Wind Turbine Rotor Research Articles

Comparison of different cross-sectional approaches for the structural design and optimization of composite wind turbine blades based on beam models

Abstract. During the preliminary design phase of wind turbine blades, the evaluation of many design candidates in a short period of time plays an important role. Computationally efficient methods for the structural analysis that correctly predict stiffness matrix entries for beam models including the (bend–twist) coupling terms are thus needed. The present paper provides an extended overview of available approaches and shows their abilities to fulfill the requirements for the composite design of rotor blades with respect to accuracy and computational efficiency. Three cross-sectional theories are selected and implemented to compare the prediction quality of the cross-sectional coupling stiffness terms and the stress distribution based on different multi-cell test cross-sections. The cross-sectional results are compared with the 2D finite element code BECAS and are discussed in the context of accuracy and computational efficiency. The analytical solution performing best shows very small deviations in the stiffness matrix entries compared to BECAS (below 1 % in the majority of test cases). It achieved a better resolution of the stress distribution and a computation time that is more than an order of magnitude smaller using the same spatial discretization. The deviations of the stress distributions are below 10 % for most test cases. The analytical solution can thus be rated as a feasible approach for a beam-based pre-design of wind turbine rotor blades.

Initial Study of the Onsite Measurement of Flow Sensors on Turbine Blades (SOTB).

This paper presents a new framework using MEMS flow sensors on turbine blades (SOTB) to investigate unsteady flow features of a rotating wind turbine. Self-heating flow sensors were implemented by the U18 complementary metal-oxide semiconductor (CMOS) MEMS foundry provided by Taiwan Semiconductor Research Institute (TSRI). Flow sensor chips with a size of 1.5 mm × 1.5 mm were parylene-coated, output via a wireless data acquisition system (WDAQ), and mounted at the root, middle and tip of a 1.2 m diameter semi-rigid turbine blade of a 400 W horizontal axis wind turbine (HAWT). The instantaneous angles of attack (AOAs) of the SOTB were found to be 46~62°, much higher than the general stall AOA of 15°, but were accurate considering the normal detection of the flow sensors. The computational fluid dynamics (CFD) simulation of the HAWT was also compared with the SOTB output. The onsite measurement herein revealed that the 3D secondary flow increment, mostly obvious near the middle part of the turbine blades, degraded both the sensor and the turbine performance and initially justified the onsite measurement application.

Harnessing the power of thermal imagery and visual inspection- a mean for reliable damage detection of wind turbine rotor blades

Generation of green electricity as part of the energy transition is leading to a growing market in the wind energy sector all over the world. Maintenance and inspection are key to the reliability, safety and efficiency of wind turbines, the regular maintenance of rotor blades focuses on damage such as erosion on the leading edge of the profile, delamination and thermal cracks due to lightning strikes. To date, visual inspection by technicians (climbers) has been the state of the art and it is time consuming besides posing safety risk for themselves. Recently, drone-based inspections using visual cameras have become more common, enabling fast, reliable and cost-effective inspections. However, no internal damage to the rotor blades can be detected during such an inspection. Thermography is a recognised method for detecting damage beneath the surface of an object, which has been promoted and further developed at BAM for years. To enhance the accuracy and reliability of wind turbine blade inspection, the fusion of thermal and visual data has emerged as a promising technique. Here, a drone-based multisensory system is being presented that combines thermal imagery, high-resolution visual images and a three-dimensional image obtained from 3D Laser scanner to represent a comprehensive view of the blades' internal and external conditions. Besides this, flight data obtained from IMU, GPS and LiDAR sensors will be used for enhancing the inspection data. Thus, fusion of images enables a more detailed analysis of potential defects and this not only improves the overall effectiveness of the inspection process but also provides windmill operators a new dimensions for data analysis and decision making.

Continuum approach of nonlinear dynamic analysis of a rotating wind turbine blade considering hub motion

This paper reports the nonlinear dynamics of a large wind turbine blade considering the hub motion. The investigation is based on a continuum mathematical model, which is derived using the extended Hamilton principle. The proposed model is validated via comparison with both published results and numerical simulations obtained from open source software. The steady state amplitude curves as well as the dominating frequencies of the modal forces are compared and analyzed with discussions on the effects of the hub motion amplitude, yaw angle and rotational frequency of the blade. Comparison of dominant frequencies of modal forces shows that the skewed inflow can strengthen the periodic excitation whose frequency equals to the rotating frequency of the blade. The steady state amplitude curves show that vibration of the edge-wise mode is more significant than that of the flap-wise mode when under the hub motion excitation. Moreover, internal resonance between the first three modes is also observed. The skewed inflow has a significant effect on the first mode in terms of vibration amplitude and the most unfavorable direction of hub motion. Finally, recommendation on how to avoid the resonance and internal resonance of the blades under skewed inflow are given.

Wind turbine rotor blade encoding marker recognition method based on improved YOLOv8 model

Wind turbine rotor blade encoding marker recognition method based on improved YOLOv8 model

Development of a portable small wind turbine for integration into a mobile cooling technology

In this study, a small wind turbine prototype was developed to provide electric power for a mobile cooling unit The aim of this study was to design and develop a 600-W small wind turbine that can generate electric energy to power a mobile cooling unit used for the storage of fruits and vegetables, mainly for the benefit of smallholder farmers. Smallholder farmers suffer from high postharvest losses, approximated at 50%, some of which can be avoided by using efficient low-cost cooling units, rather than open transport. Cooling slows down the metabolic rate which consequently extends the produce's shelf life and prevents spoilage, allowing farmers to provide high-quality produce to the market. This could potentially increase the farmers’ monetary returns. The study was conducted in KwaZulu-Natal on the road that stretches between Pietermaritzburg and Estcourt. The wind turbine is made of a 600-mm-diameter rotor with three PVC blades, a permanent magnet synchronous generator, a bridge rectifier, a 230-V AC inverter and a battery for energy storage. The wind turbine was tested against three vehicle speeds of 60, 80, and 100 km h−1, and the two opening levels, level 1 at 45∘\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^\\circ$$\\end{document} and level 2 at 80∘\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$^\\circ$$\\end{document} relative to the louvre mechanism frame. The results of this study revealed that the power generated by the wind turbine is greatly influenced (p < 0.001) by both the vehicle travelling speed and louvre opening level. The power output of 113.4, 159.6 and 210.0 W per hour was observed for the vehicle speeds of 60, 80 and 100 km h−1, respectively, on louvre opening level 1. The power output of 142.8 W h−1, 268.8 W h−1 and 294.0 W h−1 were observed for a wind speed of 60 km h−1, 80 km h−1, and 100 km h−1, respectively, on Louvre opening level 2. This shows that higher wind speeds (vehicle speeds) produce high-power output which accounts for the small size of the wind turbine rotor. A maximum power coefficient of 0.49 was achieved for this study. The wind turbine can generate the power required to run a cooling technology to a limited extent, thus must have a backup power supply from the diesel engine or be used in a hybrid system.

Bladeless Wind Turbine Triboelectric Nanogenerator for Effectively Harvesting Random Gust Energy

AbstractIn the environment, there is an abundance of gust energy which is challenging to harvest with conventional rotating wind turbines, such as the gusts generated by passing vehicles along roadsides. Addressing the irregular and low‐frequency characteristics of gusts, a bladeless wind turbine triboelectric nanogenerator (BWT‐TENG) with enhanced aerodynamic performance is proposed, enabling effective harvesting of random gust energy. First, a bladeless wind turbine with a cylindrical bluff body shape is designed, and its aerodynamic principles under gust‐driven conditions are elucidated through the computational fluid dynamics method. Subsequently, parameter optimization is conducted for the multilayered TENG. Systematic experiments demonstrated that the BWT‐TENG achieved a peak power density of 4.08 W m−3 driven by a gust of 10 m s−1, and can even operate at frequencies as low as 0.1 Hz. Finally, experiments showcased the BWT‐TENG powering a warning light in a simulated rainfall environment and harvesting gust energy from vehicles passing by real roadside to power wireless gyroscopic sensors, thereby achieving self‐powered structural health monitoring of roads or bridges. This work provides a novel strategy for utilizing TENGs in the harvest of environmental gust energy and demonstrates the vast potential of TENGs in the field of self‐powered structural health monitoring.

Effect of Induction and Blade Elasticity Modelling on Wind Turbine Rotor Performance Predictions

This study investigates the impact of blade induction modelling on the accuracy of wind turbine rotor aeroelastic predictions. It extends the capabilities of AEOLIAN (AErOeLastic sImulAtioN), a Fluid Structure Interaction (FSI) solver based on Blade Element Momentum Theory (BEMT) coupled with a Lumped Mass approach to represent the blade structure. Herein, AEOLIAN’s analytical wake induction engineering model is replaced with the outcomes of a physically-consistent three-dimensional Free-Vortex Wake (FVW) formulation initially employed in AeroROTOR. This versatile aeroelastic simulation tool is implemented within the framework of MATLAB Simulink/Simscape-Multibody©, a modular environment suitable for industry analysts, researchers, and academic users focusing on wind turbine aero-servo-elastic applications. Furthermore, it serves to lay the groundwork for the development of advanced control laws for multi-megawatt rotors, fostering innovation in the design and optimization of the next-generation wind turbines. The presented analyses focus on predicting the aeroelastic behavior of the bottom-fixed NREL 5MW rotor in uniform axial flow over the operating range, complemented by more detailed investigations at the rated condition undergoing inflow with/without wind misalignment (yaw). The study on key performance parameters is conducted by comparing with the higher-fidelity data from available Computational Fluid Dynamics (CFD) and Computational Structural Dynamics (CSD) coupled with CFD.

Static Testing Using OFDR-based Quasi Continuous Fiber Optic Strain Sensing on the WiValdi Wind Turbine Rotor Blade

The German research wind farm WiValdi was taken into service in 2023 and enables researchers and industry to conduct full-scale experiments. It currently consists of two highly instrumented wind turbines and multiple wind measurement masts. In one of the six wind turbine rotor blades, four Fiber Optic Strain Sensors (FOSS) are installed [1].The strain which a wind turbine rotor blade exhibits under load is a key factor in the design, the certification and the assessment of its remaining service life as well as its further development and optimization. The aerodynamic optimization of wind turbine rotor blades leads to longer, and increasingly slender rotor blades and therefore increasingly relevant static and dynamic structural mechanical challenges.Usually the strain of wind turbine rotor blades during static deflection tests is measured with electrical strain gauges (SGs) or Fiber-Bragg-Gratings (FBGs). In the present study, quasi-continuous Fiber-Optic-Strain-Sensors (FOSS) are used in parallel with FBGs. The underlying technology is called optical frequency domain reflectometry (OFDR) and since only glass-fibers are used, it is inherently lightning protected [2]. Besides that, it is commercially available and delivers strain results at a spatial resolution of 2.6 mm. For this study, there are four 50-meter-long glass fibers bonded to the wind turbine rotor blade on suction and pressure side of each main spar of a 57 m rotor blade. This contribution shows how the FOSS measurements are post-processed and validated. The validated data is then used to obtain the relation between strain and introduced force over the length of each FOSS. With the accurate prediction of root-bending-moments, one of many possible use-cases of the FOSS results, is presented.Besides the computational model updating of the corresponding Finite-Element model in future investigations, the results will also be used to judge the current structural mechanical state of the wind turbine rotor blade at recurring service inspections. In that, results of the then to be acquired state at a given time and the shown base-line state will be compared. This will lead to high resolution insight into the deterioration of the structural integrity of the wind turbine rotor blade over time.Considering that standard telecommunication glass fibers can be used as sensors, the cost per sensor (i.e. fiber) is relatively low (circa 1500 € for 50 m). Despite the high cost of the required measurement system, the low prices of the fibers make FOSS measurements not only a scientifically interesting but also a potentially economically viable solution for measuring strain on wind turbine rotor blades.

Rotor Performance and Wake Interaction of Controlled Dual Surging FOWT Rotors in Tandem

Using Large Eddy Simulation (LES) with Actuator Line Model (ALM), this work investigates the system of two surging wind turbine rotors operating under realistic turbulent inflow conditions (TI = 5.3%). The two rotors are placed in tandem with a spacing of 5D and the surging motions are harmonic. A widely used torque controlling strategy, MPPT (Maximum Power Point Tracking), is implemented to ensure a maximium power extraction under all conditions. The rotor performances as well as the field data are surveyed to examine the effectiveness and impacts of the controller. It is found that the power performances of the surging rotors are benefited by the controller with a small margin (∼ 1%) when the surging motions are moderate. The results also show that the controller reacts much slower than the considered surging frequency, making the power performances of the rotors worse than the quasi-steady predictions (targeted values) and complicating the system dynamics. In general, the implementation of the controller has minor impacts on the wake characteristics; however, the strengths of Surging Induced Periodic Coherent Structures (SIPCS) are found to be enhanced.

Investigation of a modified IEA 15MW wind turbine including a winglet with a new mid-fidelity acoustic toolchain

This paper presents the applicability of a new hybrid tool chain for wind turbine rotor turbulent boundary-layer trailing-edge noise (TBL-TEN) to complex rotor geometries. The research introduces a rapid, high-fidelity 2D CFD/CAA method, which enables the assessment of innovative low-noise profile designs concerning TBL-TEN, coupled with a mid-fidelity extrapolation technique for the 3D analysis, aiming to minimize computational time essential for design processes. This procedure is applied here to compare the changes in performance, load distribution and noise reduction if a winglet is added to a large-scale wind turbine, using the example of the IEA Wind TCP Task 47 15-MW-reference turbine [12]. In the first step, a 3D CFD calculation of the entire turbine is performed in order to derive the local flow parameters and thus the input variables for the 2D CAA. These results for the 2D TEN are then extrapolated for the entire turbine using the fast summation and extrapolation method TAP (Turbine acoustic prediction) [1]. As only data for the TEN is included here, only this noise source for the winglet can be analyzed. For a more complete comparison, a 3D scale-resolving simulation was set up for a static wind tunnel situation in order to be able to qualitatively classify the tip/vortex interaction noise that was not recorded by the procedure mentioned above.

Aeroelastic analysis of a very large wind turbine in various atmospheric stability conditions

With the growing trend towards larger wind turbine rotor diameters, the impact of wind shear on rotor performance and loads becomes increasingly significant. Atmospheric stability strongly influences wind shear, leading to higher wind shear under stable atmospheric conditions. In this study, the aeroelastic performance of the IEA 22 MW rotor is assessed under inflow conditions generated by different methods. Inflow conditions were generated using turbulence models specified in the IEC Standards and also by Large Eddy simulations. Standalone OpenFAST simulations were conducted with the respective inflow conditions.It was found that at rated and above-rated wind speeds, the time-averaged wind turbine design loads were higher in stable atmospheric conditions, in comparison to the IEC NTM inflow conditions, while the opposite held for below-rated wind speeds. Specifically, the time-averaged root flapwise bending moment and rotor thrust were found to be higher by up to 7% in stable atmospheres. However, maximum design and fatigue loads were considerably higher in the IEC NTM case due to elevated turbulence levels. Compared to the IEC NTM case, the damage equivalent root flapwise bending moment was found to be 30% to 70% lower in the different scenarios.

Experimental and numerical comparison of aerodynamics for a wind turbine rotor model under turbulent inflow conditions

Computational prediction of aerodynamics for a two-bladed wind turbine rotor model at the profile, blade, and rotor scale are compared to wind tunnel experimental data. Wind tunnel tests are carried out in the large subsonic wind tunnel at the University of Orleans with an instrumented rotor model to obtain power and thrust coefficients, bending moments and pressure chordwise distributions. A passive uniform turbulence grid is used to obtain a turbulence intensity of 3.8% at the front of the rotor. A k − ω SST unsteady Reynolds-averaged Navier-Stockes (URANS) simulation with the ISIS-CFD flow solver is conducted with a fully resolved rotor, with and without automatic grid refinement. The focus of this study is to examine the prediction capabilities of the simulation for various physical variables. The simulation demonstrates a relatively good prediction of power and thrust coefficients compared with experimental data with a maximum scatter of 8.5%. The prediction of the pressure coefficient distribution at three different radial positions is satisfactory, but shows discrepancies when it comes to accurately predicting flow separation and Reynolds effects.

How does the blade element momentum method see swept or prebent blades?

The blade element momentum (BEM) method is among the mostly used engineering aerodynamic models for wind turbine rotors. Even though the BEM method is strictly only applicable to straight blades forming a planar rotor, it is in practice used for load calculation and design optimization of blades with relatively large sweep, prebend or coning. The present work aims to answer the question: How does the BEM method see swept or prebent blades? The blade element part of the method can be adapted for swept or prebent blades by projecting the velocity and loads between the 2-D airfoil section and the 3-D blade geometry under the cross-flow principle. However, momentum theory considers the 3-D wake and the induction of a planar actuator disc with straight blades. There could be significant differences between the results calculated from different implementations of the BEM method, while the details of the implementations are often omitted in the literature. In the present work, a consistent implementation including the definition of airfoil section alignments and the coordinate systems is provided. It is shown that for a given circulation distribution, the BEM method will predict approximately the same local aerodynamic loads for non-straight blades as if the blades were straight and the rotor was planar. This means using the BEM method to perform aerodynamic optimization of modern rotors with curved blades and/or rotors that are not planar can only have little meaningful result. In order to model the impact of the wake geometry on the induction, more advanced models that take into account the wake geometry on the induction are necessary.

Effects of wind turbine rotor tilt on large-scale wind farms

Recent studies have explored the use of rotor tilt adjustments to reduce wake losses in wind farms. While downward wake deflection in aligned wind farms has shown promise for significant power gains, the impact of wind farm layout on the effectiveness of tilt strategies is not yet fully understood. Additionally, the effect on downstream farms remains unclear. Our large eddy simulations reveal that a rotor tilt of 20 degrees significantly reduces wake losses in aligned wind farms. For wind farms with 8 turbine rows, we observe an overall increase in wind farm productivity of up to 11%. However, tilting the rotors may decrease power production in staggered wind farms, where wake losses are inherently lower due to the increased spacing between turbines. Our findings also suggest that a downstream wind farm might benefit from an upstream farm implementing rotor tilt, although this advantage is primarily observed in the first row of the downstream farm.

Improved performance of k − ω SST turbulence model in predicting airfoil characteristics for a wide range of airfoil thicknesses

A variety of airfoils designed for wind turbine rotor blade applications were simulated with k − ω shear stress transport (SST) turbulence model in its standard form, the k − ω SST turbulence model with a modified a 1 constant (referred to as the a 1 method), the Spalart-Allmaras turbulence model, and the panel method XFoil. To assess their performance, the results of airfoil lift, drag, and coefficient of pressure were compared against available wind tunnel data, where available. The standard k −ω SST turbulence model is found to over-predict Reynolds shear stresses, delay the flow separation, and under-predict the separated-flow region on the airfoil’s suction side. Airfoil thicknesses between 11% and 36% of the chord length were studied. Using a modified a 1 constant, some airfoils exhibited up to a 20% improvement in the prediction of lift and drag coefficients within the post-stall range of angle of attack (AoA). It is important to note that the a 1 method’s applicability and effectiveness is specifically tested for airfoils, and its performance is highly dependent on the airfoil geometry.

A Hybrid Testing Framework for Wind Turbine Mechanical Components

As wind turbines continue to increase in size, hybrid testing is emerging as a key-enabling solution for experimental assessment of their large mechanical components. Hybrid testing is conducted using a hybrid model which combines physically tested and numerically simulated components. In this work, the Kane method is proposed to formulate the equation of motion for the hybrid model of a wind turbine rotor system where one single blade is physically tested. The Kane method allows for formulating the equation of motion of multi-body-dynamic models efficiently and, therefore, it is widely used in state-of-art simulation software. The hybrid model of the rotor is successfully implemented on a single-degree-of-freedom test bench with a cantilever steel beam serving as the physical substructure. The performance of the implemented hybrid model is assessed through a comparison with a pure numerical simulation of the same system. The main finding of the study emphasizes the efficiency of incorporating physically measured restoring force as model parameters while formulating the equation of motion of a hybrid model.

Buckling behavior of rotor blade sandwich panels with spatially distributed material uncertainties

The study evaluates the impact of material uncertainties on the buckling behavior of sandwich panels in wind turbine rotor blades. The analysis is limited to linear buckling and is performed using stochastic finite element Monte Carlo simulation on a rectangular and flat submodel of the rotor blade’s trailing edge panel. The finite element model of the panels is simply supported on all edges. To generate the spatial material property distributions, the Karhunen-Loève expansion is used in combination with Latin hypercube sampling. The results compare the effects of various correlation lengths of the spatial distributions. The buckling loads vary in correlation to the average panel stiffness caused by the random distributions. The spatial distribution has a less dominant effect, reducing the mean value of the buckling load results. The amount of reduction in buckling load is highest when the correlation length of the distribution is close to the harmonic half-wave of the dominant buckling shape.

A grid deformation method for wind turbine CFD applications

The desire for low cost of energy is driving the development of modern wind turbines toward more and more flexible designs. With respect to aerodynamic evaluation of these designs based on CFD this pose a requirement for the capability to compute the flow on deforming grids around heavily deforming geometries. The present paper describes an efficient explicit method suited for deforming geometries in a structured grid approach, here illustrated for airfoil and a wind turbine rotor simulations. Based on the present investigation, it is illustrated that the solution on the deformed grids deviates less than a fraction of a percent from the solution on the original grids both for the airfoil and the rotor case.

Towards LES-RANS/LES coupling for simulation of a wind turbine rotor in a nocturnal atmospheric condition

A segregated approach with either synthetically generated velocity fluctuations or a precursor Large-Eddy Simulation (LES) of the atmospheric boundary layer (ABL) and a subsequent hybrid RANS/LES (HRLS) of the resolved rotor is presented in this paper. At first the approach is validated with synthetically generated velocity fluctuations that resemble neutrally stratified conditions. Subsequently the same turbine is artificially exposed to a stably stratified atmospheric inflow (nocturnal boundary layer) with a comparable mean velocity but different shear and veer characteristics. The HRLS simulations with the neutrally stratified and stably stratified inflow are compared to LESs with an implemented actuator disc (AD) model. The presented work is a first important step towards a digital twin for research wind parks.