that bearings face in actual applications during the bearing design stage. The gluing failure of bearings is often not caused by a single steady-state factor, but is the result of the combined action of multiple dynamic working conditions. Constant load and speed represent the basic working condition when the equipment is running stably, while variable load and its frequency simulate the dynamic response of the equipment when it is subjected to periodic impact working conditions or working cycles. This alternating stress significantly affects the stability of the lubricating oil film in the contact surface. The acceleration and deceleration process changes the relative motion relationship between the rolling element and the raceway, directly affecting the instantaneous temperature rise in the contact area. The lubricating oil viscosity, oil supply rate and pollution level jointly constrain the actual working state boundary of the lubricating system, among which the pollution level directly relates to the damage degree of abrasive wear to the surface integrity. The environmental temperature serves as a thermal boundary condition, affecting the thermal equilibrium state of the entire bearing system. By systematically combining these working condition parameters, the gluing risk of bearings in different application scenarios can be comprehensively evaluated, and the adaptability of bearing structural parameters to complex working conditions in operation can be improved.

[0092] Specific implementation method five: combination Figure 1In this embodiment, the numerical simulation in step one three is as follows: a bearing model containing detailed internal structure is established by combining elastohydrodynamic lubrication theory, thermal analysis or micro contact model, and the thermal elastohydrodynamic lubrication effect and surface interaction are considered.

[0093] In step one, the numerical simulation method combining elastohydrodynamic lubrication theory, thermal analysis and micro contact model is adopted, and the core purpose is to accurately reproduce the process of bearing gluing damage in actual operation in a virtual environment. Gluing failure is essentially an extreme result caused by mechanics, heat and surface interaction together, and traditional simplified calculation cannot accurately summarize this complex phenomenon. By establishing a bearing model containing detailed internal structure, the geometric relationship and motion constraint between the rolling body and the raceway can be truly reflected. The introduction of elastohydrodynamic lubrication theory can accurately calculate the oil film thickness and pressure distribution in the contact area, and judge whether the lubrication state is good. The thermal analysis module can simulate the temperature field change caused by friction heat, and extract the local high temperature to trigger the risk state point of material surface welding. The above multi-physical field coupling simulation can evaluate the gluing failure risk of the bearing under different parameter combinations in the design stage, and provide a solid data basis for the subsequent establishment of a reliable gluing risk prediction model.

[0094] Specific implementation method six: combining Figure 1 In this embodiment, the gluing failure characteristic data in step two one of this embodiment includes:

[0095] The minimum load at which the bearing begins to gluing under this working condition, that is, the critical load F critical for gluing to occur;

[0096] The minimum rotational speed at which the bearing begins to gluing under this working condition, that is, the critical rotational speed N critical for gluing to occur;

[0097] The increment of torque at the moment of gluing relative to stable operation, that is, the friction torque mutation ΔT when gluing occurs;

[0098] The proportion of the area of gluing damage to the total contact area is A ratio ;

[0099] The time from the start of operation to the occurrence of gluing is t failure .

[0100] In step two, these gluing failure characteristic data are selected in order to quantitatively characterize the gluing failure process from multiple key dimensions. The critical load F critical and the critical rotational speed N criticalThe load and rotation speed threshold of bearing gluing are defined, which directly reflects the bearing load limit under specific working conditions. The friction torque jump ΔT captures the mechanical response characteristics at the moment of gluing, and its sharp change is a direct reflection of the surface material welding and tearing. The gluing area proportion A ratio quantifies the severity of failure and reveals the macroscopic extension range of damage. The gluing time t failure records the duration from normal operation to failure, reflecting the durability of the bearing under harsh working conditions. These characteristic parameters together constitute the gluing failure evaluation system, providing comprehensive and reliable quantitative basis for subsequent construction of comprehensive gluing risk factor.

[0101] Specific embodiment seven: in combination Figure 1 In this embodiment, the calculation formula of the gluing risk factor γ in step two is in the form of weighted sum:

[0102] Wherein, F ref , N ref , ΔT ref , t ref are reference values set according to bearing type and industry standard;

[0103] k1, k2, k3, k4, k5 are weight coefficients reflecting the importance of each feature, and Σki=1; the larger the value of γ, the higher the risk of gluing under the combination of parameters and working conditions.

[0104] In step two, the calculation formula of gluing risk factor γ is in the form of weighted sum, which integrates multiple failure features with different dimensions and physical meanings into a unified and comparable comprehensive evaluation index. Gluing failure is a complex process affected by multiple factors, and single feature parameter cannot fully reflect its risk level. By introducing reference values set according to bearing type and industry standard, the feature data is first converted into dimensionless relative values, solving the problem of non-uniform dimension between different parameters. On this basis, each feature parameter is assigned a weight coefficient reflecting its importance, so that critical load, critical speed and other decisive factors can be reasonably integrated with torque jump, damage area and failure time, according to their contribution to gluing risk. This weighted sum method ensures the systematization and comprehensiveness of risk evaluation, and through the adjustment of coefficients, different application scenarios can be analyzed. The final value of γ as a single and clear quantitative index enables designers to intuitively compare the gluing risk of different design schemes, providing clear and reliable basis for subsequent optimization decision.

[0105] Specific embodiment eight: in combination Figure 1In this embodiment, the process of defining the multi-objective optimization problem in step four is as follows: taking the minimum bonding risk factor γ, the maximum bearing rated dynamic load C, and the maximum bearing operation stability index St as the multi-objectives.

[0106] In step four, the multi-objective optimization problem of minimizing the bonding risk factor γ, maximizing the bearing rated dynamic load C, and maximizing the operation stability index St is defined, and the ultimate purpose is to break through the limitation of the traditional design criterion of maximizing the calculated life and seek the overall optimal solution of the comprehensive performance of the bearing. In actual engineering, these key performance indicators often restrict each other: simply pursuing low bonding risk may lead to a decrease in load-carrying capacity; overemphasizing high load capacity may affect the operation stability. By simultaneously incorporating these three core objectives into the optimization system, the optimization algorithm can weigh and compromise between multiple contradictory design requirements. This multi-objective optimization mechanism can systematically explore the entire design space and automatically find design solutions that achieve the best balance between bonding resistance, load-carrying efficiency, and operation stability. The final Pareto optimal solution set provides a series of "non-inferior solutions" for designers, each representing a feasible design solution that reasonably balances different performance indicators, ensuring that the optimized bearing not only has excellent bonding resistance but also has sufficient load-carrying capacity and good operation stability, meeting the complex demands of actual applications.

[0107] DETAILED DESCRIPTION NINE: COMBINATION Figure 1 In this embodiment, the process of solving the Pareto optimal solution set Pareto_Set in step four is as follows:

[0108] (1) Adaptive inertia weight w: cosine decreasing strategy is adopted:

[0109]

[0110] wherein, w max , w min are the maximum and minimum inertia weights, respectively; t is the current iteration number; T is the maximum iteration number;

[0111] (2) Dynamic learning factor: individual learning factor c1 and group learning factor c2 are updated using an exponential decay strategy:

[0112]

[0113]

[0114] wherein, c 1max , c 2maxmax(c1, c2) respectively; c 1min max(c1, c2) respectively; c 2min min(c1, c2) respectively; c

[0115] (3) After the algorithm runs, a set of non-dominated solutions, i.e. a Pareto optimal solution set Pareto Set, is obtained.

[0116] The embodiment effectively balances the conflict between the design objectives through an intelligent optimization mechanism. The algorithm combines a cosine-decreasing adaptive inertia weight with an exponentially decaying dynamic learning factor to enhance the global exploration ability in the early iteration and avoid premature convergence, and to strengthen the local development precision in the later period and ensure stable convergence. This improvement significantly improves the search efficiency and solution set quality of the algorithm in the complex design space, and can systematically generate a series of design schemes that achieve the best balance between the anti-bonding performance, bearing capacity and running stability, providing sufficient basis for engineering design optimization.

[0117] In combination with Figure 1 The embodiments of the present application are described:

[0118] A bearing anti-bonding design method based on internal structure parameter optimization, comprising the following steps:

[0119] Step 1, build a basic database: select the target bearing type (such as cylindrical roller bearing, angular contact ball bearing, etc.), determine the key internal structure parameter set S = {S1, S2, …, Sm} affecting the bonding failure (for example: inner ring raceway curvature radius coefficient fi, outer ring raceway curvature radius coefficient fe, rolling element diameter Dw, rolling element number Z, cage pocket gap δc, cage guide gap δg, roller modification parameters such as logarithmic curve modification amount Cp and modification length Lp, surface roughness Ra, coating type and thickness Tc, etc.). Set multiple typical service condition sets W = {W1, W2, …, Wn} (for example: constant load F const , variable load (amplitude F amp , frequency f load ), constant speed N, acceleration / deceleration (acceleration a_acc / deceleration a_dec), lubricating oil viscosity η, oil supply rate Q, pollution level, environment temperature T_env, etc.). For each bearing structure parameter combination (S combination) and service condition (Wj), the bonding failure characteristic data of the bearing is obtained through physical bench test (such as FZG test machine modification, special bearing bonding test bench) or high-precision numerical simulation (combined with EHL, thermal analysis, micro contact model). The simulation needs to establish a bearing model containing detailed internal structure, considering the thermal elastohydrodynamic lubrication effect and surface interaction.

[0120] Step 2: Calculate the scuffing risk factor: Based on the scuffing failure characteristic data obtained in Step 1, calculate the scuffing risk factor γ under the corresponding parameter combinations and operating conditions. γ is a dimensionless or physically meaningful quantitative index used to comprehensively characterize the tendency or risk level of bearing scuffing failure under specific conditions. Characteristic data may include: the critical load F at which scuffing occurs. critical (Minimum load at which bearings begin to scuff under this operating condition), critical speed N critical (The lowest speed at which bearings begin to scuff under this operating condition), the sudden change in frictional torque ΔT at the time of scuffing (the increase in torque at the instant of scuffing relative to stable operation), and the proportion of the scuffed area A. ratio (The proportion of the area where adhesive damage occurred to the total contact area), and the time t for adhesive damage to occur. failure (Time from start of operation to the occurrence of adhesion). The formula for calculating the adhesion risk factor γ can be designed as a weighted sum:

[0121]

[0122] Among them, F ref N ref ΔT ref , t ref These are reference values ​​set according to bearing type and industry standards; k1, k2, k3, k4, and k5 are weighting coefficients reflecting the importance of each characteristic, and satisfy Σki=1. The larger the γ value, the higher the risk of seizure under this parameter combination and operating condition.

[0123] Step 3: Establish a relational database: Associate the bearing internal structural parameter combination S and service condition W from Step 1 with the galling risk factor γ calculated in Step 2, and establish a basic database DB_base for structural parameters, service conditions, and galling risk factors. This database forms the basis for subsequent optimization design.

[0124] Step 4, Parameter Optimization Design:

[0125] a) Set optimization objectives: Define a multi-objective optimization problem. The primary objective is to minimize the adhesion risk factor γ. Simultaneously, the basic performance requirements of the bearing must also be considered, such as maximizing the bearing's rated dynamic load C (reflecting its load-carrying capacity) and maximizing the bearing's operational stability index St (e.g., maximizing the minimum oil film thickness h_min to improve lubrication, minimizing the maximum operating temperature T_max to control temperature rise, minimizing vibration intensity V_rms, etc.). The objective function vector can be represented as: Minimize [γ, -C, -St] or an equivalent form.

[0126] b) Define design variables: select a subset X ⊆ S from the set of key internal structure parameters S that significantly affect the bonding and can be optimized as design variables (e.g. X = {fi, fe, Dw, Z, δc, Cp, Lp, Ra, Tc}).

[0127] c) Set constraints: including:

[0128] Design variable range constraints: each design variable Xi needs to satisfy its engineering feasible range Xi_min ≤ Xi ≤ Xi_max (e.g. fi between 0.51~0.54).

[0129] Bearing basic performance constraints: such as static safety factor S0 ≥ S0_min, basic rated life L10 ≥ L10_req, etc.

[0130] Manufacturing process constraints: such as machining precision accessibility, cost limitations, etc.

[0131] d) Apply improved multi-objective optimization algorithm for solution: use the improved multi-objective particle swarm optimization (MOPSO) algorithm or other advanced multi-objective algorithms to optimize the design variables X under the constraints. The standard MOPSO algorithm is improved in this application to improve convergence and avoid premature:

[0132] Adaptive inertia weight w: use cosine decreasing strategy:

[0133]

[0134] where, w max (0.9, for example), w min (0.4, for example) are the maximum and minimum inertia weights, respectively; t is the current iteration number; T is the maximum iteration number. This strategy can smoothly transition global exploration and local development.

[0135] Dynamic learning factor: individual learning factor c 1 and group learning factor c 2 Use exponential decay strategy to update:

[0136]

[0137]

[0138] where, 1max , 2max (2.5, for example, c 1max =2.5, c 2max =2.5) are the maximum values of c1 and c2, respectively; c1min , c 2min (like c 1min =0.5, c 2min =0.5) are the minimum values of c1, c2, respectively. This strategy makes the algorithm focus on individual experience (exploration) at the beginning and focus on population experience (convergence) at the end.

[0139] After the algorithm runs, a set of non-dominated solutions, i.e. the Pareto optimal solution set Pareto_Set, is obtained. Each solution in this set represents a design scheme that achieves a good trade-off among multiple objectives.

[0140] Step five, scheme selection and verification: according to the specific requirements of practical application (such as paying more attention to anti-bonding, or bearing capacity, or temperature control), select a set of optimal solutions from the Pareto optimal solution set Pareto_Set that best meet the requirements. Determine the corresponding optimal internal structure parameter combination X_opt according to the optimal solution. According to the parameter values of X_opt, design and manufacture bearing samples. Finally, verify whether the actual anti-bonding performance of the optimized bearing meets the expected target through strict bench tests (simulating actual or more severe working conditions).

[0141] Through the above steps, the present application provides a bearing anti-bonding design method based on internal structure parameter optimization. This method establishes a quantitative correlation model between bearing internal structure parameters, service conditions and bonding risk factors, and applies an improved multi-objective optimization algorithm to realize the collaborative optimization of key structure parameters, thereby significantly reducing the bonding risk and improving the reliability of bearings under harsh working conditions.

[0142] The above examples are only used to illustrate the technical solutions of the present application, but not to limit it; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing examples, or make equivalent substitutions for part of the technical features; and these modifications or substitutions do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A bearing anti-galling design method based on internal structural parameter optimization, characterized in that: It includes the following steps Step 1: Build the basic database: Step 11: Select the target bearing type and determine the set of key internal structural parameters S = {S1, S2, ..., Sm} that affect the adhesion failure; where Si represents the i-th structural parameter and m is the total number of parameters; Steps 1 and 2: Define a set of multiple typical service conditions W = {W1, W2, ..., Wn}; where Wj represents the j-th service condition and n is the number of service conditions; Step 13: For each bearing structural parameter and service condition, obtain bearing adhesion failure characteristic data through physical bench tests or numerical simulations; Step 2: Calculate the bonding risk factor: Based on the adhesive failure characteristic data obtained in step one, calculate the adhesive risk factor γ under the corresponding parameter combination and working condition; whereby the adhesive risk factor γ is defined as the tendency or risk level of the bearing to undergo adhesive failure under this specific condition. Step 3: Establish a relational database: The bearing internal structural parameter combination S and service condition W in step one are correlated with the galling risk factor γ calculated in step two, and a basic database DB_base of structural parameters-service condition-gag risk factor is established. Step 4: Parameter Optimization Design Step 41: Set optimization objectives: Define a multi-objective optimization problem; Step 42: Define design variables: Select a subset from the set of key internal structural parameters S that has a significant impact on bonding and can be optimized. As a design variable: Step 43: Set constraints: S1: Design variable range constraint: Each design variable Xi must satisfy its engineering feasible range Xi_min ≤ Xi ≤ Xi_max; S2: Basic bearing performance constraints: Static safety factor S0 ≥ S0_min, basic rated life L10 ≥ L10_req; S3: Manufacturing process constraints: Achievable machining accuracy, cost limitations; Step 44: Solve using a multi-objective optimization algorithm: Under the constraints, the design variable X is optimized and solved; a set of non-dominated solutions is obtained, namely the Pareto optimal solution set Pareto_Set, where each solution represents a design scheme that achieves a good trade-off among multiple objectives. Step 5: Solution Selection and Verification Step 51: Based on the specific requirements of the actual application, select a set of optimal solutions that best meet the requirements from the Pareto optimal solution set Pareto_Set; determine the corresponding optimal combination of internal structure parameters X_opt based on the optimal solution; Step 52: Based on the parameter values ​​of X_opt, design and manufacture bearing prototypes; verify whether the actual anti-galling performance of the optimized bearing meets the expected target through rigorous bench testing.

2. The bearing anti-galling design method based on internal structural parameter optimization according to claim 1, characterized in that: The target bearing type in step one is a cylindrical roller bearing or an angular contact ball bearing.

3. The bearing anti-galling design method based on internal structural parameter optimization according to claim 2, characterized in that: The key internal structural parameters affecting adhesive failure in step one are: Inner ring raceway curvature radius coefficient fi, outer ring raceway curvature radius coefficient fe, rolling element diameter Dw, number of rolling elements Z, cage pocket clearance δc, cage guide clearance δg, roller shaping parameters such as logarithmic curve shaping amount Cp and shaping length Lp, surface roughness Ra, coating type and thickness Tc.

4. The bearing anti-galling design method based on internal structural parameter optimization according to claim 3, characterized in that: In steps one and two, several typical service conditions are set as follows: Constant load F_const, amplitude F_amp and frequency f_load in variable load, constant speed N_const, acceleration a_acc or deceleration a_dec in acceleration and deceleration, lubricating oil viscosity η, oil supply rate Q, pollution level and ambient temperature T_env.

5. The bearing anti-galling design method based on internal structural parameter optimization according to claim 4, characterized in that: The numerical simulation in steps one and three involves combining elastohydrodynamic lubrication theory, thermal analysis, or microscopic contact models. The simulation requires establishing a bearing model that includes a detailed internal structure and considering the thermoelastohydrodynamic lubrication effect and surface interaction.

6. The bearing anti-galling design method based on internal structural parameter optimization according to claim 5, characterized in that: The adhesive failure characteristic data in step 2.1 include: The minimum load at which bearings begin to scuff under this operating condition, i.e., the critical load F for scuffing to occur. critical ; The minimum speed at which bearings begin to scuff under this operating condition, i.e., the critical speed N for scuffing. critical ; The increment of the instantaneous torque during adhesive bonding relative to stable operation; the sudden change in frictional torque ΔT at the time of adhesive bonding. The proportion of the area where adhesive damage occurred to the total contact area is A. ratio ; The time from the start of operation to the occurrence of bonding is t. failure .

7. The bearing anti-galling design method based on internal structural parameter optimization according to claim 6, characterized in that: The formula for calculating the adhesion risk factor γ in step two is a weighted sum: Among them, F ref N ref ΔT ref , t ref These are reference values ​​set according to bearing type and industry standards; k1, k2, k3, k4, and k5 are weighting coefficients reflecting the importance of each feature, and satisfy Σki=1; the larger the γ value, the higher the risk of adhesive bonding under this parameter combination and operating condition.

8. The bearing anti-galling design method based on internal structural parameter optimization according to claim 7, characterized in that: Step 4.1 defines the process of defining a multi-objective optimization problem as follows: The objectives are to minimize the adhesion risk factor γ, maximize the rated dynamic load C of the bearing, and maximize the bearing operating stability index St.

9. The bearing anti-galling design method based on internal structural parameter optimization according to claim 8, characterized in that: The process of solving for the Pareto optimal solution set Pareto_Set in step four is as follows: (1) Adaptive inertia weight w: adopts a cosine decreasing strategy: in, w max , w min These are the maximum and minimum inertia weights, respectively; t This represents the current iteration number; T This represents the maximum number of iterations. (2) Dynamic learning factors: Individual learning factor c1 and group learning factor c2 are updated using an exponential decay strategy. Among them, c 1max c 2max The maximum values ​​of c1 and c2 are respectively; c 1min c 2min Let c1 and c2 be the minimum values, respectively. (3) After the algorithm runs, a set of non-dominated solutions is obtained, namely the Pareto optimal solution set Pareto_Set.

Images