## How does it work

The AVATAR (AdVanced Aerodynamic Tools of lArge Rotors) project is organized in different Work Packages (WP’s). Apart from the dissemination and coordination Work Packages there is Work Package 2 which deals with the advanced aerodynamic modeling of all aspects which are expected to play a role in the design of large 10MW+ wind turbine blades. This Work Package excludes the modeling of flow devices which is included in the separate Work Package 3. The modeling of aero-elastic effects on large and flexible rotor blades also needs a separate which is Work Package 4. Moreover Work Package 1 is added which integrates and evaluates the results and which provides the reference turbines on which the modeling is tested.

As such there are 4 technical Work Packages. A detailed description of every work package can be found in the links below. The work structure is shown graphically in Figure 1.

• Integration and Evaluation 10 MW Rotor (WP 1),

• Advanced Aerodynamic Modeling (WP 2),

• Models for Flow Devices and Flow Control (WP 3),

• Aeroelastic Analysis of Large and Flexible Blades (WP4 ).

WP1 is presented in vertical direction and WP’s 2 to 4 are presented horizontally. WP1 is led by CRES, WP2 by DTU, WP3 by CENER and WP4 by NTUA.

^{Figure 1: AVATAR's four technical WP's}

The focus of WP1 lies on the (integrated) design and evaluation of a 10MW reference rotor. Thereto a design of a reference rotor is delivered to the horizontal work packages for further analysis. The design of the reference rotor is closely related to activities carried in the adjacent INNWIND.EU project where a 10 MW reference rotor is designed too. However AVATAR adds another reference rotor which is intended to be more challenging, i.e. more extreme in terms of aerodynamic modeling.

The design of the AVATAR reference rotor takes place in the first 6 months of the project after which it is, together with the INNWIND.EU reference rotor, delivered to the WP’s 2 to 4. Thereafter in the period from month 6 to month 36 of the project, the emphasis of the activities lies at WP’s 2 to 4 which are run in parallel.

In WP2 the aero-tools are improved and calibrated for all aspects, which play a role at the design of large wind turbines. In WP3 the models for flow devices and flow control are developed and improved in aerodynamic terms, basically on a sectional level. Then WP4 considers the aeroelastic aspects of large scale rotors where it should be noted that aerodynamics and aeroelasticity are inextricably connected.

After the delivery of the WP’s 2 to 4 results (i.e. the improved models) in month 36, the emphasis of activities moves again to WP1 where the behavior of the reference rotor is re-evaluated based on the newly developed models. This is followed by a redesign of the turbine using the advanced control options. Finally WP1 will also develop aerodynamic/aero-elastic design guidelines on the aerodynamic related actions which are needed to attain a further upscaling towards 20 MW. These guidelines include a list of models which need further adjustments when applied to 20 MW turbines and a definition of a large scale experiment (See below).

The improvement and validation of aerodynamic models for 10MW+ turbines obviously also requires suitable experimental data but this is complicated by the fact that 10 MW+ turbines are not on the market yet by which measurements on such turbines are impossible to accomplish. Nevertheless AVATAR will apply a long list of experimental data which either focus on specific submodels or they are taken at a smaller scale by which they anyhow facilitate the improvement and validation of models.

The list of foreseen experiments consists of:

• Pressurized DNW HDG wind tunnel

Airfoil measurements at Reynolds numbers up to 15 Million and low Mach number (< 0.2)

• LM: Wind tunnel airfoil measurements at dynamic conditions

• Forwind: Wind tunnel airfoil measurements at specified turbulence inflow

• DTU : Danaero: Aerodynamic field experiments on a 2.3 MW turbine and supporting 2D wind tunnel measurements

• TUDelft: 2D (possibly rotating) wind tunnel experiments on airfoils with vortex generators in LST and Open jet facility.

• NTUA: Wind tunnel experiments on airfoils with/without vortex generators

Note that further experiments may be added (in-kind) at a later stage.

The provision of experimental data is closely linked to similar activities going on in the subgroup aerodynamics of EERA which clearly stated that the time is ripe for a new joint field aerodynamic measurement program on a scale which is at large as possible (at least 5MW scale, increasing to 10MW+ in the near future). The subgroup aerodynamics also stated that the data of this experiment should be made publicly available to the entire European research society. Thereto an extensive set of aerodynamic data should be collected on a large scale state-of the art turbine using the most advanced measurement techniques. A more detailed definition of the experiment will be carried out in AVATAR and forms part of the above mentioned guidelines on the required aerodynamic actions needed to make 20 MW turbines possible.

## Approach: aerodynamic model improvement and calibration

Roughly speaking, three types of models can be distinguished for evaluation of wind turbine aerodynamics, ranging from low complexity/computational efficient models to high complexity/computational demanding models, with intermediate models in between, see also Figure 2. Low complexity models are based on the so-called Blade Element Momentum theory with engineering add-ons, see e.g. [Schepers 2012]. High complexity (high fidelity) models basically consists of Computational Fluid Dynamics tools see eg [Sorensen, 2011] or possibly models based on viscous-inviscid interaction schemes. Figure 2 also indicates intermediate tools mainly consisting of Free Vortex (or possibly prescribed) wake methods. Now it should be known, see [Schepers 2012], that the role of calculation; time for wind energy applications is much more crucial than it is for most other areas of technology and that the number of time steps for design calculations is more than 5 million. This makes the use of computational efficient methods (i.e. BEM with engineering methods) inevitable for routinely design calculations but it is inparticular this class of models which has a large number of calibration constants, the validity of which is unknown for 10 MW+ turbines. The improvement of low and intermediate complexity models in the present project is then largely achieved by calibrating them with results from high fidelity models. This however does not include an enhancement of these high fidelity modles too, e.g. for the modeling of transition, flow devices etc. Obviously models are also improved and validated using the experiments described above. A description of all tools available in the project, is given here.

^{Figure 2: Aerodynamic models in terms of increasing complexity}