• Energy Research

The results of surface pressure measurements are presented in this paper to gain further insight into the lift changing influence of finite width micro-tabs, especially in adjacent airfoil sections. Micro-tabs are a promising concept for load control on wind turbines. Local pressure distributions were measured in several rows of pressure taps in the vicinity of the finite width micro-tab attached to a FX 63-137 profile at low Reynolds numbers. The investigation focuses on length dependency, chordwise position, and interaction between two micro-tabs. Additionally, stereo Particle-Image-Velocimetry measurements were conducted to study the structure, sense of rotation, and influence of tab-induced tip vortices, as well as the impact of a finite width micro-tab on the model’s near wake. Experiments reveal relative changes of more than 30 % in the pressure coefficient distribution upstream of several micro-tab configurations. Furthermore, increments of 20 % are recorded in neighbouring sections not directly controlled by micro-tabs. Even higher changes are obtained in the region between two tabs. These improvements are attained due to local and global changes in the effective camber.

Abstract An aerodynamic load control concept termed “adaptive blowing” was successfully tested on a NACA 0018 airfoil model at Reynolds numbers ranging from 1.5·10 5 to 5·10 5 . The global objective was to eliminate lift oscillations typically encountered on wind turbine blade sections. Depending on the jet momentum flux, steady blowing from a control slot in the leading-edge region can be utilized to either enhance or reduce lift by suppressing or inducing boundary layer separation respectively. Furthermore, high momentum blowing effectively eliminated the dynamic stall vortex during deep dynamic stall conditions. Based on these previous findings, the present work explores the feasibility of controlling unsteady aerodynamic loads by dynamically varying the jet momentum flux to compensate for transient changes of the inflow. Various scenarios including high amplitude pitching, rapid freestream oscillations and combinations of both were investigated in a custom-built unsteady wind tunnel facility. An iterative control algorithm was implemented which successfully identified the momentum coefficient time profiles required to minimize the lift excursions. The combination of fully suppressing dynamic stall and dynamically adjusting the lift coefficient provided an unprecedented control authority, producing virtually constant phase averaged lift in all cases.

This paper describes the introduction of an unsteady aerodynamics model applicable for horizontal and vertical axis wind turbines (HAWT/VAWT) into the advanced blade design and simulation code QBlade, developed at the HFI of the TU Berlin. The software contains a module based on lifting line theory including a free vortex wake algorithm (LLFVW) which has recently been coupled to the structural solver of FAST to allow for time-resolved aeroelastic simulations of large, flexible wind turbine blades. The aerodynamic model yields an accuracy improvement with respect to Blade Element Momentum (BEM) theory and a more practical approach compared to higher fidelity methods such as Computational Fluid Dynamics (CFD) which are too computationally demanding for load case calculations. To capture the dynamics of flow separation, a semi-empirical method based on the Beddoes-Leishman model now extends the simple table lookups of static polar data by predicting the unsteady lift and drag coefficients from steady data and the current state of motion. The model modifications for wind turbines and the coupling to QBlade’s vortex method are described. A 2D validation of the implementation is presented in this paper to demonstrate the capability and reliability of the resulting simulation scheme. The applicability of the model is shown for exemplary HAWT and VAWT test cases. The modelling of the dynamic stall vortex, the empiric model constants as well as the influence of the dynamic coefficients on performance predictions are investigated.

Wind turbines generally suffer from unsteady inflow caused by yaw misalignment, gusts, and turbulence which induce fatigue loads. Spanwise distributed active micro-tabs at the mid and outer blade regions are able to countervail these unsteady loads. However, during the actuation process of these devices, transient effects play an important role. This work aims to give a deeper insight in the process of the tab deployment and retraction to evaluate the effectiveness of active micro-tabs for load control on wind turbines. Wind tunnel experiments on a two-dimensional NACA 0018 airfoil with an active micro-tab were conducted. The tab deployment- and retraction time was varied for an application on the suction or the pressure side of the airfoil. Time resolved surface pressure measurements were performed at Reynolds numbers of Re = 7 · 105 and 1 · 106. Transient responses showed a significant delay and post deployment behavior of the lift which strongly depend on the actuation time.

AbstractThis paper presents an experimental assessment of a blended fatigue‐extreme controller for load control employing trailing edge flaps on a lab‐scale wind turbine. The controller blends between a repetitive model predictive controller that targets fatigue loads and a dedicated extreme load controller, which consists of a simple on‐off load control strategy. The Fatigue controller uses the flapwise blade root bending moments of the three blades as input sensors. The Extreme controller additionally uses on‐blade angle of attack and velocity measurements as well as acceleration measurements to detect extreme events and to allow for a fast reaction. The experiments are conducted on the Berlin Research Turbine within the large wind tunnel of the TU Berlin. In order to reproduce test cases with deterministic extreme wind conditions that follow industry standards, the wind tunnel was redesigned. The analyzed test cases are extreme direction change, extreme coherent gust, extreme operating gust and extreme coherent gust with direction change. The test cases are analyzed by on‐blade angle of attack and velocity measurements. The load control performance of the Blended controller is compared to the pure fatigue oriented and the pure extreme load controller. The Blended controller achieves a maximum flapwise blade root bending moment reduction of 23%, which is comparable to the reduction achieved by the Extreme controller.

Abstract Dynamic stall on both horizontal axis and vertical axis wind turbine blades is accompanied by simultaneous changes in pitch and surge, but this simultaneous effect has never been documented. Using a unique unsteady wind tunnel, synchronous oscillations in angle of attack and flow speed were considered on two prototypical wind turbine blades. At a steady freestream, the concept of matched pitch rate was observed to be valid for large positive and negative pitch angles. In the presence of an unsteady stream, matching the flow speed as well as the pitch angle and its time derivative during pitch-up produced excellent correspondence between lift, drag and moment coefficients throughout the entire dynamic stall event.

This paper presents an evaluation of XFOIL and a commercially available CFD solver to predict the two-dimensional lift and drag coefficients of wind turbine airfoil sections for attached and separated flows. The computational solutions are correlated with the experimental data for the DU 96-W-180 airfoil that has been generated from wind tunnel testing performed at TU Delft and TU Berlin. CFD solutions are presented for turbulent calculations using the Shear-Stress Transport (SST) [1] and Spalart-Allmaras [2] turbulence models and transition calculations using the SST γ-θ model. Transition points from the CFD simulations are compared to the results obtained using XFOIL [3]. The paper culminates in a quantitative analysis identifying the deviations of the XFOIL and CFD solutions from the experimental data. This analysis uses the experimental polars generated by TU Delft and TU Berlin as a baseline for the comparison with the end goal of determining the best computational source of design polars for use in industry.

Wind turbines are classically designed for an extremely long lifetime in machinery design terms, for example, the Siemens SWT-6.0-154 was recently certified for 25 years [1]. The implication is that wind turbines accumulate damage via a number of mechanisms. A primary concern is naturally fatigue, exacerbated by a long life (high number of cycles); however, environmental effects such as bio-fouling and leading edge erosion damage the structure but also modify the Lift and Drag characteristics, particularly the stall behaviour. Vortex Generators (VGs), more commonly known from the aviation industry, have been demonstrated to delay stall and improve the stall region characteristics. This restoration of properties has been associated with reduced fatigue loading following the logic that the rotor blade will undergo stall less severely and less often. This hypothesis was tested in this study using a newly developed fatigue tool ALBdeS (named after W. Albert the first author to write a paper considering fatigue) to post-process aeroelastic simulations conducted in FAST (from NREL/NWTC) [2]. The post processing tool is an extension of the PMV custom section rotorblade analysis tool of SMART BLADE GmbH. The ALBdeS tool calculates the cumulative damage value in each individual layer of the blade section laminates and determines whether or not failure will occur over the course of 20 years, following the GL Guidelines [3]. Sensitivity studies showed that by de-constructing the main oscillation into 30 analysis points, the accumulated damage converges to a stable result, thus increasing confidence in the stability of the method. The FAST simulations were conducted with modified versions of the NREL 5MW reference turbine. The inboard lift and drag polars of the 5MW were modified in order to simulate the effect of adding VGs to the design. The polar modifications were made in the absence of 3D stall delay effects although literature does indicate the effects are somewhat additive. However, the resulting simulations did demonstrate that VGs did in fact change the fatigue characteristics of the rotor blades but by an inconsiderate amount.