Saarblitz, Wind farm in HDR, flickr.com, creative commons by-nc-sa 2.0

AVATAR 2D High Reynolds

Managed by

Scope

Within the AVATAR project 2D airfoil measurements have been taken in the wind tunnel at conditions similar to the conditions of 10MW+ turbines. The pressurized DNW HDG wind tunnel measurements have been carried out on a DU00-W-212 airfoil where the pressurized environment enabled Reynolds numbers up to 15M (note that the high tunnel pressures allow such high Reynolds numbers to be taken at a relatively low Mach (M) number of say 0.1 in correspondence with wind turbine conditions). Such high Reynolds/low M measurements on a wind turbine airfoil are very unique.

Data Accessibility

Creative Commons Attribution-NonCommercial-NoDerivatives

Input data

Model and Geometry
Model used during the tests has 0.15m chord and 0.6m span. It was located at the center of the test section which is 0.6mx0.6m with a length of 1.0m. The model has 90 pressure taps along the mid span. Pressure readings are integrated in order to obtain lift and pitching moment coefficients. A wake rake with 118 total and 8 static pressure probes is installed 3.5 chords downstream of the trailing edge of the model. The complete wake rake can be side-wise traversed along the starboard half of the test section. The drag coefficient is calculated from the flow loss of momentum by integrating the total and static pressures in the airfoil wake.

Figure 1 DNW-HDG wind tunnel test section used in the measurement campaign and the position of the airfoil inside.

The coordinates of the airfoil to be used in the simulations are given in the appendix and delivered with separate .txt files. The first file holds the original airfoil coordinates which are used as input to design the wind tunnel model. Although the difference between the given coordinates and the manufactured airfoil is within the specified tolerances, it is decided to also deliver the actual measured model coordinates for this comparison. Therefore, the second attachment holds the measured airfoil coordinates at the center of the span. This location is in the middle of the pressure tap position. The third attachment has the measured coordinates at the spanwise location where the wake rake was mostly used to measure the drag (100 mm. to the starboard side from the mid-section).
We suggest using the original airfoil coordinates (first one) for the blind test comparisons. However, in case you would like to compare, you can use the measured coordinates. The deviation of the measured section and the nominal coordinates are shown in Figure 2 and Figure 3. The dimensions are in mm units.
500mm300mm425mmWake rake150mm


Figure 2. Deviations between the manufactured & assembled geometry and the CAD model. Measured section is located at the center of the span.


Figure 3. Deviations between the manufactured & assembled geometry and the CAD model. Measured section is located 100 mm. to the starboard from center of the span.

Validation data

Ceyhan Ozlem, Pires Oscar, & Munduate Xabier. (2017). AVATAR HIGH REYNOLDS NUMBER TESTS ON AIRFOIL DU00-W-212 [Data set]. Zenodo. http://doi.org/10.5281/zenodo.439827

Model runs

For all tests, three different turbulent conditions should be simulated. The results, therefore, should be provided for these three turbulence levels. Turbulent conditions are given in terms of turbulence intensity (Ti) in percentage. Different turbulent levels are obtained from the (turbulence) measurements performed in DNW-HDG wind tunnel in different periods. These measurements are performed at the model location or very close to it by hot wire.

Where
Pt = total pressure of the tunnel [bars]
Vtunnel= tunnel speed [m/s]
ρ= density [kg/m3]
T= temperature [⁰K]
Ti= turbulence intensity [%]
Simulations should be performed for every 2 degrees of angle of attack from -16 to 20 degrees. It is suggested to pay attention in capturing the corners of the laminar drag bucket (if possible or applicable) which might mean refinement in angle of attack values used for simulations only to capture laminar drag bucket corners. Since the specific angle of attack values for this purpose might differ from code to code, no further specification is possible at this point.

Output data

The output data necessary for the comparisons are the following:
Cl, Cd, Cm(0.25), XTRpres, XTRsuc for every two degrees of angle of attack values between -16 to 20 degrees.
Cp and Cf vs. X/c and Y/c (non-dimensionalized!) for all angle of attack values.
Ti% at 0.5c in front of the airfoil leading edge and 2c behind the airfoil (outside the wake, when possible) trailing edge for each test and each specified turbulence level (assuming that these values won’t change significantly with angle of attack). Turbulence values should be delivered only for the models that depend on the turbulence intensity.
To be consistent with the earlier delivered data formats, output data should be delivered with the following formats:

Three set of files are required per condition.
1- “Polar.dat” : Polar files. The data inside should be stored as:
# AOA CL CD CM XTRpres XTRsuc
-15 … … … … ...
-14 … … … … …

2- “APxx.cpcf” or “AMxx.cpcf” : to store pressure coefficient and skin friction values for every angle of attach value. If positive angle of attack, choose AP, if negative then choose AM. Next two digits should come from the corresponding angle of attack. For instance. If -8 degree angle of attack Cp and cf data will be stored, then the name should be AM08.cpcf. The data inside should look like:
# X/C Y/C CP CF
X1/c y1/c Cp1 Cf1
X2/c y2/c Cp2 Cf2

3- “Turb.dat” : store the turbulence intensity information. It is only one line data and it should look like
# Ti[%]LE(0.5c) Ti[%]TE(2.16c)
Ti1 Ti2

These files above should be stored in the directories in the following order:
Organisation Name_SoftwareName/Grid/Turb or Tran/TestX/TurbLevX/filename
Grid name should be either CGF(common grid file) or OG(own grid)
For example: a polar file from RFOIL fully turbulent simulation performed by ECN for Test 4 for the turbulence level 2 should be stored as
ECN_RFOIL/OG/Turb/Test4/TurbLev2/polar.dat
In this case grid name is chosen as OG since RFOIL doesn’t use grid.

Remarks

Further information about the experiment. Literature survey

Public

Documents

Loading widget...
No files to show
{{node.name}}
({{node.children.length}})
{{node.date}}
{{node.modified}}
{{node.filesize}}
{{node.filename}}