S805 Wind Tunnel Testing

AERSP 305WAerospace Technology Laboratory

Laboratory Section 4

Laboratory Experiment Number 3Wind Tunnel Testing of a S805 Airfoil

March 2, 2012Performed in Room 8 Hammond Building

Connor Hoover

Lab Partner’s Names:Ethan Corle

Kaitlynn HetrickStephen PrichardAnthony Parente

Lab TA: Kylie Flickinger

Course Instructor: Richard Auhl

Page 1 of 15

AbstractThe objective of this lab was to determine the coefficients of lift and drag of a S805 airfoil at

various angles of attack. This was achieved in three different sections. First was the

collection of pressure tap data on the surface of the airfoil using a manometer bank, next a

wake velocity profile using a hot-wire anemometer, and finally a general observation of stall

effects on an airfoil. The calculating of the coefficients was done using equations introduced

in the next section. The pressure data will yield the coefficient of lift, while the wake velocity

profile produced the coefficient of drag. These values once calculated were compared to

NREL published data for the S805 airfoil. The experimental data was found to be accurate

when compared and that the goals of calculating the values of the coefficients were achieved.

The one main source of error was that the airfoil used in the experiment was half the span of

the airfoil of the published data.

Page 2 of 15

RsTPTT

PP r

rr /

(1)

(2)

UcRe

IntroductionThe objective of this experiment was to determine the coefficients of lift and drag of a two dimensional

airfoil at a range of angles of attack. The airfoil tested was the S805 airfoil; Figure 1 shows the

dimensions of this airfoil in inches.

Figure 1. Test S805 Airfoil

To complete the objective stated above, certain equations are needed to determine key values. The

first is the ideal gas law, which will give the density of the air in the wind tunnel with the simple

measurement of temperature and pressure. Equation 1 below shows the ideal gas law rearranged to

solve for density:

Another important value needed to understand the data collected is Reynolds number. This non-

dimensional value will help to control the experiment. Equation 2 below will allow for the calculation of

the velocity needed to keep the wind tunnel at the Reynolds number specified in the procedure.

The next set of equations deal with the calculation of the coefficient of lift for the airfoil. This is done

via the integration of the pressure distribution on the airfoil. Using a manometer bank, at a given angle

θm and having a known specific gravity, equations 3a and 3b can be used to calculate the pressure

coefficients of the upper and lower surfaces of the airfoil. It should be noted that h∞, the reference

height will be measured by a specific port in the manometer.

C pu=Pu−P∞

q∞=S .G. sin θm¿¿ (3a) C pl

=Pl−P∞

q∞=S .G . sin θm¿¿ (3b)

Page 3 of 15

By averaging the coefficients of pressure determined above across the chord length the normal force

coefficients can be written as equation 4a and 4b below.

cniUPPER=¿ (4a)

cni LOWER=¿ (4b)

Remaining in coefficient form, the total normal force acting on the airfoil can be found by subtracting

the sum of each of the individual normal force coefficients found in the equations above. Equation 5

below shows this relationship.

cn= ∑i=1

(n) panels

cni LOWER− ∑

i=1

(n ) panels

cniUPPER (5)

Finally, the coefficient of lift for the airfoil can be determined by equation 6 below, where α is the angle

of attack of the airfoil.

c l≅ cn cosα (6)

The coefficient of drag will be calculated via the integration of the wake profile. This profile will be

measured by a hot-wire anemometer mounted downwind of the airfoil. This data will give us the profile

drag which can be manipulated to find the coefficient of drag.

Profile drag has two components, pressure drag and skin friction drag. By looking at the momentum

flux ahead and aft of the airfoil, profile drag can be determined as the difference in the x-component of

this flux. When considering only the linear momentum, the air suffers a loss of momentum flux equal

to the drag of the airfoil. In this case, drag can be described as equation 7 below.

F=D= ddt

(mV )=m∆V (7)

Where mass flow rate can be written as,

Page 4 of 15

m=∬ ρVdA (8)

Substituting this expression into equation(7), drag force is put in terms of the free stream velocity of

the wake (V 0), the local velocity of the wake (V), and cross-sectional area (dA).

D=∬ ρV (V 0−V )dA (9)

Non-dimensionalizing the above equation gives the section drag coefficient of the airfoil. Again c is

the chord length of the airfoil while b is equal to 1 for unit span.

Cd=2∫(√ qq0

− qq0

) dyc = D12ρV o

2bc (10)

The use of the hot-wire anemometer can yield a velocity profile wake as shown in Figure 2 below. If

the traverse moves an even number of steps to collect the data, Simpson’s Rule for integration can be

applied to find the drag coefficient.

Figure 2. Wake Velocity Profile

The final part of this experiment is to observe the airfoil as it moves to stall; the angle of attack will be

increased until a large pressure drop is observed across the surface of the airfoil. Also, the airfoil will

then be brought back from this stall angle until normal flow begins again. This is expected to be at a

lower angle than the original stall angle due to flow circulation in the wake.

Page 5 of 15

Experimental ProcedureTo begin the temperature and pressure of the laboratory were measured. Using the calibration given

for Transducer B the velocity needed to achieve a Reynolds number of 800,000 was calculated. This

was found to be around 81 ft/s. The wind tunnel was dialed up until Transducer B read a voltage of

2.37, which corresponds to this velocity. Afterwards, the airfoil was rotated using LabView to the first

angle of attack, -5 degrees. As shown below in Figure 3, the S850 airfoil was hooked up to a

manometer bank to record the pressures felt on the surface of the airfoil.

Figure 3. Airfoil Set Up in Wind Tunnel

The manometer bank’s 36 ports all recorded pressures from pressure tap connections on the airfoil.

Two ports, 18 and 19, were used to record the atmospheric and T.S. Static pressures. The angle of

the bank was also recorded to interpret the data recorded later. These pressure readings were later

used to calculate the coefficient of lift for the airfoil. After finishing these readings for -5 degrees the

airfoil was rotated to an angle of attack of 15 degrees where another set of manometer data was

taken. Figure 4 below shows a sample reading of the manometer bank.

Page 6 of 15

Figure 4. Sample Manometer Bank Data Reading

The second part of the lab recorded the wake profile behind the airfoil. This was done using a hot-wire

anemometer as stated before. The hot-wire recorded the wakes for both angles at specified points.

These two profiles take were later used to determine the coefficient of drag of the S850 airfoil. Also,

the hot-wire took a time trace at each location with a sample rate of 2000 samples per second. To

correct the voltages taken by the hot-wire during the experiment to fit with the temperature of the lab

the following equation was used:

Ecorrected=Emeas .(T wire−Tcalib .

T wire−Tmeas .)

12 (11)

This allows for the hot-wire voltages to be correctly converted to velocity without the problem of the

temperature the wire was calibrated at being different than the laboratory environment.

Page 7 of 15

For the final portion of the experiment the wind tunnel was taken down to 50% power. The airfoil angle

of attack was increased until it reached stall, then reverted back until it had recovered from stall. The

angle of attack at which stall was observed was determined by watching the manometer bank.

Page 8 of 15

Results and Discussion In lab experiment 2 it was shown that the hot-wire anemometer was able to accurately map a flow

field and be used to find the flow’s turbulence intensity. This was done again for the S805 airfoil at

angles of attack of -5 and 15 degrees. The plot below shows the velocity measured by the

anemometer versus the time. The turbulence intensities were calculated and showed that the flow

over the airfoil at 15 degrees angle of attack is slightly more turbulent than the -5 angle. This is to be

expected since flow is known to separate earlier when the angle of attack is greater. The turbulence

intensities of both angles are shown in Table 1.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 186

86.5

87

87.5

88

88.5

89

89.5

Alpha = -5Alpha = 15

Time (sec)

Velo

city

(ft/s

)

Figure 5. Velocity vs. Time at Section 4 α’s

Table 1. Turbulence Intensity At Section 4 α’s

Angle of Attack (Deg) Ti-5 0.0023115 0.00301

The wake profiles for the airfoil for each of the angles of attack are shown in figures 6 and 7. It should

be noted that the voltages of the hot-anemometer were corrected by use of equation 11. These plots

were later used to determine the initial dynamic pressures used to calculate the coefficient of drag. As

expected the angle of attack of 15 has a larger wake profile than the -5 angle of attack. Again this can

be explained by the greater separation caused by the magnitude of the angle. This will create a larger

velocity profile for the 15 degree alpha; hence it makes sense that figures 7 shows a larger wake Page 9 of 15

region than figure 6. This shows that the 15 degree angle of attack will cause more drag. As stated in

equation (9) and shown in figure (2) the area under the wake profile curve is equal to the resultant

drag. So since the wake profile is bigger for 15 degrees the drag is greater when the airfoil is in this

configuration.

60 65 70 75 80 85 900

5

10

15

20

25

30

Uncorrected Corrected

Velocity (ft/s)

Wak

e Tr

aver

se D

istan

ce

Figure 6. Wave Traverse Distance vs. Hot-Wire Velocity (α = -5 Deg.)

60 65 70 75 80 85 900

5

10

15

20

25

Uncorrected Corrected

Figure 7. Wave Traverse Distance vs. Hot-Wire Velocity (α = 15 Deg.)

The pressure distribution calculated from the manometer readings at a Reynolds number of 800,000

are shown below in figures 8 and 9. For the angle of attack of -5 degrees the S805 airfoil the upper

Page 10 of 15

surface shows a favorable pressure gradient until .2c at which point it becomes slightly adverse until

leveling out at almost zero at .8c. The lower surface on the other hand for the most part remains

adverse until becoming slightly favorable at .8c as well. This shows that this angle is favorable to

reduce drag, due to its ease of transition from laminar to turbulent flow. But, lift will not be generated

enough to combat weight at this angle of attack.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

Lab Data, UpperLab Data, Lower

x/c

Cp

Figure 8. Coefficient of Pressure vs. x/c (α = -5 Deg.)

Figure 9 below shows that at an angle of attack of 15 degrees the upper surface of the airfoil has an

adverse pressure gradient the entire chord, while the lower surface has a favorable pressure gradient

over the chord. This allows for better lift than shown in the in figure 8 for the angle of attack of -5

degrees.

Page 11 of 15

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

-6

-5

-4

-3

-2

-1

0

1

2

Lab Data, UpperLab Data, Lower

x/c

Cp

Figure 9. Coefficient of Pressure vs. x/c (α = 15 Deg.)

After all the sections were done collecting their data for the various angles of attack the next two plots

were created. Figure 10 is a plot of Cl vs. α while figure 11 is the plot of Cl vs. Cd. Both of these plots

have overlays of NREL published data for the S805 airfoil to compare the resulting calculations from

the experimental data. As shown in figure 10, the coefficient of lift curve fits fairly well with the NREL

data for Reynolds numbers of 700,000 and 1,000,000. Only at the two extremes of the data, when

angle of attack is nearing stall do we see any discrepancies worth noting. The error involved between

the three curves can be attributed to human error reading the NREL plots, equipment calibration

mishaps, and the relative size of the air foil with respect to the wind tunnel.

Page 12 of 15

-20 -15 -10 -5 0 5 10 15 20

-1

-0.5

0

0.5

1

1.5

Section 4 Points Lab Data, (Re = .8E6)NREL (Re = 1.0E6) NREL (Re = .7E6)

Alpha (degrees)

Cl

Figure 10. Coefficient of Lift vs. α for S805 Airfoil

Figure 11 as mentioned above plots Cl vs. Cd. Now, the NREL data used at a Reynolds number of

700,000 was difficult to read, and therefore only the far left data was able to be plotted. This section

of the data fit well with the experimental data, showing that using Simpson’s rule to calculate the area

under the curve of the wake velocity profile is a viable way to calculate the drag coefficient of an

airfoil. The only problem with this method that could cause error is the selection of the ends of the

wake region in the flow. This was very arbitrarily done in this case, one way in which to improve this

would be to plot all the wake velocity profiles to process the data collected.

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35

-1

-0.5

0

0.5

1

1.5

Section 4 PointsLab Data, Re = .8E6NREL, Re = .7E6

Cd

Cl

Figure 11. Coefficient of Drag vs. α for S805 Airfoil

Page 13 of 15

The final portion of the lab was the observation of the stall. As the airfoil’s angle of attack was

increased a large pressure drop on the upper surface of the airfoil was seen at an angle of 19

degrees, this pressure drop coincides with the definition of stall. Stall is when the flow detaches from

the upper surface of an airfoil. After being stalled out, the angle was decreased again until the flow

reattached, this happened at 17 degrees. The difference between the stall angle and the angle at

which its flow reattached was due to the wake vortices causing an adverse pressure gradient along

the airfoil, as well as possible momentum forcing the flow to remain separated until reaching a lower

angle of attack.

Page 14 of 15

ConclusionsThe objectives of this experiment were completed. The coefficients of lift and drag for various

angle of attack were calculated using the data collected in the wind tunnel testing of the S805

airfoil at a Reynolds number of 800,000. The coefficients of lift was calculated using the

manometer bank’s pressure readings while the coefficients of drag were calculated by

applying Simpson’s rule to the hot-wire anemometer’s wake velocity profiles. After being

calculated the values were plotted and found to be accurate when compared to the published

NREL data. The only discrepancies arose from the human error or the size of the airfoil itself,

which was cut down to fit in the wind tunnel. It was also found that the positive angle of attack

caused a favorable pressure gradient on the lower surface of the airfoil, while an adverse

gradient was seen on the upper surface. The negative angle was found to be the opposite,

causing the positive angles to have better coefficients of lift which is to be expected.

In future experiments of this nature can be improved by the use of various different airfoils.

The lab experiment did have a time limit, but there is much to be learned from comparing the

different lift and drag aspects of all kinds of airfoils and what each is best suited for.

Page 15 of 15