A set of linear equations describing the motion of an operating 3-bladed HAWT are obtained from the dynamic characteristics of the stationary turbine by adding rotating frame effects. The approach makes use of the Coleman multi-blade transformation to present all results relative to the fixed frame. The formulation is in terms of a selected number of stationary, real, mode shapes. The formulation is applied to the expression of both the aerodynamic loading and the displacement response in terms of the operating mode shapes. This technique is applied to the conditions of vertical wind shear and off-yaw operation of a hypothetical 46-m wind turbine. The principal objective of the paper is to enable the characteristic of the inflow to be related to the nature of the response. A second objective is to illustrate a method of extracting linearized models from general aeroelastic codes such as ADAMS™.

Completion of the full-scale wind tunnel tests of the NREL Unsteady Aerodynamics Experiment (UAE) Phase VI allowed validation of the AeroDyn wind tuxbine aerodynamics software to commence. Detailed knowledge of the inflow to the UAE was the bane of prior attempts to accomplish any in-depth validation in the past. The wind tunnel tests permitted unprecedented control and measurement of inflow to the UAE rotor. The data collected from these UAE tests are currently under investigation as part of an effort to better understand wind turbine rotor aerodynamics in order to improve aero-elastic modeling techniques. Preliminary results from this study using the AeroDyn subroutines are presented, pointing to several avenues toward improvement. Test data indicate that rotational effects cause more static stall delay over a larger portion of the blades than predicted by current methods. Despite the relatively stiff properties of the UAE, vibration modes appear to influence the aerodynamic forces and system loads. AeroDyn adequately predicts dynamic stall hysteresis loops when appropriate steady, 2-D airfoil tables are used. Problems encountered include uncertainties in converting measured inflow angle to angle of attack for the UAE phase VI. Future work is proposed to address this angle of attack problem and to analyze a slightly more complex dynamics model that incorporates some of the structural vibration modes evident in the test data.

The effect of varying turbulence levels on long-term loads extrapolation techniques was examined using a joint probability density function of both mean wind speed and turbulence level for loads calculations. The turbulence level has a dramatic effect on the statistics of moment maxima extracted from aeroelastic simulations. Maxima from simulations at lower turbulence levels are more deterministic and become dominated by the stochastic component as turbulence level increases. Short-term probability distributions were calculated using four different moment-based fitting methods. Several hundred of these distributions were used to calculate a long-term probability function. From the long-term probability, 1- and 50-year extreme loads were estimated. As an alternative, using a normal distribution of turbulence level produced a long-term load comparable to that of a log-normal distribution and may be more straightforward to implement. A parametric model of the moments was also used to estimate the extreme loads. The parametric model predicted nearly identical loads to the empirical model and required less data. An input extrapolation technique was also examined. Extrapolating the turbulence level prior to input into the aeroelastic code simplifies the loads extrapolation procedure but, in this case, produces loads lower than the empirical model and may be non-conservative in general.

This paper presents a numerical method for performance predictions of wind turbines immersed into stable, neutral, or unstable atmospheric boundary layer. Tile flowfield around a turbine is described by the Reynolds’ averaged Navier-Stokes equations complemented by the k-ε turbulence model. The density variations are introduced into the momentum equation using the Boussinesq approximation and appropriate buoyancy terms are included into the k and ε equations. An original expression for the closure coefficient related to the buoyancy production term is proposed in order to improve the accuracy of the simulations. The turbine is idealized as actuator disk surface, on which external surficial forces exerted by the turbine blade on the flow are prescribed according to the blade element theory. The resulting mathematical model has been implemented in FLUENT. The results presented in the paper include the power output and wake development under various thermal stratifications of an isolated wind turbine. In stable stratification, the power output is 4% lower than in neutral condition, while in unstable situation, the power is 3% larger. The predicted wake velocity defects are qualitatively in agreement with experimental observations.

Active flow control and load mitigation concepts developed for traditional aeronautical applications have potential to decrease torque, bending and fatigue loads on wind turbine blades and to help increase turbine life. Much of the early work in flow control focused on steady aerodynamic benefits. More recent technologies have focused on unsteady flow control techniques which require a deeper understanding of the underlying flow physics as well as sensors to record the various time-dependent aerodynamic phenomena and fast actuators for control. This paper identifies some developmental control concepts for load mitigation along with a new translational microfabricated tab concept available for active flow and load control on lifting surfaces and explores their applicability for wind turbine rotor blades. Specifically, this paper focuses on experimental results based on an innovative microtab approach for unsteady, active load control. Previous papers on this effort by Yen et al. focused on the multi-disciplinary design methodology and the significant lift enhancement achieved using these micro-scale devices. The current research extends the effort to include dynamic results with discontinuous tab effects, effects on drag, and lower (pressure side) and upper surface (suction side) tab deployment effects for the prototype airfoil as well as for the S809, a representative wind turbine airfoil. Results show that the microtab concept can provide macro-scale load changes and is capable of offering active control of lift and drag forces for load alleviation.

The Long-term Inflow and Structural Test (LIST) program is collecting long-term, continuous inflow and structural response data to characterize the spectrum of loads on wind turbines. A heavily instrumented Micon 65/13M turbine with SERI 8m blades is being used as the primary test turbine for this test. This turbine is located in Bushland, TX, a test site that exposes the turbine to a wind regime representative of a Great Plains commercial site. The turbine and inflow are being characterized with 60 measurements: 34 to characterize the inflow, 19 to characterize structural response, and 7 to characterize the time-varying state of the turbine. In this paper, the inflow and structural data from this measurement campaign are analyzed to determine the correlation of various inflow descriptors with fatigue loads. The inflow is described by various parameters, including the mean, standard deviation, skewness and kurtosis of the wind speed, turbulence intensity, turbulence length scales, Reynolds stresses, local friction velocity, Obukhov length and the gradient Richardson number. The fatigue load spectrum corresponding to these parameters is characterized as an equivalent fatigue load. A regression analysis is then used to determine which parameters are correlated to the fatigue loads. The results illustrate that the vertical component of the inflow is the most important of the secondary inflow parameters on fatigue loads. Long-term fatigue spectra illustrate that extrapolation of relatively short-term data to longer times is consistent for the data reported here.

The aim of this work is two-fold. Firstly, 28 sets of airfoil (widely used for wind turbine applications) measurements were compared with numerical results from a 2D Navier-Stokes solver and a panel method code. These results have been collected into an airfoil catalogue that has been separately published. Secondly, based on the previous results, criterions for evaluating the airfoils are derived. Thereby, the performance of the Navier-Stokes solver is evaluated. Further analysis of the results determines geometrical and flow properties that may cause problems when computing airfoil flows with a Navier-Stokes solver, and some recommendations are given.

This paper presents an extension of the work described in Reference 1, which dealt only with tensile fatigue loads spectra. Experimental results and modeling have now been extended to simple spectra which cover a full range of loads including compression and reversed tension-compression, as well as WISPERX and other spectra representative of wind turbines, containing a full range of R-values. Experimental fatigue data (over 1100 tests) have been generated for a fiberglass laminate which is typical of wind turbine blade construction, tested under a variety of loads spectra. Constant amplitude data typical of the DOE/MSU fatigue database are then used with various modeling schemes to predict the lifetime under spectrum loading. Several areas of current concern in blade lifetime prediction are explored. First, it is demonstrated that the Miner’s Rule cumulative damage law results in significantly non-conservative lifetime predictions for most loads spectra. Linear and nonlinear residual strength models are easily applied and are more accurate in predicting lifetime; they also provide a prediction of remaining strength as a function of the history of loading prior to failure. Prediction accuracy is also sensitive to the model (exponential or power law) used to represent and extrapolate the constant amplitude S-N fatigue data. A power law representation appears to provide more accurate predictions of lifetime under spectral loading for longer lifetime cases.

A numerical algorithm for acoustic noise generation is extended to 3D flows. The approach involves two parts comprising a viscous incompressible flow part and an inviscid acoustic part. In order to simulate noise generated from a wind turbine, the incompressible and acoustic equations are written in polar coordinates. The developed algorithm is combined with a so-called actuator-line technique in which the loading is distributed along lines representing the blade forces. Computations are carried out for the 500kW Nordtank wind turbine equipped with three LM19 blades.

Possible alternative fatigue formulations to predict lifetime under variable-amplitude loading are investigated. Test results of WISPER and WISPERX variable-amplitude tests on a material representative for wind turbine rotor blades are used. All fatigue calculations are performed using Rainflow counting of the WISPER(X) load histories and employing the Miner summation. The formulation of the SN-curve and the constant-life diagram are varied. Commonly, a log-log SN-curve is used in combination with a linear Goodman constant-life relation. However, in previous work, it was found that these formulations overestimate lifetime of specimens subjected to the variable-amplitude WISPER and WISPERX load histories. This previous work suggested that the SN-formulation be changed and also used an alternative constant-life formulation with parallel lines. These formulations and variations on them are investigated. Also, constant-amplitude data for R = 0.1 are included to construct an alternative constant-life diagram. Including R = 0.1 constant-amplitude data in the lifetime predictions for WISPER(X) seems to improve the accuracy of the calculation. The alternative constant-life formulation might remove the non-conservatism from the lifetime prediction and account for the difference in lifetime between WISPER and WISPERX.

In the IEC 61400-12 standard “Wind turbine power performance testing”, the availability of a wind turbine is defined as: “ the ratio of the total number of hours during a certain period, excluding the number of hours that the wind turbine could not be operated due to maintenance or fault situations, to the total number of hours in the period, expressed as a percentage ” [1]. It seems to be a clear definition, however in practice all parties involved, and often commercially bounded by guarantees, cannot interpret it unambiguously. This has all to do with the fact that the actual measuring and tracking is not done in a structured manner, mainly because it is not clear what the various (sub-)states of a wind turbine can be, and which sub-states have to be considered as available or not, and also for discerning who is liable. Moreover, due to heavy operation conditions and potential unintended built-in weaknesses, the ultimate availability of a wind turbine measured over the lifetime of 20 years, can easily tail off by 10% to 25%. Directly related to availability, the return on investment (ROI) will even decrease with a higher proportional part. Consequently, availability is a serious issue for manufacturers and investors, and may have extensive commercial implications. To better understand and improve the availability, from the beginning of the design process, this article presents the ‘ State Model for Wind Turbines ’. It underpins the strategy to optimize the availability versus the total costs. Furthermore the article provides some practical tools for specifying, design-in and measuring of the final availability.

The design of a robust frequency converter controller for high dynamic performance of a synchronous generator requires an accurate dynamic model of the electromagnetic part. In this paper a new procedure for identifying the transfer functions of Park’s dq -axis model of a synchronous generator has been developed. It will be shown that the parameters of this model can be easily identified from standstill time-domain data. The validity of the theoretical model has been verified by comparing time-domain simulations with measurements taken from the Lagerwey LW-50/750 direct-drive synchronous generator. It can be concluded that a consistent model estimate of the electromagnetic part of the LW-50/750 generator has been obtained. Ultimate validation, however, will follow after the implementation of the designed frequency converter controller in this wind turbine.

Due to the increasing size of wind turbine rotors, especially for offshore wind turbines, the power output and the blade mass are becoming more important. A new type of cheaper carbon fibres is expected to result in a lower blade mass and marginally cheaper blades. In a Joule-funded project, the possibilities for economic use of carbon fibres is determined by establishing material design data, analysing production methods, developing cost-effective blade root joints and assessing blade and turbine costs. The R&D project will be finalised by the end of 2001. Up to now, production processes and promising material combinations have been reviewed, tested and ranked. Basic material design data have been established for the two most-promising material combinations. Due to disappointing fatigue results on Panex/VE, additional testing on four large-tow laminates was performed. The fatigue properties of these showed to be consistent. Later testing on a similar Panex/epoxy laminate, however, revealed much better fatigue behaviour. Joint development and cost assessment are underway and show promising results.

This paper concerns an experimental investigation of a NACA 0015 airfoil subject to harmonic one-degree-of-freedom translatory motion. Specifically, unsteady pressure distributions were measured at a range of incidences and movement directions at reduced frequencies matching real life conditions for the lead-lag motion of wind turbine rotors. From the experimental results, hysteresis loops and aerodynamic damping were computed and compared to results from linear quasi-stationary theory and unsteady potential flow theory. The maximum negative aerodynamic damping was found to take place at moderate stall and an incidence of about 15 ° , at a movement direction close to the chordwise direction. Comparison with unsteady potential flow theory showed excellent agreement with the experimental data for incidences up to 5 ° . Linear quasi-stationary theory failed to reproduce the overall features of the aerodynamic damping for incidences above 12 ° .

The aeroelastic stability of a three-bladed wind turbine is considered with respect to classical flutter . Previous studies have shown that the risk of stall-induced vibrations of turbine blades is related to the dynamics of the complete turbine, for example does the aerodynamic damping of a rotor whirling mode depend highly on the tower stiffness. The results of this paper indicate that the turbine dynamics also affect the risk of flutter. The study is based on an eigenvalue analysis of a linear aeroelastic turbine model. In an example of a MW sized turbine, the critical frequency of the first torsional blade mode is determined for which flutter can occur under normal operation conditions. It is shown that this critical torsional frequency is higher when the blades are interacting through the hub with the remaining turbine, than when all blades are rigidly clamped at the root. Thus, the dynamics of the turbine has increased the risk of flutter.

A wind power system differs from a conventional power system. In a conventional power plant, the operator can control the plant’s output. The output of a wind farm cannot be controlled because the output fluctuates with the wind. In this study, we investigated only the fixed-frequency induction generator, often used with wind turbines. We adopted the worst-case scenario and conducted a per-phase, per-turbine analysis. Our analysis showed a strong interaction among the wind farm, the utility grid, and the individual generator. In this paper, we investigate the power-system interaction resulting from power variations at wind farms using steady-state analysis. We use the characteristic of a real windsite on a known weak grid. We present different types of capacitor compensations and use phasor diagrams to illustrate the characteristics of these compensations. The purpose of our study is to provide wind farm developers with somc insights on wind farm power systems.

Many of the aerodynamic phenomena contributing to the observed effects on wind turbines are now known, but the details of the flow are still poorly understood and are challenging to predict accurately, issues discussed herein include the modeling of the induced velocity field produced by the vortical wake behind the turbine, the various unsteady aerodynamic issues associated with the blade sections, and the intricacies of dynamic stall. Fundamental limits exist in the capabilities of all models, and misunderstandings or ambiguities can also arise in how these models should be properly applied. A challenge for analysts is to use complementary experimental measurements and modeling techniques to better understand the aerodynamic problems found on wind turbines, and to develop more rigorous models with wider ranges of application.

This paper presents a framework for the dynamic and aeroelastic analysis of a horizontal axis wind turbine modeled as a multi-flexible-body system. The multi-rigid-body portions of the system, composed of the nacelle and hub, are modeled as a system of interconnected rigid bodies using Kane’s equations. The flexible portions, composed of the the blades and tower, are represented using nonlinear beam finite elements, taken from a mixed formulation for the dynamics of moving beams. Each analysis leads to a set of symbolic equations that can be coupled symbolically to represent the dynamic behavior of the wind turbine. A solution procedure is implemented to assess the dynamic stability of the system. Here the solution is divided into two parts: a set of nonlinear ordinary differential equations governing the periodic steady-state operating condition, and a set of equations that are linearized about the steady-state operating condition governing the transient dynamics. The harmonic balance method is used for the nonlinear periodic steady-state solution, and the finite element in time method is proposed as an alternative method. Linearization of the equations about the steady-state operating condition yields system equations with periodic coefficients which are solved by Floquet approach to extract the modal parameters. For the aeroelastic analysis, aerodynamic loads from an aerodynamic model to be selected in the future will be inserted into the present framework. Then, the framework can produce a symbolic system matrix, potentially useful for control design. Numerical results are presented for the dynamic characteristic of HAWT’s with flexible tower and blades.

Most small wind turbines use permanent magnet (PM) generators. The generators are usually direct-drive (i.e., no gearbox is required). Direct-drive PM generators are characterized by low maintenance and high efficiency. Small wind turbines are usually self-starting and require very simple controls. Cogging torque is an inherent characteristic of PM generators and is caused by the geometry of the generator. Cogging torque affects self-start ability and produces noise and mechanical vibration. Thus, minimizing cogging torque is important in improving the operation of small wind turbines. In this paper, we investigate three design options to minimize cogging torque: uniformity of air gap, pole width, and skewing. Although the design improvement is intended for small wind turbines, it is also applicable to larger wind turbines.

Wind turbine blade certification tests, comprising a static test, a fatigue test, and finally a residual strength test, often involve sudden audible cracking sounds from somewhere within the blade, without the operators being able to locate the noise source, or to determine whether damage (minor or major) has occurred. A current EC-funded research project is looking at the possibility of using acoustic emission (AE) monitoring during testing of fibre composite blades to detect such events and assess the blade condition. AE can both locate and characterise damage processes in blades, starting with non-audible signals occurring due to damage propagation at relatively low loads. The test methodology is discussed in the context of the blade certification procedure and results are presented from a series of static and fatigue blade tests to failure in the laboratory. Inferences are drawn about small differences in the manufacture of the nominally identical blades and conclusions are presented for the application of the methodology.