A method for optimizing a bearing for wind power

Abstract

The application relates to a kind of wind power bearing optimization method, belong to bearing optimization technical field.The application first obtains the bearing parameter that fatigue life meets the stipulation, then considers the deformation condition of inner ring raceway, obtains the bearing parameter that is more actual respectively in and under the contact stress of corresponding roller and inner ring raceway with three accurate contact stresses in this place, with the three accurate contact stresses as the target, adjust the roller surface finish amount in roller parameter, realize the optimization of wind power bearing, and the error of obtained bearing parameter, the contact stress of roller and inner ring raceway corresponding to the bearing parameter and fatigue life is smaller, to further improve the accuracy of bearing optimization result.

Description

Technical Field

[0001] This invention relates to an optimization method for wind power bearings, belonging to the field of bearing optimization technology. Background Technology

[0002] Currently, wind power technology is developing rapidly worldwide. To meet the demands of offshore wind power applications, wind turbines are becoming increasingly larger, currently exceeding 20MW. Compared to onshore wind energy resources, offshore wind energy resources are more abundant, resulting in larger and longer-lasting wind loads on offshore wind turbines. Consequently, offshore wind turbines are relatively larger in size, placing higher demands on the long-term safety performance of turbine components.

[0003] The wind turbine main shaft is a large rotating shaft connecting the impeller hub and the gearbox generator. The front end of the main shaft bearing needs to bear the complex multi-degree-of-freedom loads generated by the rotation of the front blades, while the rear end of the bearing needs to withstand the weight load of the gearbox generator and the unstable impact loads of the entire system. Due to the large size of the wind turbine and the limitations of its installation location, the cost of bearing maintenance and replacement is very high. Therefore, the reliable and stable operation and long life of the bearings play a crucial role in the overall operation of the wind turbine. To ensure the reliable and stable operation and long life of the bearings, in addition to detailed calculations of the internal parameters of the bearings during the design phase, many other factors need to be considered when verifying the bearing performance.

[0004] Chinese invention patent application CN107729597A, published on February 23, 2018, discloses a spindle bearing raceway verification tool. This tool can directly perform equivalent processing on time-series loads and LDD load spectra, specifically applying multiple sets of loads as a single load to the hub to calculate the bearing load. Then, it calculates the bearing life and contact stress according to standards. However, the bearing load calculated by applying only one set of loads has a large error compared to the actual load. This large error in the calculated bearing load results in relatively low accuracy for the bearing life and contact stress calculated using ISO standard methods, and also relatively low accuracy for bearing performance evaluations based on experience. In other words, the bearing load, bearing life, and contact stress obtained through this method all deviate significantly from reality. Furthermore, the calculation process is based on a rigid body, so the calculated results are likely to be larger than the actual engineering results. If calculated based on a flexible body, the accuracy of the results would be even worse. This solution can serve as a reference for general engineering projects, but for more complex marine conditions and for wind turbines that require higher safety and stability, more precise and rigorous algorithms are needed.

[0005] Chinese invention patent application CN107590356A, published on January 16, 2018, discloses an automatic selection method and storage device for main shaft bearings of wind turbine generators. This method establishes a bearing database and evaluates bearings based on calculated bearing loads. While this method can quickly select and evaluate bearings, it is only applicable to bearing application units because it lacks bearing optimization capabilities and is not suitable for optimizing existing bearings or designing new bearings.

[0006] Chinese invention patent application CN107688716A, published on February 13, 2018, discloses a parameter optimization method for hollow cylindrical roller bearings based on load distribution and fatigue life. This optimization method mainly targets the parameter optimization of hollow cylindrical roller bearings whose fatigue life is less than the expected fatigue life. It only gives the conditions for meeting the bearing optimization, but does not provide the optimization-related schemes.

[0007] In summary, given the complex offshore operating conditions and the trend towards larger wind turbines, an optimized bearing method that ensures reliable and stable operation and long service life has become an industry requirement.

[0008] Currently, bearing optimization can be achieved using finite element method (FEM) simulation software. While FEM-based bearing optimization schemes can consider the deformation effects of bearing interference fits on the bearing races and the expansion deformation caused by temperature rise during bearing operation, they cannot handle the calculation of multiple sets of multi-degree-of-freedom loads. Even calculating only one set requires a very long time due to the complexity of the wind turbine model and the bearing, making it inefficient to obtain the final optimization results. Therefore, it is not suitable for rapid iteration of engineering design schemes. Furthermore, when optimizing bearings based on only one or a few load calculations, the actual operating conditions retained by these loads are limited, failing to fully recreate the actual operating conditions of the bearing. This results in significant deviations between the optimized bearing parameters and the actual conditions.

[0009] While transmission simulation software such as Romaxwind can calculate multiple sets of multi-degree-of-freedom loads and quickly calculate bearing life and contact stress, its bearing model calculations do not consider the influence of internal deformation, resulting in a large deviation between the bearing parameters optimized in this way and the actual values. Summary of the Invention

[0010] The purpose of this invention is to provide an optimization method for wind turbine bearings, in order to solve the problem that the accuracy of existing optimization results for wind turbine bearings is relatively low.

[0011] To achieve the above objectives, the present invention includes:

[0012] An optimization method for wind turbine bearings according to the present invention includes the following steps:

[0013] By applying the ultimate load spectrum and fatigue load spectrum to the bearing model constructed based on the initial parameters of the bearing, the maximum load on the bearing rollers under the ultimate load spectrum is obtained. Fatigue life and fatigue stress of bearings under fatigue load spectrum;

[0014] The initial parameters were adjusted to obtain the optimized parameters with the goal of meeting the fatigue life requirements.

[0015] The inner raceway parameters after deformation are determined based on the inner raceway parameters in the optimized parameters; the times corresponding to the load steps in the fatigue load spectrum are sorted according to the magnitude of their corresponding fatigue stresses, and the maximum roller loads corresponding to the first and last fatigue load spectra corresponding to the specified proportion of the total time sorted from largest to smallest fatigue stress or the specified proportion of the total time sorted from smallest to largest fatigue stress are determined. ;

[0016] The calculations were performed based on the roller parameters and the deformed inner raceway parameters in the optimized parameters. The contact stress between the bearing roller and the inner ring raceway. ;

[0017] by With the goal of meeting the corresponding regulations, the roller surface trimming amount in the roller parameters is adjusted to obtain the final parameters of the bearing.

[0018] Furthermore, the parameters of the deformed inner raceway include the diameter and surface shape of the inner raceway, which are calculated based on the slicing method theory.

[0019] Furthermore, All results were obtained based on the finite-length line contact theory.

[0020] Furthermore, the bearing model adopts a flexible simulation model.

[0021] Furthermore, the optimization steps were implemented using transmission simulation software.

[0022] Furthermore, the deformation factors of the inner ring raceway include the bearing interference fit and the temperature rise during bearing operation.

[0023] Furthermore, the stipulated percentage is 20%.

[0024] Furthermore, the number of equivalent groups for the ultimate load spectrum ranges from 10 to 25.

[0025] Furthermore, the number of equivalent groups in the fatigue load spectrum ranges from 100 to 500.

[0026] The beneficial effects of this invention are:

[0027] This invention is groundbreaking, providing an optimization method for wind turbine bearings. The method first constructs a bearing model based on the bearing's initial parameters, and then applies ultimate load and fatigue load spectra to this model to obtain the maximum load on the bearing's rollers under actual extreme operating conditions. By adjusting the initial parameters of the bearing and considering the fatigue life and fatigue stress of the bearing under actual fatigue conditions, the optimized parameters of the bearing that meet the bearing life requirements under the aforementioned actual conditions are obtained.

[0028] Secondly, in order to make the optimized parameters of the bearing more in line with reality, the deformation of the inner ring raceway under actual working conditions is considered, and the parameters of the deformed inner ring raceway are determined.

[0029] The times corresponding to the load steps in the fatigue load spectrum are sorted according to the magnitude of the fatigue stress corresponding to that fatigue load spectrum. Based on the principle that the time exceeding the industry-specified limit for fatigue stress does not exceed a specified proportion (specified percentage), the maximum roller load corresponding to the first fatigue load spectrum is selected from the total time sorted from largest to smallest, within the first specified proportion, or from the total time sorted from smallest to largest, within the second specified proportion. The maximum roller load corresponding to the last fatigue load spectrum ;

[0030] Finally, based on the roller parameters and the deformed inner ring raceway parameters in the bearing parameters that meet the bearing life requirements, the following calculations were performed: Contact stress between the lower bearing roller and the inner ring raceway , Contact stress between the lower bearing roller and the inner ring raceway and Contact stress between the lower bearing roller and the inner ring raceway ,by , and With the goal of meeting the corresponding regulations, the roller surface trimming amount in the roller parameters is adjusted to obtain the final parameters of the bearing as the optimization result.

[0031] This invention first obtains bearing parameters that meet the specified fatigue life, and then considers the deformation of the inner ring raceway to obtain bearing parameters that are more realistic. , and The contact stress between the roller and the inner ring raceway is determined. With the goal of ensuring that the three accurate contact stresses meet the corresponding specifications, the roller surface trimming amount in the roller parameters is adjusted to optimize the bearing. The resulting bearing parameters, the contact stress between the roller and the inner ring raceway corresponding to these bearing parameters, and the fatigue life have smaller errors, thereby improving the accuracy of the bearing optimization results. Attached Figure Description

[0032] Figure 1 This is an optimization flowchart for bearings used in wind power.

[0033] Figure 2 This is a schematic diagram of a flexible simulation model;

[0034] Figure 3 This is a schematic diagram of the coordinate system and model of the inner ring raceway of the bearing;

[0035] Figure 4 It is a graph showing the contact stress before and after the roller surface of the bearing is optimized.

[0036] Explanation of reference numerals in the attached figures:

[0037] 1. Wind load; 2. Bearing A; 3. Bearing B; 4. Gearbox gravity. Detailed Implementation

[0038] To address the problems in the background art, this invention, while reproducing the actual working conditions of the bearing, also considers the impact of bearing deformation on contact stress and fatigue life during operation. Based on a load spectrum and bearing parameters that are more closely aligned with actual working conditions, the bearing parameters are optimized to improve the accuracy of the bearing optimization results.

[0039] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0040] An embodiment of an optimization method for wind turbine bearings:

[0041] An optimization method for wind turbine bearings includes the following steps:

[0042] A bearing model was constructed based on the initial parameters of the bearing. The ultimate load spectrum and fatigue load spectrum were then applied to the bearing model to obtain the maximum load on the bearing rollers under the ultimate load spectrum. Fatigue life and fatigue stress of bearings under fatigue load spectrum.

[0043] With the goal of ensuring the bearing's fatigue life meets the specified requirements, the initial parameters of the bearing are adjusted to obtain the optimized parameters for the bearing that meet the fatigue life requirements.

[0044] The parameters of the deformed inner raceway are determined based on the inner raceway parameters in the optimized parameters.

[0045] The total time is obtained by sorting the time corresponding to each load step in the fatigue load spectrum according to the magnitude of the corresponding fatigue stress. The maximum roller load corresponding to the first fatigue load spectrum with the predetermined proportion of fatigue stress in the total time (sorted from largest to smallest) is then determined. The maximum roller load corresponding to the last fatigue load spectrum Alternatively, determine the maximum roller load corresponding to the first fatigue load spectrum that corresponds to the specified percentage of fatigue stress in the total time, sorted from smallest to largest. The maximum roller load corresponding to the last fatigue load spectrum .

[0046] about , It can also be expressed as:

[0047] The fatigue stresses are sorted in order of magnitude to the corresponding time steps in the fatigue load spectrum to obtain a time scale reflecting the fatigue stress-fatigue load spectrum. The maximum roller load corresponding to the first fatigue load spectrum in the pre-defined proportion of the fatigue stresses sorted from largest to smallest within this time scale is then determined. The maximum roller load corresponding to the last fatigue load spectrum Alternatively, determine the maximum roller load corresponding to the first fatigue load spectrum in the specified proportion of fatigue stress ordered from smallest to largest within that timescale. The maximum roller load corresponding to the last fatigue load spectrum .

[0048] The calculations were performed based on the roller parameters and the deformed inner raceway parameters in the optimized parameters. Contact stress between the lower bearing roller and the inner ring raceway , Contact stress between the lower bearing roller and the inner ring raceway and Contact stress between the lower bearing roller and the inner ring raceway ;

[0049] by , and With the goal of ensuring all parameters meet the corresponding requirements, the roller surface dressing amount in the roller parameters is adjusted to achieve the desired result. , and The corresponding bearing parameters are used as the final parameters of the optimized bearing.

[0050] Among them, the ultimate load spectrum and fatigue load spectrum can restore the actual operating conditions of the bearing.

[0051] The initial parameters, optimized parameters, and final parameters of a bearing all include internal parameters such as roller parameters, inner ring raceway parameters, outer ring raceway parameters, and cage parameters. Roller parameters include roller size, number of rollers, and roller surface finishing allowance; inner ring raceway parameters include the inner ring raceway cone angle; and outer ring raceway parameters include the outer ring raceway cone angle.

[0052] When the initial parameters of the bearing are the bearing parameters to be optimized, the final parameters of the bearing are used as the optimization result for the bearing used in wind power. When the initial parameters of the bearing are the bearing parameters to be designed, the final parameters of the bearing are used as the design result for the bearing to be designed.

[0053] Specifically, the inner raceway parameters before and after deformation also include the diameter and surface shape of the inner raceway. The inner raceway parameters after deformation are calculated based on the slicing method theory.

[0054] Specifically, , and All results were obtained based on the finite-length line contact theory.

[0055] Specifically, the bearing model adopts a flexible simulation model to provide fast and reliable calculation software to meet the rapid iteration requirements of engineering design.

[0056] Specifically, the optimization method steps are implemented using transmission simulation software.

[0057] Specifically, the deformation factors of the inner ring raceway include the bearing interference fit and the temperature rise during bearing operation.

[0058] Considering that the time during which fatigue stress exceeds the industry-specified limit does not exceed the specified percentage (specified proportion), and since the industry-specified proportion is 20%, we will take 20% as the specified proportion. There is also an existing industry-specified proportion of 15%, therefore, this proportion can also be taken as 15%.

[0059] To reproduce the extreme operating conditions that occur during bearing operation as accurately as possible, the equivalent number of groups of the ultimate load spectrum is in the range of 10 to 25.

[0060] To reproduce the fatigue conditions that occur during bearing operation as accurately as possible, the equivalent number of fatigue load spectra is in the range of 100 to 500.

[0061] This optimization method for wind turbine bearings is applied to the main shaft of wind turbines. Specifically, it employs two judgments to iteratively optimize bearing parameters. The first judgment determines the bearing's lifespan (fatigue life) to iteratively optimize the bearing parameters, primarily through simulation calculations. The second judgment determines the internal stress value of the bearing (internal stress refers to the contact stress between the bearing rollers and the inner raceway) to iteratively optimize the roller surface finishing allowance in the bearing parameters. This optimization considers the influence of bearing installation interference and operating temperature on the internal structure, and is achieved through theoretical calculations combined with computer-programmed numerical analysis. This invention optimizes bearing parameters through iterative optimization using these two judgments, ultimately resulting in a safe, stable, and long-life bearing suitable for the complex load conditions of wind turbine main shafts.

[0062] The optimization of bearings for wind power applications is explained in detail below with reference to the accompanying drawings:

[0063] like Figure 1 As shown, the specific implementation steps are as follows:

[0064] Step 1: Preliminary design of bearing parameters.

[0065] The preliminary design of the bearing is based on past experience. The bearing parameters mainly include the internal roller size, number of rollers, inner and outer ring cone angles, cage, etc.

[0066] This step relies on the experience of the technicians and affects the number of iterations and computation time in subsequent steps. Therefore, bearing technicians should possess design capabilities and prior experience.

[0067] Simultaneously, this step should initially determine the roller surface modification when setting the bearing's internal parameters. Typically, this is set as a certain proportion of the bearing's calculated rated load. Subsequent iterations of the roller parameters should be based on this step. The specific roller surface modification equation depends on the experience of the bearing design unit and may be a circular arc, a circular arc plus a straight line, a logarithmic equation, a logarithmic-like equation, or other modification curve types, and is not limited to symmetrical or asymmetrical modifications. However, roller surface modification is necessary, as it has a significant impact on the bearing's internal stress and service life during application. Therefore, this step cannot be omitted.

[0068] Specifically, during the preliminary bearing design, based on past experience in wind turbine main shaft bearing design, the internal roller dimensions, number of rollers, inner and outer ring cone angles, and cage are initially determined. Roller surface modification is based on the classic Lundberg logarithmic equation, with the modification load factor taken as 0.2 times the bearing's rated dynamic load. The specific modification equation expression can be found in relevant literature. This modification method is widely considered the best and most widely used; subsequent iterations of modification parameters can achieve better computational speed based on this method.

[0069] Step 2: Prepare the ultimate load spectrum and fatigue load spectrum.

[0070] Load equivalence is a crucial step in wind turbine verification. To accurately reflect actual operating conditions, the loads should retain as many characteristics as possible. Clearly, more equivalent load groups will result in a closer approximation to reality. However, computational resources are limited; more loads mean longer computation time and greater computing power. Therefore, this scheme equips the ultimate load spectrum with 16 groups, and the fatigue load spectrum with 100–500 groups. Specific equivalence methods can utilize the existing patent technology with authorization publication number CN117592221B, which will not be elaborated here. Compared to equipping with only one group, the optimization results of this invention will be more reliable.

[0071] Specifically, the ultimate load of this scheme is provided by the wind turbine manufacturer, and the fatigue load spectrum is equivalent to the data provided by the manufacturer, resulting in 421 equivalent sets.

[0072] Step 3: Establish a flexible simulation model of the spindle bearing and load the fatigue load spectrum and ultimate load spectrum respectively.

[0073] This step is completed using computer software. This solution uses 3D software to draw the hub, spindle, and other components; a flexible mesh is created using a meshing tool; the model is imported into the transmission simulation software Romaxwind to build the bearing model, and the bearing races are then meshed with a flexible mesh. The ultimate load and fatigue load from step 2 are input for flexible calculations. This step is not limited to this modeling method. Because the model is flexible, the results are more realistic than those of a rigid model.

[0074] The model has two bearings, both of which should be calculated. Since the routes are the same, this manual will only refer to them as "bearings" and will no longer distinguish between bearing A and bearing B. The methods described below are applicable to both bearings.

[0075] The flexible simulation model for this step, such as Figure 2 As shown, it includes wind load 1, bearing A2, bearing B3 and gearbox gravity 4. Bearing A2 mainly bears the complex multi-degree-of-freedom load of the front blade caused by the rotation of wind load 1, and bearing B3 mainly bears the weight load of the rear gearbox gravity 4.

[0076] This step is not limited to this modeling and simulation method. Any commercial simulation software that can provide flexible functionality, calculate more than a hundred loads, and output bearing life and stress results can be used for this solution.

[0077] Step 4: Obtain the bearing life results (fatigue life of the bearing) and fatigue stress under the fatigue load spectrum; obtain the maximum load of the bearing rollers under the ultimate load spectrum. .

[0078] In this step, the bearing life output must conform to the ISO standard. If the output is the number of stress cycles due to the simulation software, the life should be converted.

[0079] The number of fatigue stress groups and the number of fatigue life groups both depend on the equivalent number of fatigue load spectrum groups. That is, the equivalent number of fatigue load spectrum groups equals the number of fatigue stress groups equals the number of fatigue life groups. Since there are 100 to 500 fatigue load groups, there will also be 100 to 500 fatigue stress groups.

[0080] Fatigue life and fatigue stress are obtained based on the bearing parameters set in step 1 (including the surface modification of the bearing rollers). Although the surface modification has an impact on bearing life, and iterative optimization of the surface modification will be performed in later stages, its impact on bearing life will no longer be calculated.

[0081] In this step, the maximum load on the bearing rollers under the ultimate load is obtained. This actually includes both normal and torque loads on the roller surface, generated by the contact between the roller and the inner and outer raceways, respectively. Therefore, the maximum load on the bearing roller under ultimate load... The actual values ​​include four parameters: the contact force between the roller and the inner ring, the contact force between the roller and the outer ring, the torque between the roller and the inner ring, and the torque between the roller and the outer ring. All of these results are derived from the simulation calculations in step 3.

[0082] Step 5: Determine bearing life.

[0083] If the bearing life result meets the required value (specification or user), proceed to the next step; otherwise, return to step 1, adjust the bearing parameters until the bearing life result meets the requirements, and then output the roller parameters.

[0084] Since lifespan is the primary bearing performance factor, the initial iteration focuses on optimizing the main bearing parameters with lifespan as the target. After meeting the lifespan requirement, the internal stress values ​​are iterated upon, and the surface modification parameters of the internal rollers and raceways are optimized.

[0085] While simulation results can yield internal bearing stress values, to make the calculations more realistic and valuable, subsequent stress calculations will consider the impact of bearing interference fit on the bearing race deformation, as well as the expansion deformation caused by temperature rise during bearing operation. These factors cannot be considered in the current simulation calculations, meaning that the stress values ​​calculated in this scheme are initial stress values, and further calculations are needed to correct them.

[0086] While current finite element simulation software can account for interference and temperature rise deformation, it cannot handle the calculation of 100–500 sets of multi-degree-of-freedom loads. Even calculating just one set requires a very long time due to the complexity of the wind turbine model and bearings, making it unsuitable for rapid iteration in engineering design. Transmission simulation software such as Romaxwind, while capable of calculating 100–500 load spectra and quickly determining bearing life and stress, does not consider the influence of internal deformation in its bearing model calculations. Therefore, this proposal suggests a route that first uses simulation to calculate bearing life, followed by iterative optimization of internal bearing stress through theoretical calculations.

[0087] Specifically, after performing the aforementioned 421 sets of equivalent fatigue load spectrum calculations, the bearing life results are shown in Table 1 below. In Table 1, 'e' is in scientific notation; for example, 'e5' refers to 10 to the power of 5, and 'e6' refers to 10 to the power of 6. 'Hours' in Table 1 is the unit of life.

[0088] Table 1

[0089]

[0090] According to regulations, when L10m > 130000h (hours) and L10mr > 175000h, the bearing parameters are considered to meet the life requirements.

[0091] Based on the premise that the surface stress of the roller under the maximum load under ultimate fatigue does not exceed a certain limit, the limit here is specified as 2800MPa according to industry experience.

[0092] Under fatigue load, the stress time exceeding the limit value shall not exceed the required percentage X%. According to industry experience, the limit value here is specified as 1500MPa, and the required percentage X% is specified as 20%.

[0093] The surface stress of the roller bearing the maximum load under fatigue load shall not exceed the limit value, which, according to industry experience, is specified as 1650 MPa.

[0094] Step 6: Select the top X% of the loads in terms of time percentage.

[0095] Based on the results of step 4, the time scale reflecting the fatigue stress-fatigue load spectrum is obtained by sorting the load steps in the fatigue load spectrum in descending order of fatigue stress. The load spectrum sequence number corresponding to the first X% (first X% of time percentage) of the fatigue stress in descending order of fatigue stress in this time scale is 1...k.

[0096] It should be noted that this refers to the time percentage, not the group percentage. The aforementioned fatigue load spectrum is equivalent to 421 groups, and the time corresponding to each load step in the fatigue load spectrum should be calculated. Among them, the first X% of the time percentage is preferably 20%, so the number of load groups in the first 20% of the time percentage for bearing A is 117 groups.

[0097] Step 7: Obtain the maximum load of the bearing rollers corresponding to load spectrum numbers 1 and k. and .

[0098] Same as step 4, at this point and These represent four values: the contact force between the roller and the inner ring, the contact force between the roller and the outer ring, the torque between the roller and the inner ring, and the torque between the roller and the outer ring.

[0099] Step 8: Calculate the diameter and surface shape of the deformed raceway based on the slice method, taking into account the effects of bearing interference and temperature rise.

[0100] This section, based on the thick-walled circular ring theory, first slices the bearing rings and establishes the raceway diameter equation. Considering the effects of bearing interference and temperature rise, each slice is calculated individually to obtain the raceway diameter and surface shape after total deformation.

[0101] like Figure 3 As shown, establish the coordinate system of the inner ring of the tapered roller bearing. The equation for the raceway radius is as follows:

[0102]

[0103] The equation for the deformation of the inner raceway considering the interference effect is as follows:

[0104]

[0105] The equation for the deformation of the inner raceway considering the effect of temperature rise is as follows:

[0106]

[0107] If the coefficient of thermal expansion of the shaft is the same as that of the inner ring, then the equation for the raceway radius is updated as follows:

[0108]

[0109] If the coefficient of thermal expansion of the shaft is different from that of the inner ring, the interference fit needs to be adjusted. Perform the following calculations:

[0110]

[0111] Then use The equation for inner raceway deformation that replaces the interference effect is as follows: The values ​​are then calculated again, taking into account the interference effect of the inner raceway deformation, the effect of temperature rise, the thermal expansion coefficient of the shaft, and the raceway radius when the inner raceway is the same.

[0112] In the above formula, This is the raceway radius value. The x-coordinate value of the raceway cross section. The inner raceway diameter is the large end diameter. The horizontal length of the raceway. The raceway half-cone angle, This represents the change in raceway radius due to the interference effect. Inner diameter of the inner ring Interference quantity , These are the elastic moduli of the bearing inner ring and the shaft, respectively. , These are the elastic moduli of the bearing inner ring and the shaft, respectively. Let be the inner diameter of the shaft. This represents the deformation caused by the temperature rise of the inner raceway. The coefficient of thermal expansion of the inner ring material is... Operating temperature It is at room temperature (generally 20℃). The coefficient of thermal expansion of the shaft material is . This is the radius of the raceway after thermal expansion.

[0113] Since the coefficients of thermal expansion of the shaft and the outer ring are the same, the increase in interference is 0, so there is no need to calculate the effect of interference after temperature rise. If the coefficients of thermal expansion of the shaft and the inner ring are different, the interference after temperature rise needs to be recalculated, and the effect of interference needs to be calculated.

[0114] Specifically, with an interference fit of 1.15 mm and an operating temperature of 65 °C, the inner ring raceway of bearing A is calculated to deform from the small end to the large end by 0.435–0.465 mm, and the raceway half-cone angle changes by 0°32″. Therefore, the deformed raceway needs to have the deformation amount added to the y-value to obtain the radius value of the raceway slice.

[0115] Step 9: Based on the load , and Calculate the contact stress between the bearing roller and the inner ring raceway on a single roller surface.

[0116] This section considers the original surface shaping, the deformation in step 8, and the tilting gap caused by the contact torque, comprehensively calculating the contact gap and calculating the contact stress between the bearing roller and the inner ring raceway based on the finite-length line contact theory. This calculation process requires computer programming. There are various methods in the industry for calculating the contact stress between the bearing roller and the inner ring raceway; the calculation based on the finite-length line contact theory is one of them.

[0117] This technical solution calculates the contact stress according to the flexibility coefficient method equation as follows:

[0118] The contact area of ​​the two contacting bodies is divided into equal parts. The rectangular element has a half-length of 1. Half width is .

[0119] The equilibrium equation is:

[0120]

[0121]

[0122]

[0123] In the above formula, The compliance coefficient represents the element's flexibility. In unit The impact produced above; This represents the relative displacement between the two contacting bodies; The elastic modulus of the two contacting bodies; For unit Contact stress on; The equation for the surface morphology of the contact body; This represents the total load force. For unit In Geometric center in the direction; For unit In Geometric center in the direction; For unit In the unit Geometric center in the direction; For unit In the unit Geometric center in the direction; The profile function of the contact surface with respect to The derivative; The profile function of the contact surface with respect to The derivative of the equation. Solving the equilibrium equation requires numerical computation via programming. For details on the flexibility coefficient method, please refer to Ma Jiaju's 1992 paper "Roller Convex Design".

[0124] Specifically, this step involves developing a numerical calculation program using the mathematical software MATLAB.

[0125] Step 10: Determine the contact stress.

[0126] According to the set judgment conditions, the load Below, contact stress value The load shall not exceed the required value, which is preferably 2800 MPa; Below, contact stress value The load shall not exceed the required value, preferably 1650 MPa; Below, contact stress value The pressure should not exceed the required value, preferably 1500 MPa. If any condition is not met, adjust the roller surface profile, specifically by increasing or decreasing the profile. After adjustment, return to step 9 to reconstruct the clearance equation for calculation. If all conditions are met, the iteration ends.

[0127] Contact stress curves before and after ultimate load optimization, such as...Figure 4 As shown, after optimization, the contact stress is lower than the set 2800MPa.

[0128] Step 11: By following the above steps, the iterative design of the bearing is completed, and the bearing engineering drawings can be output.

[0129] The final optimized bearing design has been tested and verified, proving suitable for the engineering conditions described. The developed bearing has achieved good economic benefits. This design is applicable to the optimization of main shaft bearings for large wind turbines and lays the foundation for the future development of deep-sea wind turbines.

[0130] Through the above steps, bearing parameters are iteratively designed based on fatigue life results. Considering the deformation effects of bearing installation interference and temperature rise, the bearing contact stress is theoretically calculated based on ultimate load and fatigue load conditions. The roller surface trimming amount (roller surface profile trimming parameters) is then iteratively designed. Finally, the bearing optimization is completed. This scheme provides evaluation conditions for each iteration, considers more comprehensive factors, and designs a more reliable bearing. It overcomes the shortcomings of existing technologies for optimizing large wind turbine main shaft bearings, making the bearing optimization more reasonable and the results more consistent with engineering practice. It is suitable for the reliability design and evaluation of future large and super-large wind turbine main shaft bearings, and can provide technical support for the safe operation of future megawatt-class wind turbines.

[0131] This method uses a relatively large number of load spectra to reproduce the actual working conditions of the bearing as much as possible. For bearings that meet the bearing life requirements, it considers the influence of bearing installation interference on the deformation of the raceway and the influence of expansion deformation caused by temperature rise during bearing operation. This results in a more realistic bearing parameter corresponding to the contact stress between the roller and the inner ring raceway. Based on the accurate contact stress, the bearing is optimized. The resulting bearing parameters, the corresponding contact stress between the roller and the inner ring raceway, and the fatigue life have smaller errors, thus improving the accuracy of the bearing optimization results.

[0132] Because the technical solution of this invention involves numerous calculation processes, the implementation methods are not described in detail. For example, modeling and simulation techniques can be found in relevant professional modeling and simulation tools and books; deformation and stress calculations involve computer programming and numerical iterative calculations. Regarding the overall technical route, any adjustments to the order, calculation coefficients, or constraints made based on understanding, or the integration of this route into design software, should all be considered within the scope of this invention. This technical solution is not limited to the specified tapered bearing.

Claims

1. An optimization method for wind turbine bearings, characterized in that, Includes the following steps: By applying the ultimate load spectrum and fatigue load spectrum to the bearing model constructed based on the initial parameters of the bearing, the maximum load on the bearing rollers under the ultimate load spectrum is obtained. F j Fatigue life and fatigue stress of bearings under fatigue load spectrum; The initial parameters are adjusted to obtain optimized parameters with the goal of achieving the specified fatigue life. Based on the inner raceway parameters in the optimized parameters, the inner raceway parameters after deformation caused by bearing interference fit and temperature rise during bearing operation are determined; the times corresponding to the load steps in the fatigue load spectrum are sorted according to the magnitude of their corresponding fatigue stresses, and the maximum roller loads corresponding to the first and last fatigue load spectra in the total time sorted from largest to smallest fatigue stress are determined, either by the specified proportion of the first time or the specified proportion of the last time sorted from smallest to largest fatigue stress. F 1, F k ; The calculations were performed based on the roller parameters and the deformed inner raceway parameters in the optimized parameters. F j , F 1, F k The contact stress between the bearing roller and the inner ring raceway. P j , P 1, P k ; by P j , P 1, P k With the goal of meeting the corresponding regulations, the roller surface trimming amount in the roller parameters is adjusted to obtain the final parameters of the bearing.

2. The optimization method for wind power bearings according to claim 1, characterized in that, The deformed inner raceway parameters include the diameter and surface shape of the inner raceway, which are calculated based on the slicing method theory.

3. The optimization method for wind power bearings according to claim 1, characterized in that, P j , P 1, P k All results were obtained based on the finite-length line contact theory.

4. The optimization method for wind power bearings according to claim 1, characterized in that, The bearing model adopts a flexible simulation model.

5. The optimization method for wind power bearings according to claim 1, characterized in that, The steps of the optimization method are implemented using transmission simulation software.

6. The optimization method for wind power bearings according to claim 1, characterized in that, F j , F 1, F k Each includes four values: the contact force between the roller and the inner ring, the contact force between the roller and the outer ring, the torque between the roller and the inner ring, and the torque between the roller and the outer ring.

7. The optimization method for wind power bearings according to claim 1, characterized in that, The stipulated percentage is 20%.

8. The optimization method for wind power bearings according to claim 1, characterized in that, The equivalent number of the ultimate load spectrum ranges from 10 to 25.

9. The optimization method for wind power bearings according to claim 1, characterized in that, The number of equivalent groups in the fatigue load spectrum ranges from 100 to 500.

Images