A nonlinear fatigue calculation method for wind turbine tower top flange
By performing static finite element calculations and post-processing in ANSYS and combining it with the ncode algorithm for fatigue damage calculation, the problem that traditional linear fatigue analysis methods cannot accurately reflect the nonlinear stress and load relationship of the wind turbine tower top flange is solved, achieving more efficient and accurate fatigue calculation results.
Patent Information
- Application Number
- CN202211233073.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-10-10
- Publication Date
- 2025-09-09
- Estimated Expiration
- 2042-10-10
AI Technical Summary
Conventional linear fatigue analysis methods cannot accurately reflect the nonlinear relationship between stress and load in the transition fillet of the wind turbine tower flange and the neck weld area, resulting in distorted fatigue calculation results. Furthermore, manual node selection and assumed stress directions introduce errors, making the evaluation process inefficient and prone to errors.
ANSYS was used for static finite element calculations, and the result files were post-processed using multi-segment fitting. Fatigue damage calculations were then performed using the NCODE algorithm. The specific steps included: performing static finite element calculations in ANSYS, post-processing to fit nonlinear load stress curves, splitting the load time series, and performing fatigue assessment using the NCODE algorithm. The calculations were performed using the critical plane method and the Miner linear damage accumulation criterion.
It improves the accuracy of fatigue calculation results, overcomes the limitations of commercial fatigue software in nonlinear analysis, improves calculation efficiency, reduces human errors, and ensures the efficiency and reliability of fatigue damage calculation of wind turbine tower top flange.
Smart Images
Figure CN115618672B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of wind power generation, and in particular to a nonlinear fatigue calculation method for a tower top flange of a wind turbine generator set. Background Art
[0002] During the design and development of large wind turbines, simulation design of key components is required, and the accuracy of the simulation directly affects the design quality of the components. In recent years, with the rapid development of computer technology, finite element numerical analysis (FEA) has been increasingly used in the design and development of wind turbines. Among them, fatigue analysis of complex components is mainly based on finite element simulation calculations, and the quality of their design directly determines the overall reliability of the wind turbine. In the industry, finite element simulation calculations and commercial fatigue software are mainly used to solve linear fatigue analysis problems. The top flange of the wind turbine tower is directly connected to the yaw bearing and the main frame, resulting in a strong nonlinear stress response of the fatigue-weak transition fillet and neck weld area to external loads. For the fatigue strength assessment of this area of the top flange, the traditional linear fatigue analysis method based on commercial fatigue software is no longer applicable. The industry's nonlinear fatigue calculation method for the transition fillet and neck weld of the tower top flange is mainly divided into the following steps: A. Select some nodes with greater potential damage based on experience; B. Assume a stress direction that contributes most to fatigue damage at each node based on experience; C. Extract the normal stress values under different external loads in the corresponding assumed direction for each node, and draw a nonlinear load-stress curve; D. Based on the Markov matrix of the external load and the nonlinear load-stress curve of each node, use the linear interpolation method to interpolate and calculate the Markov matrix of the normal stress variation range of each node, and calculate the fatigue damage of each node in combination with the SN curve.
[0003] Problems existing in the above prior art:
[0004] 1) When performing fatigue analysis on the top flange transition fillet and neck weld, if the traditional linear analysis method based on commercial fatigue software is used for fatigue analysis, when performing stress time series synthesis of the nodes in the transition fillet and neck weld area, it can only be assumed that the relationship between stress and external load is linear, and the nonlinear relationship between stress and load at the top flange transition fillet and neck weld position cannot be truly considered, which leads to distortion of fatigue calculation results;
[0005] 2) This approach uses empirically selected nodes and assumes the direction of each node's maximum damage stress to extract node stress data and plot a nonlinear load-stress curve. This is then combined with the Markov matrix to calculate fatigue damage. Because this method uses empirically selected evaluation nodes, there's a risk of missing nodes with the highest actual damage values. Furthermore, the assumed maximum damage stress direction often deviates significantly from the true maximum damage stress direction, often resulting in calculated damage that's less than the actual fatigue damage, posing a fatigue safety risk. Furthermore, this method requires a significant amount of manual work from engineers, leading to inefficient and error-prone assessments. Summary of the Invention
[0006] In response to the shortcomings of the existing technology, the present invention proposes a nonlinear fatigue calculation method for the tower top flange of a wind turbine generator set to solve the technical problems in the existing technology that the traditional linear analysis method can only assume that the stress and load changes are linear, resulting in distorted fatigue calculation results, and manual selection may omit the nodes with the largest actual damage values, which may easily lead to low evaluation efficiency and easy errors.
[0007] A nonlinear fatigue calculation method for a wind turbine tower top flange comprises: performing static finite element calculation in ANSYS and post-processing the calculation result file; dividing the non-constant load time series loaded in the finite element calculation into intervals according to the load values loaded in the loading sub-steps; performing fatigue damage calculation using the NCODE algorithm, and using the post-processed file and the split load time series as inputs to the NCODE algorithm; and performing fatigue assessment using the critical plane method and the Miner linear damage accumulation criterion in the NCODE algorithm to obtain fatigue calculation results.
[0008] In one embodiment, a static finite element calculation is performed in ANSYS, and a calculation result file is post-processed, including: setting the same number of loading sub-steps in ANSYS for the loading of constant loads and non-constant loads in three working conditions; loading the load based on the number of loading sub-steps, performing static finite element calculation on the three working conditions, and obtaining three RST result files corresponding to the three working conditions; and post-processing the three RST result files to obtain a post-processed RST result file.
[0009] In one embodiment, the non-constant load includes a minimum load and a maximum load.
[0010] In one embodiment, the three working conditions include: minimum load+constant load, maximum load+constant load, and constant load.
[0011] In one embodiment, the three RST result files are post-processed to obtain the post-processed RST result files, including: fitting the nonlinear load stress curves of each node in the three RST result files in a multi-segment form to obtain a nonlinear load stress curve equation in a multi-segment form for each node; using the nonlinear load stress and non-constant load in the nonlinear load stress curve equation as input, solving a set of binary linear equations for each straight line segment in the multi-segment, and determining the slope and intercept parameters of each straight line segment; storing the obtained slope and intercept parameters of each straight line segment in each result channel of the same RST file in order from small to large in terms of load range in units of nodes, to obtain the post-processed RST result file.
[0012] In one embodiment, the step of performing interval splitting on the non-constant load time series loaded in the finite element calculation according to the load values loaded in the loading sub-steps includes: determining the number of intervals selected for splitting according to the number of loading sub-steps for loading the non-constant load; splitting the load intervals according to the load size at each time point in the time series; and each load channel obtained after splitting corresponds one-to-one to each result channel in the rst file obtained by post-processing.
[0013] In one embodiment, the stress time series expression of the node calculated by the ncode algorithm is:
[0014]
[0015] Where S(t) is the stress time series; M1(t), M m (t) are the time series corresponding to each load component in the load time series file, S FE，1 ,…,S FE，m They are the same as M1, ..., M in the result file respectively. m The nodal finite element stress corresponding to the load; SF is the load magnification factor in ncode, Offset is the load offset factor, and Divider is the stress scaling factor.
[0016] In one embodiment, the critical plane method is used to perform fatigue analysis.
[0017] In one embodiment, the miner linear damage accumulation criterion is used to calculate fatigue damage.
[0018] It can be seen from the above technical solution that the beneficial technical effects of the present invention are as follows:
[0019] 1. This solution is applicable to both linear and nonlinear analysis, especially when the analysis model has strong nonlinearity. Post-processing methods are used to more accurately fit the load-stress relationship curve using the results of polyline interpolation. This solves the difficulty of commercial fatigue software in performing strong nonlinear fatigue analysis at the input end, enabling it to obtain fatigue stress time series and damage calculation results that are more in line with reality, thereby improving the accuracy of fatigue calculation results.
[0020] 2. During nonlinear analysis, existing manual or simple programming methods for extracting data and interpolating calculations are often unable to fully reflect the fatigue analysis status of components. However, this solution utilizes the ncode algorithm for analysis. Compared with simple programming, the ncode algorithm is more efficient when performing the same workload, significantly reducing computational costs and improving fatigue assessment efficiency.
[0021] 3. Existing nonlinear fatigue analysis methods often use manual calculations or simple programming to select a few points for calculation, and there is also the problem of artificially assuming the maximum fatigue damage stress direction. This method can efficiently calculate a large number of nodes. At the same time, when analyzing fatigue through ncode, the critical plane method can be used to calculate damage to achieve calculation results of damage in different stress directions for each node, and find the maximum fatigue stress direction. Compared with manual or simple programming to calculate the artificially assumed maximum direction, the results are more realistic, thus avoiding human errors. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly describes the drawings required for the specific embodiments or the description of the prior art. Similar elements or parts are generally identified by similar reference numerals throughout the drawings. Elements or parts in the drawings are not necessarily drawn to scale.
[0023] Figure 1 1 is a flow chart of a method for calculating nonlinear fatigue of a wind turbine tower top flange in one embodiment;
[0024] Figure 2 In one embodiment Figure 1 Schematic diagram of the process of step S1;
[0025] Figure 3 In one embodiment Figure 2 Flow diagram of step S13;
[0026] Figure 4 A schematic diagram of multi-segment fitting of a nonlinear load stress curve in one embodiment;
[0027] Figure 5 This is an illustration of the splitting effect between time series load partitions. DETAILED DESCRIPTION
[0028] The following embodiments of the technical solution of the present invention will be described in detail with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention and are therefore only examples and are not intended to limit the scope of protection of the present invention.
[0029] It should be noted that, unless otherwise specified, the technical or scientific terms used in this application should have the common meanings understood by those skilled in the art to which this invention belongs. The terms "first," "second," and so on, in the description and claims of the embodiments of the present disclosure, and in the accompanying drawings, are used to distinguish similar objects and are not necessarily used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate for the purposes of describing the embodiments of the present disclosure herein. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. Unless otherwise specified, the term "plurality" means two or more. In the embodiments of the present disclosure, the character " / " indicates that the preceding and following objects are in an "or" relationship. For example, A / B means: A or B. The term "and / or" describes an association relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A or B, or: A and B. The term "corresponding" can refer to an association relationship or a binding relationship; A and B corresponding means that A and B are in an association relationship or a binding relationship.
[0030] In one embodiment, Figure 1 As shown, a nonlinear fatigue calculation method for a wind turbine tower top flange is provided, comprising the following steps:
[0031] S1 performs static finite element calculations in ANSYS and post-processes the calculation result files.
[0032] Specifically, static finite element calculations of three working conditions, namely "minimum load + constant load", "maximum load + constant load" and "constant load", were performed in ANSYS respectively, and the RST result files corresponding to the three working conditions were uniformly post-processed.
[0033] In one embodiment, step S1 includes:
[0034] S11 Set the same number of loading sub-steps for the constant load and non-constant load in the three working conditions in ANSYS.
[0035] Specifically, the three load conditions include: minimum load + constant load, maximum load + constant load, and constant load. Non-constant loads include minimum load and maximum load. The number of loading substeps can be appropriately set based on computational cost and result accuracy.
[0036] S12 performs load loading based on the number of loading sub-steps, performs static finite element calculations on the three working conditions, and obtains three RST result files corresponding to the three working conditions.
[0037] Specifically, 1) the constant load is loaded from the first load step, and the non-constant load is loaded from the second load step. The load sizes of each sub-step when the constant load is loaded in two sub-steps and the non-constant load is loaded in three sub-steps are shown in Table 1, where Fcon is the constant load, Mmin is the minimum value of the non-constant load, and Mmax is the maximum value of the non-constant load.
[0038] Table 1 Schematic diagram of the load size of each sub-step in ANSYS
[0039]
[0040] S13 post-processes the three rst result files to obtain post-processed rst result files.
[0041] Specifically, after the solution is completed, the three RST result files obtained by the solution are extracted for post-processing. The specific post-processing is shown in the following embodiment.
[0042] In one embodiment, step S13 includes:
[0043] S131 uses a multi-segment form to fit the nonlinear load-stress curve of each node in the three RST result files, and obtains the nonlinear load-stress curve equation of each node in a multi-segment form.
[0044] Specifically, the nonlinear load stress curve of each node is fitted using a multi-segment form. Figure 4 The nonlinear load-stress curve is fitted using six line segments, and the corresponding nonlinear load-stress curve equation is shown below:
[0045]
[0046] S132 uses the nonlinear load stress and the non-constant load in the nonlinear load stress curve equation as input, solves a set of two-variable linear equations for each straight line segment in the multi-line segment, and determines the slope and intercept parameters of each straight line segment.
[0047] Specifically, the stress calculation results corresponding to each node under each load sub-step and the non-constant load value added in each load sub-step are used as input, and a set of two-variable linear equations are solved for each straight line segment in the multi-segment to determine the slope and intercept parameters of the straight line equation. Figure 4 Taking the straight line segment y1 as an example, the solution of the two-variable linear equation system with slope a1 and intercept b1 is shown as follows. In the equation, M min are the non-constant load values, They are M min Nodal stress results under load:
[0048]
[0049] S133 stores the obtained slope and intercept parameters of each straight line segment in the order of the load interval from small to large in each result channel of the same RST file in units of nodes, and obtains the post-processed RST result file.
[0050] Specifically, the slope and intercept parameters of each segmented line used for coupling the nonlinear load-stress curve of each node are stored in each result channel of the same RST file in order from small to large load intervals, with the node as the unit. If the minimum and maximum loads are both loaded for m sub-steps, the number of intervals selected during splitting is 2m. Each load interval corresponds to two columns of result channels. The first column of the result channel stores the line segment slope parameters, and the second column of the result channel stores the line segment intercept parameters. The data storage format of each result channel in the RST file generated by post-processing is shown in Table 2.
[0051] Table 2 Storage format of node data in the rst file obtained after post-processing
[0052]
[0053]
[0054] Note: Among them, a 11 with b 11 They represent the slope and intercept of the first segment of the broken line y1 of the nonlinear load stress curve of fitting node 1 respectively.
[0055] S2 divides the non-constant load time series loaded in the finite element calculation into intervals according to the load values loaded in the loading sub-step.
[0056] Specifically, the non-constant load time series (single-channel fatigue time series) loaded in the finite element calculation is divided into intervals according to the load values loaded in each loading sub-step.
[0057] In one embodiment, step S2 includes: determining the number of intervals selected during splitting based on the number of loading sub-steps for loading a non-constant load; splitting the load intervals based on the load size at each time point in the time series; and corresponding one-to-one to each load channel obtained after splitting and each result channel in the rst file obtained after post-processing.
[0058] Specifically, if the minimum and maximum loads are both loaded with m sub-steps, the number of intervals selected during splitting is 2m. Each load interval corresponds to two columns of channels. The first column of channels is the load value, and the second column of channels is the value calibration channel. The load size corresponding to each time point in the time series is in which load interval, and its value falls into the load channel of the corresponding load interval, and the value calibration channel of the load interval is assigned to 1, and the other load interval channels corresponding to the time point are assigned to 0. The load channels obtained after splitting correspond one-to-one to the result channels in the rst file obtained by post-processing. The effects before and after splitting are as follows: Figure 5 shown.
[0059] S3 uses the ncode algorithm to perform fatigue damage calculation, and uses the above-mentioned post-processed files and split load time series as inputs of the ncode algorithm.
[0060] Specifically, the ncode algorithm is essentially a commercial fatigue analysis software. It can be seen from the stress time series expression of the node calculated by the ncode algorithm in the following embodiment that the rst result file obtained by subsequent processing and the load time series obtained after splitting are used as ncode calculation inputs. Then, the node stress size corresponding to the load value on the nonlinear load stress curve can be calculated for different load values, which overcomes the limitation of commercial fatigue software that cannot directly perform strong nonlinear fatigue analysis.
[0061] In one embodiment, the principle expression of the stress time series of the node calculated by the ncode algorithm is:
[0062]
[0063] Where S(t) is the stress time series; M1(t), M m (t) are the time series corresponding to each load component in the load time series file, S FE，1 ,…,S FE，m They are the same as M1, ..., M in the result file respectively. m The nodal finite element stress corresponding to the load; SF is the load magnification factor in ncode, Offset is the load offset factor, and Divider is the stress scaling factor.
[0064] Specifically, the SF parameter defaults to 1, the Offset parameter defaults to 0, and the Divider parameter defaults to 1.
[0065] S4 uses the critical plane method and miner linear damage accumulation criterion in the ncode algorithm to perform fatigue assessment and obtain fatigue calculation results.
[0066] Specifically, the critical plane method is used for fatigue analysis. Miner's linear damage accumulation criterion is used to calculate fatigue damage. Miner's linear cumulative damage theory is a linear method for calculating cumulative damage. It holds that: a) under constant-amplitude cyclic loading, each cycle causes the same damage to the material; b) under variable-amplitude cyclic loading, the damage to the material caused by cyclic loads of different amplitudes is relatively independent and independent of the loading sequence; and c) the critical fatigue damage of the material is 1.
[0067] Based on the above embodiments, it can be seen that this solution uses post-processing means to overcome the inconsistency between the nonlinear finite element simulation results and the calculation input of the commercial fatigue software, and solves the limitation that the commercial fatigue software can only perform fatigue analysis in the area where the stress changes linearly with the external load. In addition, this solution uses the high efficiency of the commercial fatigue software to achieve rapid calculation of the fatigue damage of all nodes in the transition fillet area of the tower top flange and the neck weld area, and uses the critical plane method (considering a sufficient number of stress directions to calculate fatigue damage) to avoid the disadvantage of a large deviation between the assumed maximum damage stress direction and the actual maximum damage stress direction, thereby ensuring the efficiency and reliability of the fatigue damage calculation of the wind turbine tower top flange.
[0068] Obviously, those skilled in the art should understand that the modules or steps of the present invention described above can be implemented using a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices. Alternatively, they can be implemented using program codes executable by the computing device, so that they can be stored in a computer storage medium (ROM / RAM, magnetic disk, optical disk) and executed by the computing device. In some cases, the steps shown or described can be performed in a different order than herein, or they can be made into individual integrated circuit modules, or multiple modules or steps can be made into a single integrated circuit module for implementation. Therefore, the present invention is not limited to any specific combination of hardware and software.
[0069] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and description of the present invention.
Claims
1. A nonlinear fatigue calculation method for a wind turbine tower top flange, characterized in that: include: Perform static finite element calculations in ANSYS and post-process the calculation result files. This includes setting the same number of loading sub-steps for the constant load and non-constant load in the three working conditions in ANSYS; applying loads based on the number of loading sub-steps, performing static finite element calculations for the three working conditions, and obtaining three RST result files corresponding to the three working conditions. The three RST result files are post-processed, and the nonlinear load-stress curves of each node in the three RST result files are fitted in a multi-segment form to obtain a nonlinear load-stress curve equation in a multi-segment form for each node; the nonlinear load stress and the non-constant load in the nonlinear load-stress curve equation are used as input, and a set of two-variable linear equations are solved for each straight line segment in the multi-segment to determine the slope and intercept parameters of each straight line segment; the slope and intercept parameters of each straight line segment are stored in each result channel of the same RST file in order of load interval from small to large on a node basis to obtain a post-processed RST result file; The non-constant load includes a minimum load and a maximum load; the three working conditions include: considering the minimum load and a constant load, considering the maximum load and a constant load, and considering only the constant load; For the non-constant load time series loaded in the finite element calculation, the interval is split according to the load value loaded in the loading sub-step; The ncode algorithm is used to calculate fatigue damage, and the above-mentioned post-processed files and split load time series are used as inputs of the ncode algorithm; In the ncode algorithm, the critical plane method and miner linear damage accumulation criterion are used to perform fatigue assessment and obtain fatigue calculation results.
2. The method according to claim 1, characterized in that The step of performing interval splitting on the non-constant load time series loaded in the finite element calculation according to the load values loaded in the loading sub-steps includes: The number of intervals selected during splitting is determined based on the number of loading sub-steps for loading non-constant loads; Split the load interval according to the load size at each time point in the time series; Each load channel obtained after splitting corresponds one-to-one to each result channel in the rst file obtained after post-processing.
3. The method according to claim 1, characterized in that The stress time series expression of the node calculated by the ncode algorithm is: Where S(t) is the stress time series; 、 are the time series corresponding to each load component in the load time series file, S FE，1 ,…,S FE，m In the result file, respectively 、…、 Finite element stress of the node corresponding to the load; SF is the load magnification factor in ncode, Offset is the load offset factor, and Divider is the stress scaling factor.
4. The method according to claim 1, wherein The critical plane method is used to perform fatigue analysis.
5. The method according to claim 1, wherein The Miner linear damage accumulation criterion is used to calculate fatigue damage.