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

2D Unsteady Aerodynamics Open Data Set

Managed by

Description: 

This test case compiles an experimental 2D airfoil unsteady aerodynamic data set. The data have been obtained from tests performed at four different facilities by four different universities and research centers:

  • National Renewable Energy Center (CENER): Dynamic 2D airfoil pithing and flapping tests at the Technical University of Denmark (DTU) ‘Red’ wind tunnel
  • University of Glasgow: Dynamic 2D airfoil pitching tests at their own facility, the 5’ × 7’ working section low-speed, closed-return wind tunnel
  • National Renewable Energy Laboratory (NREL): Dynamic 2D airfoil pithing tests at The Ohio State University Aeronautical and Astronautical Research Laboratory 3x5 subsonic wind tunnel
  • University of Oldenburg: 2D static airfoil tests at different unsteady generated turbulent flow performed at their own return type acoustic wind tunnel.


From these tests, a preliminary selection of cases has been made for this data set considering application to the wind turbine aerodynamic investigation. The data selected are:

  • CENER experiments: NACA643-418 aifoil tested at sinusoidal type motion of the pitch, the flap and combined pitch and flap.
  • University of Glasgow experiments: NACA0015 and NACA0030 airfoils tested at sinusoidal type motion of the pitch.
  • NREL experiments: LS(1)0417MOD, NACA4415 and S809 airfoils tested at sinusoidal type motion of the pitch.
  • University of Oldenburg experiments: DU00W212 airfoil with laminar flow, open grid condition and one sinusoidal dynamic grid condition.

Data accessibility:

All data are availabale under Creative Commons Attribution Share-Alike 4.0 license.

They can be downloaded here: https://doi.org/10.5281/zenodo.1135424

Additionally the data and the accompanying scripts will soon be available in the Jupyter based sandbox service.

CENER

These experiments took place during two campaigns performed at May and November 2015 within the EU funded project ‘Windtrust’ under FP7 with Grant agreement number: 322449.

Site description:

The wind tunnel tests have been performed at the ‘red’ wind tunnel facility of DTU Wind Energy Department at Lyngby Campus.
This is an open loop suction wind tunnel with a contraction ratio of 12.5:1. The test section has a cross section of 0.5 m x 0.5 m and length of 1.3 m, and a maximum wind speed of 60 m/s. The main dimensions of the wind tunnel are shown in the figure below.
DTU 'RED' open loop wind tunnel scheme

The quality of the flow is increased through an inlet filter and four screens placed before the contraction. The turbulence intensity is about TI = 0.08-0.07.
The inflow is from the upper left side and the velocity of the flow is measured by a pitot tube just after the inlet (head of the probe at 0.07 meters from the test section inlet) and 0.10 m from the wall, as shown in figure below.
The model is placed vertically in the test section spanning from bottom and top walls. The model hinge point is located 0.61 m. from the test section inlet.

Test section plan view

A wind tunnel model wing representing the NACA643-418 airfoil has been used for this test. It has a chord of c=0.25 m. and a span of l=0.5 m. Surface of the model is made of carbon-fiber-reinforced plastic molded to the subscribed shape and  supported by a mechanical frame made out of six aluminum ribs.
The model has been designed to have a flexible trailing edge flap, by using thinner fiber sections in the surface at the bending positions for a continuous airfoil surface. The aim is to have a 15% flap for a wing chord section of 250 mm., i.e., 37.5 mm. The flap actuation was provided through a linear motor that moves a rudder horn on the pressure side and in the half span of the model (see sketch below).

Flap linear motor actuator system
When the flap is moved to different angular positions, the carbon fiber shell of the flap on the pressure side slides into the undeflected wing shell. The pressure side of the flap is deflected linearly. However, the flap shell on the suction side is bended, as shown below.

Flap shell on the suction side

Data Provider: 

National Renewable Energy Center of Spain (CENER)

Instrumentation: 

Model is instrumented with 63 pressure taps in line at span section 41% and connected to two 32 channel pressure transducers. The other remaining channel was used for the test section pitot tube. The static pressure from the pitot tube is used as reference pressure.
Lift , Pitching Moment and Pressure Drag Coefficients were calculated by integration of the pressure distribution over the airfoil.

Measurement Campaign:

All the tests have been performed at a Reynolds number of 0.5 million (wind tunnel speed: 30 m/s). The four types of tests included in this data base are:

  • Baseline static polars:
    • Measurements at steady conditions are taken (AoA and Flap are not moving).
    • Different Flap deflections are taken (-10⁰, -5⁰, 0⁰, 5⁰ & 10⁰). For each Flap angle, different angles of attack are measured in the same run. The measuring time at one fixed angle of attack is 10 seconds.
  • AoA sinusoidal movements (pitch motion):
    • The angle of attack is moved with a sinusoidal movement. Flap is fixed.
  • Flap sinusoidal movements:
    • The flap is moved with a sinusoidal movement. AoA is fixed.
  • AoA & Flap combined sinusoidal movement
    • Angle of attack and flap are moved with a sinusoidal motion. Both have the same frequency but can have different phase delays between them.

Data files: 

Each file contains the data organized in columns with this information:
    Time (in seconds): corresponding time of the measured values
    AoA (in degrees): Airfoil angle of attack
    Flap (in degrees): Flap deflection
    Cl: Lift coefficient
    Cdp: Pressure drag coefficient
    Cm: Pitching moment coefficient


File names: All the data files begin with N64418 followed by an underscore and a capital letter (S, P, F or C). This letter indicates the type of test as described above:
    S: Baseline static polars
    P: AoA sinusoidal movements (pitch motion)
    F: Flap sinusoidal movement
    C: AoA & Flap combined sinusoidal movement
Depending on each type of test the file name has these other additional fields separated by underscores:
    Case S: Only one field containing the Flap deflection:
        Example: N64418_S_F05 (polar with Flap 5o deflection)
    Case P: Four fields following type of test:
N64418_P_Axx_xx_kx-xxx_Fxx
    Field 1: Axx: mean angle of attack value
    Field 2: xx: angle of attack amplitude
    Field 3: kx-xxx: reduced frequency
    Field 4: Flap position
Example: N64418_P_A05_10_k0-100_F00 (Pitching motion with AoA mean value of 5o, amplitude of 10o and reduced frequency of 0.1. Flap at 0o.)
    Case F: Four fields following type of test:
N64418_F_Axx_Fxx_xx_kx-xxx
    Field 1: Axx: AoA position
    Field 2: Fxx: mean flap value
    Field 3: xx: flap amplitude
    Field 4: kx-xxx: reduced frequency
Example: N64418_F_A10_F00_05_k0-050 (Flap motion at AoA=10o with a Flap mean value of 0o and amplitude of 5o and a reduced frequency of 0.05.
    Case C: Seven fields following type of test:
        N64418_C_Axx_xx_kx-xxx_Fxx_xx_kx-xxx_Phxxx
    Field 1: Axx: mean AoA value
    Field 2: xx: AoA amplitude
    Field 3: kx-xxx:  reduced frequency
    Field 4: Fxx: mean flap value
    Field 5: xx: flap amplitude
    Field 6: kx-xxx: flap reduced frequency
            Field 7: Phxxx: phase shift between both motions in degrees
N64418_C_A10_10_k0-050_F00_10_k0-050_Ph045 (Pitching motion with AoA mean value of 10o and amplitude of 10o combined with flap motion around 0 with amplitude of 10o. Both motions with reduced frequency of 0.05 and a phase shift of 45o

University of Glasgow

These 2-D dynamic stall tests were conducted by the Department of Aerospace Engineering, University of Glasgow, Glasgow, UK, in the 1980s and 1990s.

Site description: 

All the tests were conducted in the University of Glasgow “The Hanley Page Tunnel” with a test section 7ft x 5ft (2.1m x 1.5m). Maximum speed 120mph (55m/s). Both NACA 0015 and NACA 0030 2D dynamic stall models were built with a chord length of c = 0.55m, and they spanned the shorter dimension of the tunnel (1.5 m.). Their construction was of a fiber glass skin bonded to a steel or aluminum spar. A hydraulically actuated crank and pitch link mechanism was used to pitch each airfoil model about its quarter chord, with pitch angle feedback provided by a rotational displacement transducer.
The Hanley Page Tunnel

Data Provider: 

University of Glasgow

Instrumentation: 

Each model was instrumented with high performance Kulite or Entran surface mounted pressure transducers at the model centre-span; models for 2-D dynamic stall contained 30 transducers arranged in a chordwise manner on the airfoil upper and lower surfaces. Pressure transducer signals were anti-aliased filtered and amplified before simultaneous sampling across all the transducers on the model.
Unsteady and dynamic stall pressure measurements were acquired and post processed resulting in force coefficients measured exclusively from the pressure taps, with no wake rake measurements of drag. Therefore the data available are Cl, Cdp (only the pressure drag, not the total drag) and Cm for the range of angle of attack excursions. During these tests data acquisition used a DEC MINC limited to a sampling rate of around 550 samples per second per channel for 256 samples. As part of the standard testing protocol sinusoidal oscillatory tests were sampled over ten pitch cycle oscillations after five previous cycles had been completed. Accurate sampling triggering ensured that an averaging calculation across the sampling cycles could be performed without loss of the details of the time varying aerodynamic transients.

Measurement Campaign:

For dynamic stall tests the wind tunnel was run usually at a nominal free stream speed of U∞ = 42m/s giving a chord Reynolds number of 1.5 million for a 0.55m chord, and a Mach number below 0.16. For all the testing conditions the turbulence intensity level of the flow was always below 0.5%.
The models were tested over a range of test and motion types. Here only are included the sinusoidal tests that were performed at a range of mean angles and amplitudes over a range of reduced frequencies k = ωc/U∞ ,where  ω= 2Πf and f is the oscillation frequency in Hz.
For static tests the model was held at fixed angle of attack and the unsteady pressure data sampled.
The tests included in this data base are:

For NACA 0015 model:

  • Static test cases: AoA polar from -25 to +25 deg
  • Sinusoidal test cases:
    • Mean AoA tests from 0 to 20 deg.
    • Amplitude from 4 to 32 deg.
    • Reduced frequency k, from 0.01 to 0.18.


For NACA 0030 model:

  • Static test cases: AoA polar from -24 to +25 deg
  • Sinusoidal test cases:
    • Mean Aoa tests from 0 to 20 deg.
    • Amplitude from 4 to 32.8 deg.
    • Reduced frequency k, from 0.01 to 0.40

Data files: 

Each file contains the data organized in columns with this information:
    Time (in seconds): corresponding time of the measured values
    AoA (in degrees): Airfoil angle of attack
    Flap (in degrees): not applicable for these files. Contains ‘NaN’
    Cl: Lift coefficient
    Cdp: Pressure drag coefficient
    Cm: Pitching moment coefficient


File names: data files name format begin with an eight digit number (abcdefgh_coeffs.dat)
where: ‘ab’ is the model number (05 for NACA0015 and 09 for NACA0030); ‘c’ is the test type (in this data base all of them correspond to 0); ‘d’ is the motion type (0 – static , 1 – sinusoidal oscillation, 4 – unsteady static); ‘efg’ is the test number; ‘h’ is the attempt number at this test.

NREL

The National Renewable Energy Laboratory (NREL), funded by the US Department of Energy, awarded a contract to Ohio State University (OSU) to conduct the dynamic stall wind tunnel tests on different airfoils during the 1990s.

Site description: 

The OSU ARC (Aerospace Research Center) Battelle Subsonic Wind Tunnel is an open circuit wind tunnel with a 0.91 × 1.52 m. (3 × 5 ft) test section. The length of the test section is 2.4 m. The maximum tunnel wind speed is 45 m/s, produced by a 2.44 m. diameter, 6-bladed fan located at the exit. The fan is powered by a 93.2 kW, 3-phase AC motor. Test section turbulence level measures below 0.1%.
2D models of the S809, the LS(1)0417MOD and the NACA 4415 airfoils among others, with constant chord of 457 mm. were manufactured out of a sandwiched composite skin over ribs. The main load bearing member was a 38-mm diameter steel tube which passed through the model quarter chord station. Ribs and end plates were used to transfer loads from the composite skin to the steel tube. The final surface was hand worked using templates to attain given coordinates within a required tolerance of ±0.25 mm.
To minimize pressure response time, which is important for the unsteady testing, the surface pressure tap lead-out lines had to be as short as possible. Consequently, a compartment was built into the model so pressure scanning modules could be installed inside the model. This compartment was accessed through a panel door fitted flush with the model contour on the lower (pressure) surface.

Data Provider: 

NREL

Instrumentation: 

Data were acquired and processed from 60 surface pressure taps, four individual tunnel pressure transducers, an angle of attack potentiometer, a wake probe position potentiometer, and a tunnel thermocouple. The data acquisition system included an IBM PC-compatible, 80486-based computer connected to a Pressure Systems Incorporated (PSI) data scanning system. The PSI system included a 780B Data Acquisition and Control Unit (DACU), 780B Pressure Calibration Unit (PCU), 81-IFC scanning module interface, two 2.5-psid pressure scanning modules (ESPs), one 20-inch water column range pressure scanning module, and a 30-channel Remotely Addressed Millivolt Module (RAMM-30).
Battelle Subsonic Wind Tunnel

Measurement Campaign: 

Tests were performed at Reynolds numbers of 0.75, 1, 1.25 and 1.5 million.
Data were obtained at surface clean condition of the airfoils and also applying Leading Edge Grit Roughness (LEGR). This data base only has data from the clean condition cases.
For steady state cases, the model was set to angle of attack and the tunnel conditions were adjusted. At operator request, pressure measurements from the airfoil surface taps and all other channels of information were acquired and stored by the DACU and subsequently passed to the controlling computer for final processing. The angles of attack were always set in the same progression, from negative to positive values.
For model oscillating cases, the tunnel conditions were set while the model was stationary at the desired mean angle of attack. The "shaker" was started, after approximately 10 seconds the model surface pressure and tunnel condition data were acquired. Generally, 120 data scans were acquired over three model oscillation cycles. Since surface pressures were scanned sequentially, the data rate was set so the model rotated through less than 0.50° during any data burst.
The pitch oscillations data were acquired at frequencies of 0.6, 1.2, and 1.8 Hz. Two sine wave forcing functions were used, ±5.5° and ±10°, at mean angles of attack of 8°, 14°, and 20°.
For model oscillating cases, the model surface pressure measurements were acquired and post processed resulting in force coefficients, with no wake rake measurements of drag. Therefore the data available are Cl, Cdp (only the pressure drag, not the total drag) and Cm for the range of angle of attack excursions. For the steady state cases, wake pressure data were acquired from a traversed pitot-static probe. These pressure measurements were used to calculate drag coefficient. But these drag coefficients data are not compiled in this data set, only the pressure drag data are presented.

File names: the steady state and the dynamic polars have different file names formats.
The steady state polars have these fields:
Name of the airfoil
    Character ‘C’ (standing for clean case)
    Reynolds number in tens of thousands
    Underscore followed by ‘Coef’
        “N4415C75_Coef.txt”:  Static state polar of NACA4415 airfoil (clean case) at Reynolds Number of 0.75 million.

The dynamic polar names have these fields:
    Character ‘C’ (clean case)
    AoA amplitude: 5 = 5.5 deg, 10 = 10 deg
    Pitch oscillation frequency: l = 0.6 hz, m = 1.2 hz, h = 1.8 hz
    Reynolds number in tens of thousands
    Airfoil name (or abbreviation of it)
        “C10m125_14_n4415.txt” Clean case, 10 deg amplitude, 1.2 hz, Re=1.25 million, NACA4415 airfoil

University of Oldenburg

Experimental airfoil characterizations with different inflow conditions were performed on a DU00-W-212 model in the return-type acoustic wind tunnel at the University of Oldenburg during 2015 and 2016 for the EU funded project ‘AVATAR’ with grant agreement nº: 608396.

Site description:

The model representing the DU00-W-212 airfoil has a chord of 300 mm and is vertically mounted inside the closed test section with cross-section of 1.0 m x 0.805 m (w x h) and 2.6 m length. The model spans the complete height of the test section (805 mm.) and is fixed to two rotating plates at its ends, which are fitted flush with the test section floor and ceiling. Two axes support the model at quarter chord and each axis is connected to a three-component load cell. The top axis is also equipped with a torque sensor and a stepper motor to control the geometric angle of attack, which is monitored by an angle encoder at the bottom mount.
The reference wind speed is measured with a Setra C239 pressure gauge connected to Pitot-static tubes in the wind tunnel contraction. A combined sensor for the meteorological data (ambient pressure p, temperature T, rel. humidity rH) is situated downstream the airfoil at the end of the closed test section.
An active grid was inserted between nozzle and test section in order to generate the reproducible, customized inflow patterns for the turbulent measurements. The grid features 16 individually movable shafts with attached flaps, of which only the 9 vertical shafts were moved during this experiment. All horizontal shafts remained in open position (least blockage) to render a quasi-two-dimensional turbulence pattern with customized inflow angle variations. A velocity-specific transfer function was used to relate grid movement and resulting flow angle in order to implement the generation of sinusoidal inflow angle fluctuations in the flow.

Return-type acoustic wind tunnel at the University of Oldenburg

Data Provider: 

AVATAR

Instrumentation: 

Model is instrumented with 48 pressure taps distributed in a span-wise staggered alignment to avoid wake interference. One tap is located at the leading edge, 25 taps along the upper surface, 21 taps along the lower surface and one tap is located at the trailing edge. The pressure taps with 0.3 mm diameter are connected to a system of three synchronized multi-channel scanners, which record the pressure at 100 Hz sampling frequency. The data acquisition with all sensors (pressure scanners, load cells, torque sensor, met sensors) are synchronized by means of a LabVIEW software and measurements are started upon a common trigger.

Measurement Campaign: 

The experiments were performed at two different Reynolds numbers of 0.5 and 0.9 million.
Airfoil model was tested at surface clean condition and also tripped at 1.5% of the chord in the suction side and 10% of the chord in the pressure side.
Laminar inflow conditions were obtained by mounting the closed test section of the wind tunnel directly to the outlet nozzle, in order to obtain classical airfoil polars as a baseline.
In this data set, two different customized turbulent flow cases are included: the open-grid case (all shafts of the grid remain in open position) and sinusoidal case (grid vertical axes have a sinusoidal motion of frequency 5Hz and amplitude 30º)

Data files: 

Two types of files are presented from these tests.
Pressure files:
These files contain the data of the aerodynamic coefficients (lift and pressure drag) obtained from the integration of the pressure distribution measured over the airfoil model.
Each file contains the data organized in columns with this information:
        Time (in seconds): not applicable for these files. Contains ‘NaN’
        AoA (in degrees): Airfoil angle of attack
        Flap (in degrees): not applicable for these files. Contains ‘NaN’
        Cl: Lift coefficient
        Cdp: Pressure drag coefficient
        Cm: not applicable for these files. Contains ‘NaN’
Force files:
These files contain the data of the aerodynamic coefficients (lift and drag) obtained from measurement of the force sensors.
Each file contains the data organized in columns with the following information:
        Time (in seconds): not applicable for these files. Contains ‘NaN’
        AoA (in degrees): Airfoil angle of attack
        Flap (in degrees): not applicable for these files. Contains ‘NaN’
        Cl: Lift coefficient
        Cd: Drag coefficient
        Cm: not applicable for these files. Contains ‘NaN’

File names: The file names are composed of 4 fields separated by underscores and end with a score followed by ‘Pressure’ or ‘Balance’ indicating if they are Pressure files or Force files respectively. The 4 fields show this information:
    Field 1: number of original polar
    Field 2: type of flow
        laminar: no grid installed
        opngrid: grid in open static position
        sinusoi: grid with vertical axis moving in sinusoidal mode
    Field 3: ‘cl’ for clean surface polars. ‘tr’ for tripped polars
    Field 4: Reynolds number (‘500k’ for 0.5 million and ‘900k’ for 0.9 million)

Public