A method for calculating fatigue life of in-service blades based on stress concentration coefficient of damaged blades and rotating speed of wind wheels

Abstract

The application discloses an in-service blade fatigue life calculation method based on a damaged blade stress concentration coefficient and a wind wheel rotating speed, and comprises the following steps: collecting unit design parameters, material S-N curve parameters and SCADA wind wheel rotating speed time series; detecting a blade defect type and a spanwise position, combining damage interval correction to obtain an equivalent stress concentration coefficient; based on the actual rotating number of a reference period of the rotating speed data, combining a field area wind frequency to estimate the total rotating number of the blade design life; accumulating the total rotating number of the unit from operation to the present, relying on the Miner theory to combine the stress concentration coefficient to calculate the equivalent damage number, and converting to obtain the remaining service days of the blade. The application does not need complex finite element time domain simulation, can consider the stress amplification of various defects and the real operation condition of the unit, greatly reduces the calculation cost, and quickly and accurately quantifies the remaining life of the in-service damaged blade.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, specifically to a method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed. Background Technology

[0002] As a key wind-catching component of wind turbine generators, wind turbine blades endure complex and variable alternating loads over long periods. Their structural integrity and reliability directly determine the overall operational safety and power generation efficiency of the turbine. During manufacturing, blades are prone to initial defects such as wrinkles and delamination. During service, they may also suffer damage due to lightning strikes, fatigue damage, or foreign object impacts. These manufacturing defects and operational damage can cause localized fiber bending and resin enrichment, significantly reducing the mechanical properties of the composite material layer and creating significant stress concentration points. Such defects and damage accelerate fatigue crack propagation, and in severe cases, can lead to blade fracture, directly affecting the accuracy of fatigue life assessments for in-service blades. Therefore, achieving accurate life assessment of damaged blades is crucial for ensuring the safe and stable operation of the turbine generator set.

[0003] Currently, blade fatigue life assessment mainly relies on theoretical calculations and certification tests during the design phase. These methods are based on standard load conditions and ideal, defect-free structural models, estimating the theoretical life of the blade through rainflow counting and Miner's linear cumulative damage theory. For example, the paper "Fatigue Load and Life Analysis of Megawatt-Class Wind Turbine Blades" proposes obtaining time-series fatigue loads based on the Kaimal turbulent wind model, processing the load spectrum using rainflow counting and Goodman mean stress correction, and then combining it with Weibull wind speed distribution to estimate the theoretical fatigue life of the blade. However, this approach focuses on extrapolating the standard life of intact blades, completely neglecting the accelerated damage accumulation characteristics of blades with defects during service. It is only suitable for new product design verification and lacks the engineering feasibility for assessing the operation and maintenance of damaged blades in service. The paper "Research on Wind Turbine Blade Life Analysis and Fault Diagnosis Methods" constructs a whole-machine GH-Bladed load model and a blade finite element model, extracts the stress spectrum of critical nodes through transient dynamics, and calculates the full-life fatigue damage of the blade based on a multi-axis Miner model. However, it also does not consider the coupling effect between the damaged blades in service and the actual speed operation data of the unit, and cannot be directly used for quantitative assessment of the remaining life of damaged blades in service.

[0004] In summary, existing blade fatigue life assessment methods are mostly geared towards the design verification of brand-new, undamaged blades, and have the following significant shortcomings for in-service damaged blades: First, they cannot consider the specific defects that occur on the blade during actual operation and the resulting local stress concentration effects; second, the assessment process often uses assumed load spectra under design conditions, failing to integrate with the actual operating environment of the blade, thus failing to accurately assess the lifespan of the blade after operation with damage. If traditional ideal defect-free models are used for calculations, the stress amplification effect caused by defects will be underestimated, resulting in an overly optimistic remaining life assessment result, which can lead to delayed operation and maintenance decisions and increase the safety risk of sudden blade fracture.

[0005] However, existing technologies have two major gaps: First, in-service blades exhibit a wide variety of defect types, with highly random sizes and distributions, and stress interference coupling exists between multiple damage types. Traditional Miner damage frameworks can only input uniform nominal stress, failing to quantitatively convert on-site measured defect characteristics into life correction parameters to directly participate in damage iteration. Second, the operating conditions of units at different stations vary significantly. To recreate the true load, current methods can only rely on specialized load simulation software for full-time-domain numerical simulation, resulting in long calculation cycles and high computational costs, making it unsuitable for rapid inspection and assessment of large numbers of damaged in-service blades. Therefore, there is an urgent need for a method to calculate the fatigue life of in-service blades that can combine actual defect characteristics with real unit operating data to achieve efficient and accurate quantitative assessment of the remaining blade life. Summary of the Invention

[0006] To address the technical problems of traditional blade fatigue assessment, such as relying on design loads, ignoring actual defects and operating conditions, and yielding conservative results that are difficult to guide on-site operation and maintenance, this invention aims to provide a method for calculating the fatigue life of in-service blades based on the stress concentration factor (SCF) of damaged blades and the rotor speed. This method combines the stress concentration effect caused by blade defects with actual turbine operating data. By introducing the SCF, it accurately reflects the accelerating effect of defects on local fatigue damage, overcoming the limitations of traditional defect-free models that do not match actual structural conditions. Simultaneously, it fully utilizes the time-series data of rotor speed collected by the SCADA system to statistically count the actual number of rotations, ensuring that the life calculation is based on the actual operating intensity of the turbine, rather than idealized design conditions, significantly improving the authenticity and reliability of the assessment results.

[0007] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows:

[0008] A method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed, the method comprising the following steps:

[0009] Step 1: Obtain the design life of the target wind turbine, the SN curve parameter m of the blade material, and collect time series data of the wind turbine rotation speed with timestamps from the unit's SCADA system;

[0010] Step 2: Detect and determine the type and spanwise location of the defect in the target blade, and match the stress concentration factor (SCF) corresponding to the defect type;

[0011] Step 3: Based on the wind turbine rotation speed time series data obtained in Step 1, extract a continuous and complete reference time period, count the actual number of wind turbine rotations Nref within the reference time period, and record the total number of days Dref corresponding to the reference time period.

[0012] Step 4: Combine the unit design parameters with the actual wind frequency distribution of the wind farm, calculate the annual number of blade rotations for each wind speed interval, and convert it to obtain the theoretical total number of blade rotations Ntotal within the design life Ldesign.

[0013] Step 5: Calculate and obtain the cumulative number of operating days Dmaint of the blade from the commissioning date or the last maintenance date Tmaint to the current SCADA data statistics cutoff date Tnow, and obtain the actual total number of rotations Nmaint of the wind turbine during this period based on SCADA data statistics, and retrieve the historical cumulative number of rotations Nmaint0 before Tmaint.

[0014] Step 6: Based on Miner's damage theory and the SN curve parameter m, combined with the stress concentration factor SCF, the actual total number of rotor rotations Nmaint, and the historical cumulative number of rotations Nmaint0, calculate the equivalent used rotations Neq including damage; subtract the equivalent used rotations Neq from the theoretical total number of blade rotations Ntotal to calculate the remaining usable rotations Nres under the influence of defects; then use the actual number of rotor rotations Nref and the total number of days Dref during the reference time period to characterize the current operating intensity of the unit, and convert the remaining usable rotations Nres into the remaining blade life days Lres.

[0015] Furthermore, in step 1, the design life Ldesign of the target unit and the SN curve parameter m of the main load-bearing material of the blade are obtained from the wind farm information system; at the same time, historical operating data containing timestamps and rotor speed are exported from the unit's SCADA system to construct the rotor speed time series data.

[0016] Step 2 further includes:

[0017] Acquire images and dimensional data of surface and internal damage of the blade, identify the defect type, which includes at least surface pits, leading edge cracks, trailing edge cracks, interlayer delamination, bond debonding, lightning strike ablation pits, and hole edge microcracks; collect the position coordinates of the defects along the blade spanwise, determine the structural zone to which the defects belong, which includes at least the main beam region, web region, blade tip leading edge region, and shell skin region.

[0018] Based on the defect geometry, the structural region where the defect is located, and the load direction, the analytical calculation formula for composite material defects is used, or a local refined finite element simulation model of the blade is built to apply the load, and the basic stress concentration factor corresponding to a single damage is directly calculated without relying on the defect coefficient lookup table. If there are two or more close-range damages in the same area, the presence of stress interference effect is determined based on the ratio of damage spacing to damage feature size. The corresponding interference correction coefficient is obtained through experimental calibration, and the basic stress concentration factor is amplified and corrected using the corresponding interference correction coefficient, finally obtaining the stress concentration factor SCF of the coupled multi-damage effect.

[0019] Step 3 further includes:

[0020] A reference time period with good continuity and SCADA data integrity reaching a preset integrity threshold is selected, and the total number of days corresponding to this reference time period is denoted as Dref. For the ideal scenario of continuous rotational speed variation, an integral operation is performed on the rotational speed data of all sampling points within the reference time period to calculate the actual number of wind turbine rotations Nref within the reference time period.

[0021]

[0022] For SCADA discrete sampling scenarios, the actual number of wind turbine rotations Nref within the reference time period is represented as:

[0023]

[0024] in The start time of the reference period The unit is seconds, n(t) is the rotational speed at time t, and Dref is the number of days in the reference period.

[0025] 1 day = 86400 seconds, n i Let dt represent the rotational speed of the i-th sampling point, dt represent the time interval of a single sampling, Nref represent the total number of wind turbine rotations during this time interval, and M represent the number of valid sampling points.

[0026] Step 4 further includes:

[0027] The wind speed is divided into several intervals from the cut-in wind speed to the cut-out wind speed. For each wind speed interval i, the probability fi of the occurrence of the wind speed interval within the design life and the number of annual occurrence hours Ti = fi × 8766 are calculated based on the wind frequency distribution of the wind farm.

[0028] Determine the unit speed rpmi corresponding to each wind speed range, with the unit of speed being revolutions per minute; below the rated wind speed, the speed is directly proportional to the wind speed; above the rated wind speed, the speed remains constant at the rated speed until the maximum blade tip speed limit is reached;

[0029] After converting the annual occurrence hours to minutes, and combining this with the annual occurrence time and turbine rotation speed for each wind speed range, the number of rotations corresponding to that wind speed range within a year is calculated as Ci = rpmi × Ti × 60:

[0030] The total number of rotations Ci across all wind speed ranges is summed to obtain the annual total number of rotations Nyear. Introducing the wind farm availability rate Uti and the design life years Ldesign, the total number of blade rotations within the design life cycle is calculated.

[0031]

[0032] Where Ntotal is the estimated total number of rotations of the blade during its design life, and the wind farm availability rate Uti is used to deduct invalid operating periods, including maintenance shutdowns and fault shutdowns.

[0033] Furthermore, in step 5, if Tmaint is the unit commissioning date, then Nmaint0=0; otherwise, the last maintenance date Tmaint is obtained by querying historical SCADA data or unit operation records.

[0034] Step 6 further includes:

[0035] Combining the stress concentration factor SCF, the actual total number of rotor rotations Nmaint, and the historical cumulative number of rotations Nmaint0, calculate the equivalent number of rotations used, Niq, including damage:

[0036] ;

[0037] The remaining usable number of revolutions Nres under the influence of defects is calculated by subtracting the equivalent used number of revolutions Neq from the theoretical total number of revolutions Ntotal of the blade:

[0038] ;

[0039] The current operating intensity of the unit is characterized by the actual number of rotor rotations Nref and the total number of days Dref during the reference time period. Combined with the stress concentration factor SCF, the remaining usable rotations Nres are converted into the remaining blade life days Lres.

[0040] .

[0041] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0042] First, the present invention provides a method for calculating the fatigue life of in-service blades based on the stress concentration factor (SCF) of damaged blades and the rotor speed. This method combines the SCF corresponding to the damage with the actual number of rotor rotations based on SCADA system rotor speed data, establishing a rapid method for calculating blade fatigue life. This method overcomes the shortcomings of traditional design evaluations that ignore actual defects and actual operating loads, making the evaluation results closer to the actual service state of the blades and providing a more scientific basis for wind farm operation and maintenance decisions.

[0043] Secondly, the fatigue life calculation method for in-service blades based on the stress concentration factor of damaged blades and the rotor speed of the present invention can be combined with the actual environment by inputting SCADA instead of input load. The calculation process is simple and fatigue life prediction can be completed in a short time, which significantly reduces the cost of engineering applications and is suitable for batch units or offline evaluation in wind farms.

[0044] Third, the method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed of the present invention introduces the reference number of rotations based on SCADA data as the benchmark operating intensity, transforming the abstract remaining usable number of rotations into an intuitive remaining lifespan in days, thus overcoming the pain point of traditional methods that can only output the number of cycles and are difficult to know the remaining fatigue life.

[0045] Fourth, the fatigue life calculation method for in-service blades based on the stress concentration factor of damaged blades and the rotor speed of the present invention forms a complete calculation system from design parameters, defect characteristics, operating data to fatigue life, realizing the combination of design data and actual operating data of the unit, which can accurately reflect the accelerating effect of defects on fatigue damage, effectively improve the life assessment of in-service damaged blades, and has high engineering application value and promotion prospects. Attached Figure Description

[0046] Figure 1 This is a flowchart of a method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed, according to the present invention.

[0047] Figure 2 This is a schematic diagram illustrating the statistical reference number of rotational revolutions based on the time series data of wind turbine speed from the SCADA system.

[0048] Figure 3 This is a schematic diagram of the Weibull distribution wind frequency curve for a wind farm.

[0049] Figure 4 This is a schematic diagram of the speed-power characteristic curve of a wind turbine.

[0050] Figure 5 This is a schematic diagram showing the relationship between the stress concentration factor (SCF) and the remaining fatigue life of the blade. Detailed Implementation

[0051] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0052] This invention discloses a method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed. The method includes the following steps:

[0053] Step 1: Obtain the design life of the target wind turbine, the SN curve parameter m of the blade material, and collect time series data of the wind turbine rotation speed with timestamps from the unit's SCADA system;

[0054] Step 2: Detect and determine the type and spanwise location of the defect in the target blade, and match the stress concentration factor (SCF) corresponding to the defect type;

[0055] Step 3: Based on the wind turbine rotation speed time series data obtained in Step 1, extract a continuous and complete reference time period, count the actual number of wind turbine rotations Nref within the reference time period, and record the total number of days Dref corresponding to the reference time period.

[0056] Step 4: Combine the unit design parameters with the actual wind frequency distribution of the wind farm, calculate the annual number of blade rotations for each wind speed interval, and convert it to obtain the theoretical total number of blade rotations Ntotal within the design life Ldesign.

[0057] Step 5: Calculate and obtain the cumulative number of operating days Dmaint of the blade from the commissioning date or the last maintenance date Tmaint to the current SCADA data statistics deadline Tnow, and obtain the actual total number of rotations Nmaint of the wind turbine during this period based on SCADA data statistics, and retrieve the historical cumulative number of rotations Nmaint0 before Tmaint.

[0058] Step 6: Based on Miner's damage theory and the SN curve parameter m, combined with the stress concentration factor SCF, the actual total number of rotor rotations Nmaint, and the historical cumulative number of rotations Nmaint0, calculate the equivalent used rotations Neq including damage; subtract the equivalent used rotations Neq from the theoretical total number of blade rotations Ntotal to calculate the remaining usable rotations Nres under the influence of defects; then use the actual number of rotor rotations Nref and the total number of days Dref during the reference time period to characterize the current operating intensity of the unit, and convert the remaining usable rotations Nres into the remaining blade life days Lres.

[0059] This invention proposes a method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed. The specific process is as follows: Figure 1 As shown, this study aims to address the technical challenge of assessing the remaining lifespan of in-service wind turbine blades after manufacturing or operational damage, providing a highly accurate quantitative method for wind farm operation and maintenance decisions and blade life extension assessments.

[0060] First, the method of the present invention is based on the following pre-established input data:

[0061] 1. Unit design parameter library: contains basic parameters such as rated power, rotor diameter, optimal tip speed ratio, cut-in and cut-out wind speeds, drive train efficiency, and wind farm availability for the target model.

[0062] 2. Damage and Defect Input: Determine the characteristics of the blade defect and the stress concentration factor, obtain the characteristic parameters of the target blade defect, including the defect type and spanwise location; determine the stress concentration factor (SCF) corresponding to the damage, and input the stress concentration factor.

[0063] 3. Fatigue Load Safety Factor Assessment: Based on the actual location of the defect on the blade, its structural region, and the fatigue stress characteristics of that region, a fatigue load safety factor assessment is performed on the defect location to obtain the corresponding fatigue load safety factor. This factor is used to reduce the results in the subsequent remaining service life calculation process to reflect the damage differences of different structural regions under fatigue load, making the service life assessment results more closely reflect the actual service condition of the blade.

[0064] 4. Wind farm wind frequency model: Based on measured data of wind farms or standard Weibull distribution, determine the annual average wind speed and shape parameters to estimate the theoretical power generation and number of revolutions of the unit.

[0065] Based on the above input, the method of the present invention is implemented according to the following steps:

[0066] Step 1: Obtain unit design parameters and operating data

[0067] Obtain the target turbine's design life (Ldesign, in years) and the SN curve parameters m of the main load-bearing materials of the blades from the wind farm information system (e.g., m=10 for glass fiber and m=14 for carbon fiber). Simultaneously, export historical operating data, including timestamps and rotor speed (rpm), from the turbine's SCADA system. This data will be used for subsequent statistics on the actual number of rotations.

[0068] Step 2: Input stress concentration factor

[0069] Acquire images and dimensional data of surface and internal damage of the blade, identify the defect type, which includes at least surface pits, leading edge cracks, trailing edge cracks, interlayer delamination, bond debonding, lightning strike ablation pits, and hole edge microcracks; collect the position coordinates of the defects along the blade spanwise, determine the structural zone to which the defects belong, which includes at least the main beam region, web region, blade tip leading edge region, and shell skin region.

[0070] Based on the defect geometry, the structural region where the defect is located, and the load direction, the analytical calculation formula for composite material defects is used, or a local refined finite element simulation model of the blade is built to apply the load, and the basic stress concentration factor corresponding to a single damage is directly calculated without relying on the defect coefficient to look up the value in the table. If there are two or more close-range damages in the same area, the presence of stress interference effect is determined based on the ratio of the damage distance to the damage characteristic size. The corresponding interference correction coefficient is obtained through experimental calibration. The basic stress concentration factor is amplified and corrected using this interference correction coefficient, and finally the stress concentration factor SCF of the coupled multi-damage effect is obtained.

[0071] Step 3: Calculate the actual reference rotation number based on SCADA speed data.

[0072] Extract the most recent continuous and complete runtime segment from the SCADA data as a reference time period (e.g., Figure 2 (As shown). The number of days in the reference period, Dref, can be set according to the data quality, for example, 30 days. Integrate the rotational speed data at each sampling point within this period to obtain the total number of wind turbine rotations, Nref, within that period. The integration formula is:

[0073]

[0074] Discrete sampling calculation formula:

[0075]

[0076] Where n(t) is the rotational speed (rpm) at time t, and Dref is the number of days in the reference period (1 day = 86400 seconds). i Let Ni be the rotational speed (rpm) at the i-th sampling point, Nref be the total number of rotor rotations during that time period, and M be the number of valid sampling points. In practical applications, the rotational speeds of discrete sampling points are usually multiplied by the sampling interval and then summed.

[0077] Step 4: Estimate the total number of rotations for the blade's design life (in conjunction with...) Figure 3 , Figure 4 )

[0078] Based on the unit design parameters and the wind frequency distribution of the wind farm (in combination with...) Figure 3), calculate the expected total number of blade rotations Ntotal during the design life Ldesign. The specific method is as follows:

[0079] First, based on the wind frequency distribution of the wind farm, the probability of occurrence of different wind speed ranges within the design life is calculated.

[0080] The cumulative probability function of the Weibull distribution for wind speed at hub height is:

[0081]

[0082] Where: V hub Where is the wind speed at hub height, C is the Weibull distribution scale parameter, and k is the Weibull distribution shape parameter PW(V) hub Wind speed at wheel hub height does not exceed V hub The cumulative probability.

[0083] When wind farm data only provides the annual average wind speed V ave When k=2, the Rayleigh distribution (a special form of the Weibull distribution, i.e., k=2) can be used for simplified calculation, and its cumulative probability function is:

[0084]

[0085] Where: V ave This represents the average annual wind speed at the hub height.

[0086] Annual average wind speed V ave The general conversion relationship between the parameters of the Weibull distribution and the parameters of the Weibull distribution is as follows:

[0087]

[0088] Where Γ is the gamma function, and k=2.

[0089] Then, the wind speed from the cut-in wind speed Vin to the cut-out wind speed Vout is divided into several intervals (step size 0.5 m / s). For each wind speed interval [vi, vi+1]:

[0090] Calculate the probability fi of this wind speed range, which is the cumulative probability difference of the Weibull distribution.

[0091] Determine the unit rotational speed rpmi corresponding to this wind speed range. Below the rated wind speed, the rotational speed is proportional to the wind speed (rpm=λ.v / (πD)); above the rated wind speed, the rotational speed remains constant until the maximum tip speed limit is reached.

[0092] Calculate the occurrence time of this wind speed range within one year: Ti = fi × 8766 hours.

[0093] Calculate the number of rotations Ci = rpmi × Ti × 60 for this wind speed range within a year.

[0094] Sum the Ci values ​​across all wind speed ranges to obtain the total number of rotations Nyear over the design life of one year. Finally, multiply by the design life in years and consider the wind farm availability Uti:

[0095]

[0096] Where Ntotal is the estimated total number of rotations of the blade during its design life, Nyear is the total number of rotations in one year of its design life, Ldesign is the number of years of the design life, and Uti is the wind farm availability rate.

[0097] Step 5: Calculate the cumulative number of damage cycles from the time the unit was put into operation or the last maintenance to the current time.

[0098] Obtain the date of the last blade maintenance (Tmaint) (or the date the unit was put into operation), and the current SCADA data statistics cutoff date (Tnow). Calculate the number of days between the two dates (Dmaint), and extract all rotational speed records within this time period from the SCADA data. Similarly, use integration to obtain the actual total number of rotor rotations (Nmaint) within this time period.

[0099] At the same time, it is necessary to obtain the cumulative number of revolutions Nmaint0 from the start of unit operation to Tmaint.

[0100] If Tmaint is the unit commissioning date, then Nmaint0 = 0; otherwise, it can be obtained through historical SCADA data or unit operation records.

[0101] Step 6: Calculate the equivalent number of used revolutions and remaining life under the influence of defects (in combination with...) Figure 5 )

[0102] The presence of defects amplifies local stress, thereby accelerating fatigue damage. According to Miner's linear cumulative damage theory and the material's SN curve, a stress amplitude amplified by a factor of SCF corresponds to a fatigue life reduced to 1 / (SCF)m of the original value. Therefore, the equivalent number of damage cycles caused by N rotations under the influence of defects is N×(SCF)m.

[0103] Therefore, the equivalent number of damage cycles from the last maintenance to the current time is Nmaint × (SCF)m. Adding the number of cycles used before maintenance, Nmaint0, we get the total equivalent number of cycles used:

[0104]

[0105] The remaining usable number of revolutions is obtained by subtracting the equivalent number of revolutions used from the total design life revolutions Ntotal.

[0106]

[0107] Where Neq is the total equivalent used cycles, Nmaint×(SCF)m is the equivalent damaged cycles from the last maintenance to the current time, Nmaint0 is the used cycles before maintenance, Nres is the remaining available cycles, and Ntotal is the total number of cycles for the design lifetime.

[0108] The remaining available revolutions do not directly reflect calendar time and need to be converted in conjunction with the actual operating intensity of the current unit. Using the reference period revolutions Nref and days Dref obtained in step 3 as the baseline operating intensity, and considering that defects also have a magnifying effect on future operations, the formula for calculating the remaining lifetime days Lres is as follows:

[0109]

[0110] Where Lres represents the remaining lifetime days, Nres represents the remaining available laps, and Nref represents the number of laps for the reference period.

[0111] Dref represents the number of days in the reference period.

[0112] The meaning of this formula is: under the premise of maintaining the operating intensity of the reference period, the equivalent number of damage cycles generated each day in the future is Nref / Dref×(SCF)m. The remaining number of days is obtained by dividing the remaining available cycles by the damage rate of that day.

[0113] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0114] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed, characterized in that, The method includes the following steps: Step 1: Obtain the design life of the target wind turbine, the SN curve parameter m of the blade material, and collect time series data of the wind turbine rotation speed with timestamps from the unit's SCADA system; Step 2: Detect and determine the type and spanwise location of the defect in the target blade, and match the stress concentration factor (SCF) corresponding to the defect type; Step 3: Based on the wind turbine rotation speed time series data obtained in Step 1, extract a continuous and complete reference time period, count the actual number of wind turbine rotations Nref within the reference time period, and record the total number of days Dref corresponding to the reference time period. Step 4: Combine the unit design parameters with the actual wind frequency distribution of the wind farm, calculate the annual number of blade rotations for each wind speed interval, and convert it to obtain the theoretical total number of blade rotations Ntotal within the design life Ldesign. Step 5: Calculate and obtain the cumulative number of operating days Dmaint of the blade from the commissioning date or the last maintenance date Tmaint to the current SCADA data statistics deadline Tnow, and obtain the actual total number of rotations Nmaint of the wind turbine during this period based on SCADA data statistics, and retrieve the historical cumulative number of rotations Nmaint0 before Tmaint. Step 6: Based on Miner's damage theory and the SN curve parameter m, combined with the stress concentration factor SCF, the actual total number of rotor rotations Nmaint, and the historical cumulative number of rotations Nmaint0, calculate the equivalent used rotations Neq including damage; subtract the equivalent used rotations Neq from the theoretical total number of blade rotations Ntotal to calculate the remaining usable rotations Nres under the influence of defects; then use the actual number of rotor rotations Nref and the total number of days Dref during the reference time period to characterize the current operating intensity of the unit, and convert the remaining usable rotations Nres into the remaining blade life days Lres.

2. The method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed as described in claim 1, characterized in that, In step 1, the design life Ldesign of the target unit and the SN curve parameters m of the main load-bearing material of the blade are obtained from the wind farm information system. At the same time, historical operating data containing timestamps and rotor speed are exported from the unit's SCADA system to construct the rotor speed time series data.

3. The method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed as described in claim 1, characterized in that, Step 2 further includes: Acquire images and dimensional data of surface and internal damage of the blade, identify the defect type, which includes at least surface pits, leading edge cracks, trailing edge cracks, interlayer delamination, bond debonding, lightning strike ablation pits, and hole edge microcracks; collect the position coordinates of the defects along the blade spanwise, determine the structural zone to which the defects belong, which includes at least the main beam region, web region, blade tip leading edge region, and shell skin region. Based on the defect geometry, the structural region where the defect is located, and the load direction, the stress concentration factor of the foundation corresponding to a single damage is calculated using analytical formulas or local refined finite element simulation. If there are two or more close-range damages in the same region, the stress interference effect is judged based on the ratio of the damage spacing to the damage characteristic size. The corresponding interference correction coefficient is introduced to amplify and correct the stress concentration factor of the foundation, thus obtaining the final stress concentration factor SCF.

4. The method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed as described in claim 1, characterized in that, Step 3 further includes: A reference time period with good continuity and SCADA data integrity reaching a preset integrity threshold is selected, and the total number of days corresponding to this reference time period is denoted as Dref. For the ideal scenario of continuous rotational speed variation, an integral operation is performed on the rotational speed data of all sampling points within the reference time period to calculate the actual number of rotor rotations Nref within the reference time period. For SCADA discrete sampling scenarios, the actual number of wind turbine rotations Nref within the reference time period is represented as: in The start time of the reference period The unit is seconds, n(t) is the rotational speed at time t, and Dref is the number of days in the reference period. 1 day = 86400 seconds, n i Let dt represent the rotational speed of the i-th sampling point, dt represent the time interval of a single sampling, Nref represent the total number of rotations of the wind turbine during this time interval, and M represent the number of valid sampling points.

5. The method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed as described in claim 1, characterized in that, Step 4 further includes: The wind speed is divided into several intervals from the cut-in wind speed to the cut-out wind speed. For each wind speed interval i, the probability fi of the occurrence of the wind speed interval within the design life and the number of annual occurrence hours Ti = fi × 8766 are calculated based on the wind frequency distribution of the wind farm. Determine the unit speed rpmi corresponding to each wind speed range, with the unit of speed being revolutions per minute; below the rated wind speed, the speed is directly proportional to the wind speed; above the rated wind speed, the speed remains constant at the rated speed until the maximum blade tip speed limit is reached; After converting the annual occurrence hours to minutes, and combining this with the annual occurrence time and turbine rotation speed for each wind speed range, the number of rotations corresponding to that wind speed range within a year is calculated as Ci = rpmi × Ti × 60: The total number of rotations Ci across all wind speed ranges is summed to obtain the annual total number of rotations Nyear. Introducing the wind farm availability rate Uti and the design life years Ldesign, the total number of blade rotations within the design life cycle is calculated. Where Ntotal is the estimated total number of rotations of the blade during its design life, and the wind farm availability rate Uti is used to deduct invalid operating periods, including maintenance shutdowns and fault shutdowns.

6. The method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed as described in claim 1, characterized in that, In step 5, if Tmaint is the unit commissioning date, then Nmaint0 = 0; otherwise, the last maintenance date Tmaint is obtained by querying historical SCADA data or unit operation records.

7. The method for calculating the fatigue life of in-service blades based on the stress concentration factor of damaged blades and the rotor speed as described in claim 1, characterized in that, Step 6 further includes: Combining the stress concentration factor SCF, the actual total number of rotor rotations Nmaint, and the historical cumulative number of rotations Nmaint0, calculate the equivalent number of rotations used, Niq, including damage: ； The remaining usable number of revolutions Nres under the influence of defects is calculated by subtracting the equivalent used number of revolutions Neq from the theoretical total number of revolutions Ntotal of the blade: ； The current operating intensity of the unit is characterized by the actual number of rotor rotations Nref and the total number of days Dref during the reference time period. Combined with the stress concentration factor SCF, the remaining usable rotations Nres are converted into the remaining blade life days Lres. 。

Images