Professional Documents
Culture Documents
AME 441
Group 39
Geoff Larson
Maxworthy
Fall 2009
[FANWING]
Exploring the benefits of an active‐lift‐generating airfoil design, featuring an integrated tangential fan
Abstract
A fanwing is an experimental type of airfoil designed to utilize a tangential fan mounted within
its leading edge to improve performance characteristics. Tangential fans offer opportunities for
distributed propulsion and flow control due to the ability to install them horizontally within an airfoil
perpendicular to the flow, accelerating air across the entire span of the airfoil. The inlet intakes portions
of the boundary layer, and exhausts expel over the upper camber of the airfoil, producing lift, and
through the wake at the trailing edge, reducing drag and inhibiting stall. By actively accelerating air, a
thrust is also produced. It was found that the fanwing had a higher coefficient of lift than the control
airfoil at all tested air speeds and fan angular velocities, over all tested angles of attack. At constant air
speed, there also existed a trend of increased lift, stall inhibition, and thrust with increased angular
velocity.
1
Table of Contents
Introduction …………………………………………………………………………………………………………………..2
Experimental Technique ……..……………..…………….…………………………………………………………………3
Fabrication …………………………………………………………………………………………………………………..3
Laboratory Procedure ………………………………………………………………………………………………………6
Results and Discussion ………………………………………………..…………………………………………………….8
Lift Support ………………………………………………………………..………..……………………………………….8
Stall Inhibition ………………………………………………………………..………..…………….…………………….14
Thrust Production ……………………………………………………..……….……………….…………………..16
Conclusion ……………………………………………………………………………………..…………..….………………………18
Appendix ……………………………………………………………………………………..………………………………………19
2
Introduction
A fanwing is an experimental type of airfoil designed to utilize a tangential fan mounted within its
leading edge to improve performance characteristics. Tangential fans offer opportunities for distributed
propulsion and flow control due to the ability to install them horizontally within an airfoil perpendicular
to the flow, accelerating air across the entire span of the airfoil. The inlet intakes portions of the
boundary layer, and exhausts expel over the upper camber of the airfoil, producing lift, and through the
wake at the trailing edge, reducing drag and inhibiting stall.
The increase in flow speed over the upper surface of the airfoil creates lift by reducing pressure
according to Bernoulli’s principle for an airfoil:
1/2 ^2 1/2 ^2
Equation 1
where ρ is fluid density, V is the velocity of the airflow relative to the airfoil, and p is the static pressure.
Figure 1. Per Bernoulli's equation, lift is produced by a pressure differential between
the upper and lower surfaces of a wing.
To produce lift according to Bernoulli’s principle, the fanwing must do one or both of two things:
1. Increase the flow rate over the upper camber, or
2. Decrease the flow rate over the lower camber.
Both increase the static pressure differential between the upper and lower cambers of the airfoil, but to
see why the first method is preferable to the second, the tangential fan is considered as a rotating
cylinder subject to the Magnus effect. As the cylinder exposed to air rotates, skin friction “pulls” air
around with it perpendicular to its axis of rotation. This increases flow speed over the upper surface,
but also increases drag over the lower surface. For more efficiency, the
fanwing features an integrated shield along the underside of the
leading edge of the airfoil, eliminating the drag associated with the
Magnus effect. Accordingly, the fanwing can accelerate air over the
upper surface of the airfoil without inducing drag over the lower
surface, resulting in an improved lift‐to‐drag ratio / by increasing
while maintaining the same as a traditional airfoil. 1
Figure 2. Basic fanwing profile.
1
http://www.fanwing.com/FanWing0202b.gif
3
Experimental Technique
I. Fabrication
A. Airfoil Construction
Two identical airfoils were made for this laboratory investigation,
but a longitudinal cutout was removed from the leading edge of
one of them to accommodate a fanwing’s tangential fan. First, high
density polyurethane foam was purchased, from which to cut the
body of the airfoils. To get the correct profile, the proper NACA
airfoil curve was put into Dassault SolidWorks, scaled to the desired
size, and printed out onto plywood templates. The templates were Figure 3. Fanwing airfoil during construction.
used to hot wire cut the foam with precision. From there, the
airfoils were coated with a shell of fiberglass cloth topped with mylar, and encased in a vacuum back for
curing. The end result left accurate airfoils with smooth surfaces and structural integrity. The airfoil
dimensions were as follows:
• NACA 0015 profile
• Chord length of 10 in.
• Span of 12 in.
B. End Plates
Since the experiment was to test airfoil sections rather than
entire wings, end plates were attached to simulate a section
of an infinite wing. When air flows over a wing, it “spills”
over the ends of the wing tip. This increases pressure over
the upper tip surface and decreases pressure over the lower
tip surface, which both work against the lift being generated
by the wing near its end. By adding proper end plates, this
leakage does not occur, and the wing behaves as if it were a
section of an infinitely‐long span. From research done on
Formula‐1 race car rear spoilers, end plate geometry was
calculated from an optimal height‐to‐span ratio. The final
dimensions were:
• Height of 7.2 in. 2
2
http://history.nasa.gov/SP‐367/fig51.jpg
4
• Length of 12 in. (10 in. chord, plus 1 in. extension
Figure 4. Pressure‐driven "spillage" without end plates.
front and rear)
C. Tangential Fan
An LTG TA‐40/300/24V tangential fan was chosen for use in the fanwing for several reasons:
3
Figure 5. Schematic sketch of tangential fan chosen.
o Diameter. We were able to acquire 2 in. thick polyurethane foam from which to cut the airfoils.
This limited out maximum chord length, but also the thickness by scale. We chose a fan with a
diameter of 40 mm because its radius most nearly approximated that of the lower leading edge
of the largest airfoil we could construct from available materials.
o Span. The interior width of the Biegler Hall wind tunnel measured 18 in. We chose a fan with a
length of 12 in. to span a majority of the wind tunnel width, but leave clearance for an exterior
motor and mounting bracket assembly. In this way, the effective airfoil span could be
maximized while ensuring accurate data.
o DC Connectivity. To achieve useful results, the fanwing had to be tested with its fan rotating at
constant, observable angular velocities. However, with different air speeds and shifting angles
of attack, the fan’s angular velocity would not remain constant without feedback control. With
a 24 volt DC motor, a linear potentiometer wired in a simple voltage divider circuit could
manually adjust the potential across the motor’s terminals and control its speed.
o Price. The tangential fan chosen can be purchased online for less than $10.00 USD.
3
http://www.ltg‐inc.net/fileadmin/_temp_/downloads/dokumentationen/prozesslufttechnik/
hochleistungsventilatoren/querstrom‐ventilatoren/Tangential_Fans_GA‐TA‐TEt_25...60‐US.pdf
5
D. Force Balance Mount
A customized bracket was designed for this experiment for mounting the fanwing to the wind tunnel
force balance. Because the fan wing was larger than most airfoils tested and covered the regular
mounting hardware holes on the force balance, the piece also served as an end plate mounting adapter.
The adapter was designed in SolidWorks and 3‐D printed specifically for this laboratory investigation.
However, the thin tubular extension required reinforcement with several layers of fiberglass cloth to
handle the forces of this experiment.
Figure 6. Mount 3‐D printed in Biegler Hall. The piece bolts to the force balance through the recessed holes.
6
II. Laboratory Procedures
A. Electronic Control Module
As mentioned, in order to take accurate measurements for ω at different states, angular
velocity had to be controlled independently of other variables. To accomplish this, a custom
control panel was assembled, featuring a Venom hobby speedometer with digital LCD readout.
The device read a small magnet attached to the side wall of the fan with an inductive pickup
sensor, and calculated its speed from that signal. The sensor was countersunk into the
endplate. The controller read instantaneous fan speed in MPH, from which RPM values were
later calculated.
4
Figure 7. Magnet‐sensor orientation. Figure 8. Venom speedometer LCD readout.
The speed was visually monitored, and constant angular velocity was maintained by actively
adjusting a potentiometer wired in a voltage divider orientation. Because the chosen fan motor
was DC, voltage from an external power supply could be precisely controlled in this manner.
The motor’s electrical potential was governed by the equation:
Equation 2
The control panel was moderated by an automotive relay, and powered by a DC power supply.
4
http://www.venom‐group.com/mmVENOMGROUP2/Others/VSM‐instructions%20v2.0%20VEN%200615.pdf
7
Figure 9. Motor adjustment and fan speed feedback control panel, with a linear potentiometer resistive voltage divider
circuit wired in‐line.
B. Test Procedure
The laboratory experiment was divided into two phases. Phase one tested the control NACA 0015 airfoil
in the wind tunnel. The data‐acquisition procedure was as follows:
1. Air speed was set at 4 m/s.
2. Angle of attack was set at (‐5) degrees.
3. An automated Labview data collection program ran the airfoil through a range of angles of
attack between (‐5) degrees and 35 degrees —40 total data points—in steps of one degree,
recording force balance readings at each step.
4. When the program completed, the stepper motor returned to its original position at (‐5)
degrees. The air speed was then increased to 6 m/s.
5. Steps 1 – 4 were repeated for 8 m/s, 10 m/s, and 12 m/s.
Phase two tested the fanwing in the wind tunnel. The data acquisition procedure was repeated in a
similar manner. However, at each air speed, the program was run once for each of several fan angular
velocities:
1. Air speed was set at 4 m/s.
2. Fan angular velocity was set at 2,000 RPM.
3. Angle of attack was set at (‐5) degrees.
4. An automated Labview data collection program ran the airfoil through a range of angles of
attack between (‐5) degrees and 35 degrees —40 total data points—in steps of one degree,
recording force balance readings at each step.
5. When the program completed, the stepper motor returned to its original position at (‐5)
degrees. The fan angular velocity was then increased to 2,500 RPM.
6. Steps 1 – 5 were repeated for 3,000 RPM, 3,500 RPM, 4,000 RPM and maximum fan speed.
7. The air speed was increased to 6 m/s.
8. Steps 2 – 7 were repeated for 8 m/s, 10 m/s, and 12 m/s.
Data could not be taken below 2000 rpm because the fan could not be predictably controlled. Data
collection was attempted for all available scenarios, but it was found that as air speed was increased
beyond 8 m/s, the motor was no longer powerful enough to reach comparable angular velocities.
Therefore, data was not collected above 3000 RPM at 8 m/s, or above 2,000 RPM at 10 m/s and 12 m/s.
8
RESULTS & DISCUSSION
I. Lift Support
A. Background
Modern aircraft employ a number of techniques for reducing drag and increasing lift, including slats,
slots, flaps, and spoilers. They facilitate lower take off and landing speeds, which improve the safety of
commercial flights and the capabilities of personal aircraft and military UAVs (unmanned aerial vehicles).
For a typical wing to sustain level flight, it must create a lift force equal to the total weight of the
aircraft. To minimize flight speed, the dimensionless coefficient of lift must be maximized, according
to the equation:
1
2
Equation 3
where is the lift force, ρ is the density of the fluid flow, V is the relative velocity of the airfoil through
that fluid, and S is the planform area of the airfoil.
After rearranging, the relationship between minimum flight speed and maximum can be expressed
by:
2
,
Equation 4
Since is equal to the weight of the plane in this flight situation, and the density is considered to be
constant, the can only be decreased by an increase in either , or the wing area. Traditionally,
this problem has been alleviated by the aerodynamic devices listed above, but the fanwing offers a new
option.
9
B. Control Results
was plotted against angle of attack α of an airfoil. For this experiment, the control NACA 0015 airfoil
could be plotted against α at three air speeds with the following results:
2.5
1.5
4 ms
C_L
1 6 ms
8 ms
0.5
0
‐5 0 5 10 15 20 25 30
‐0.5 α
Figure 10. Coefficient of lift plotted against angle of attack for the control airfoil, at various air speeds.
Several interesting observations were made from Figure 10 and the data taken in the first phase of the
laboratory investigation:
1. As air speed increased, the value increased at all angles of attack.
2. The curves began rising with positive linear slopes, until reaching an angle of attack at which
they suddenly dropped off. This is known as the “critical” or “stall” angle of attack, and provides
a graphical representation of the stall angle of the airfoil at a given air speed. Stall
characteristics will be discussed later.
3. After “falling off,” the curves began rising again in a relatively linear manner, but now at angles
of attack beyond expected values of stall.
Observation 1:
There are two components to the drag on an airfoil, one of which is the drag induced by the creation of
lift, according to:
Equation 5
where K is a term determined by the airfoil design.
Recall from Equation 3 that is inversely proportional to , so increasing the velocity term decreases
the induced drag. This is why there is an increase in as air speed increases, for the control airfoil.
10
Observation 3:
Stall characteristics for Observation 2 will be discussed later in detail, but the second increase in
beyond stall characteristics is worth noting. As angle of attack increased significantly, an unexpected
increase in lift coefficient continued. As will be seen, this trend existed for the control and the fanwing,
at all air speeds. It was determined to be the result of ground effects on the airfoil at higher angles of
attack, in part due to the chord length of the airfoil being large relative to the wind tunnel width. Figure
11 shows a general observation of the airfoil’s proximity to the wind tunnel wall at higher angles of
attack. The effect was more pronounced as air speed increased.
Air flow Air flow
Force
Figure 11. Ground effects caused skew in lift data at higher angles of attack and higher air speeds.
11
C. Fanwing Results
With the control airfoil now tested and analyzed, the fanwing was tested in the wind tunnel. Upon
graphing of the data acquired for air speed of 4 / , the difference in performance of the fanwing
was visually obvious:
3
2.5
2 Control
2000 RPM
1.5
C_L
2500 RPM
1 3000 RPM
Fan Max
0.5
0
‐5 0 5 10 15 20 25 30
α
‐0.5
Figure 12. Coefficient of lift plotted against angle of attack at air speed ν=4 m/s.
The fanwing was significantly higher than the control. To quantify the improvement, comparisons
will be made between the control , and the fanwing at the same angle of attack. Table 1
displays the results that were found. For the minimum tested angular velocity of 2,000 RPM, the
more than doubled. For maximum fan speed, the more than tripled that of the control airfoil.
Table 1. C_Lmax of the control airfoil and the fanwing, for ν=4 m/s.
Airfoil RPM α
Control 0.48
2000 1.09
2500 1.21 13 deg.
Fanwing
3000 1.30
Max 1.61
12
To check the validity of these results, data was collected for two more air speeds:
3
2.5
2 Control
2000 RPM
1.5
C_L
2500 RPM
1 3000 RPM
Fan Max
0.5
0
‐5 0 5 10 15 20 25 30
‐0.5 α
Figure 13. Coefficient of lift plotted against angle of attack at air speed ν=6 m/s.
3
2.5
2 Control
2000 RPM
1.5
C_L
2500 RPM
1 3000 RPM
Fan Max
0.5
0
‐5 0 5 10 15 20 25 30
‐0.5 α
Figure 14. Coefficient of lift plotted against angle of attack at air speed ν=8 m/s.
Table 2 shows the coefficient of lift at the control , for a range of fan speeds. The fanwing
consistently has a larger , supporting the legitimacy of the data from the slower air speed.
13
Table 2. C_Lmax of the control airfoil and the fanwing, for ν=6 m/s and ν=6 m/s.
C_L
C_L
1 1 1
4 ms 4 ms 4 ms
0 0 0
‐5 0 5 10 15 20 25 30 ‐5 0 5 10 15 20 25 30 ‐5 0 5 10 15 20 25 30
‐0.5 α
‐0.5 α ‐0.5 α
Figure 15. Coefficient of lift plotted against angle of attack over range of air speeds, for three fan ω values.
Figure 15 shows the plotted against angle of attack at different air speeds for additional fan angular
velocities. When ω increases, so does for all air speeds tested. As Equation 1 predicted, the
increasing lift coefficient with increased fan angular velocity shows a direct correlation between the
acceleration of the flow over the upper surface of the airfoil and the production of lift.
As with the control airfoil, Equation 5 explains why lift coefficients are larger for faster air speeds.
However, with the fanwing, this is only true up until approximately the airfoil stall angle. Above the
traditional stall angle, the fanwing actually performed better at lower air speeds.
14
II. Stall Inhibition
Accelerating air over the upper surface of the airfoil does more than produce lift, it also inhibits stall.
Returning to the Observation 2 made when taking data on the control airfoil, there was consistent
evidence found to support such a conclusion. At 6 m/s, the control curve in Figure 16 begins with a
distinct positive linear slope, followed by a sharp drop in at an alpha of 16 degrees:
3
2.5
2
Control
1.5 2000 RPM
C_L
2500 RPM
1
3000 RPM
0.5
0
‐5 0 5 10 15 20 25 30
‐0.5 α
Figure 16. Coefficient of lift vs. α for ν=6 m/s, clearly demonstrating control stall.
The fanwing curves continue their linear increase for longer, and do not “drop off” like the control
within the studied range of α. This shows two effects:
• The fanwing “smoothes” over the stall angle of the control, and
• The fanwing moves its critical stall point to a higher angle of attack.
When a wing increases its angle of attack, the turbulent wake behind it grows in size and moves forward
toward it. When the wake reaches the wing’s trailing tip, a condition called “separation” occurs as the
flow over the top and bottom surfaces are separated from each other and no longer converge
downstream. The point of separation moves gradually up the wing’s upper surface, increasing drag.
When the wing reaches critical stall angle, the point of separation suddenly shoots up the remaining
length of the chord, causing a condition where lift drops off sharply while drag suddenly spikes. The
control’s drop in lift can be seen in Figure 16, while the complimentary 6 m/s graph of below of
against α shows the spike in drag begin at the same 16 degree angle of attack:
15
0.8
0.6
Control
0.4 2000 RPM
C_D
2500 RPM
0.2 3000 RPM
0
‐5 0 5 10 15 20 25 30
‐0.2 α
Figure 17. Coefficient of drag vs. α for ν=6 m/s, clearly demonstrating control stall.
In more detail, the boundary layer is the region of a flow passing nearest the surface of the airfoil. At
the surface, skin friction effects reduce the velocity to zero. As distance from the surface increases, the
velocity increases until it reaches the maximum velocity on the upper camber. Stall occurs when the
velocity of the airflow has been reduced so much that pressure effects cause it to reverse direction. On
a larger scale, this causes flow separation and a loss of lift.
5
Figure 18. Propagation of turbulent wake as alpha increases.
The fanwing does not experience this loss of lift and spike in drag, though. The reason for this lies in
another method for increasing , called boundary layer control. The process replaces slower low‐
energy segments of the boundary layer along the upper surface of the wing with high‐energy flow
moving at a faster speed. By accelerating air over its upper camber, the fanwing fights off the
propagation of turbulence. As a result, a more‐laminar flow stretches down the surface of the airfoil
longer than the control, delaying separation and stall until a larger angle of attack.
5
http://history.nasa.gov/SP‐367/fig48.jpg
16
III. Thrust Production
One of the causal effects of accelerating airflow toward the trailing edge of the airfoil is that a small
vectored thrust component is generated. Thrust causes asymptotic behavior in the relationship of CL/CD.
When graphed against α, asymptotic peaks and valleys appear. What these represent is the angle of
attack at which the fanwing’s drag exceeds its produced thrust. A traditional airfoil should not have this
asymptote because it does not produce any thrust.
For each plot of CL / CD, asymptotic behavior occurred where drag began to exceed thrust, and the
theoretical value of the lift‐to‐drag ratio approached infinity. At lower air speeds, the fanwing’s thrust
was able to overcome drag associated with larger angles of attack. As the flow velocity increased,
however, the asymptotes approached α = 0.
30
25
20
15
10
4 ms
C_L / C_D
5
6 ms
0
‐5 0 5 10 15 20 25 30 8 ms
‐5
α
‐10
‐15
‐20
‐25
‐30
Figure 19. L / D vs. angle of attack for fan angular velocity of 2,000 RPM.
30
25
20
15
10
5 4 ms
C_L / C_D
0 6 ms
‐5 0 5 10 15 20 25 30
‐5 8 ms
α
‐10
‐15
‐20
‐25
‐30
Figure 20. L / D vs. angle of attack for maximum fan angular velocity.
17
The reason why this occurs is that increasing the fan angular velocity increases thrust by moving more
air and at a higher velocity. This increases thrust by the equation:
Equation 6
where T is vectorized thrust and v is vectorized exhaust velocity relative to the flow.
Increasing the mass of air moved increases thrust. Increasing the velocity of air also produces thrust. By
moving additional amounts of air over the upper surface of the airfoil at speeds greater than the rest of
the flow, a fanwing produces a measurable amount of thrust in addition to lift.
18
Conclusion
The fan wing design has benefits over a traditional airfoil. The primary benefits include:
o Generation of additional lift,
o Producing a nominal thrust, and
o Inhibition of stall characteristics at high angles of attack.
By increasing the coefficient of lift at a given air speed, the fanwing has the ability to take off and land at
lower speeds. The additional lift at all speeds also means that a fanwing aircraft can carry a significantly
larger payload than a traditional fixed‐wing aircraft.
By inhibiting stall characteristics, the fanwing can maintain flight at larger angles of attack than a
traditional wing. The ability to fly at larger angles of attack, combined with the existence of lift at near‐
zero air speed also means that the fanwing design has the potential to achieve nearly vertical take off
and landing (VTOL) with improved design.
By generating thrust, the fanwing has the potential to eliminate the need for other propulsion sources,
such as propellers.
The characteristics of a fanwing may have many practical applications. Military unmanned aerial
vehicles (UAVs) may benefit from the shorter take off distance and slower flight speed capabilities. On a
larger scale, remote locations may improve their access to others, including emergency services, with
fanwing aircraft. They may benefit from the ability to take off and land on shorter runways due to near
VTOL and more extreme terrain due to reduced flight speed. A fanwing aircraft may be more capable of
landing in forest clearings, on dirt roads, or even on water. Some cargo transportation may also be done
more efficiently because of the higher lift coefficient.
However, the energy consumption of the fanwing was not tested in this laboratory investigation, so the
practicality of larger scale fanwing use relies on continued research. Some changes to consider for
future investigation include:
o Monitor electrical power consumption.
o Find exact location of separation as it occurs.
o Use more powerful motor to collect data at higher air speeds and angular velocities.
o Increase diameter of fan used.
o Investigate alternative airfoil profiles and fan mount locations along them.
19
Appendix
I. Determining Fan Speed
The Venom Speed Meter utilizes an inductive magnetic sensor to determine RPM. However, because
the meter was designed for vehicles, the readout it provides is in mph rather than RPM. As a result, it
was necessary to convert the readout from mph back to RPM.
To calibrate the meter, the circumference of the vehicle’s wheels must be selected. For our purposes, a
circumference of 10.56 inches was selected. The reasons for this will soon become apparent.
5280 12
1 1
1 100
60 10.56
1 1
Equation 7
Therefore, the readout in miles per hour could simply be multiplied by 100 to attain RPM. For example,
a readout of 20mph corresponds to 2,000RPM. By doing this, the RPM could be easily selected by
adjusting the voltage input and observing the Venom Speed Meter readout. Complicated conversions
were no longer necessary to perform on the spot.
II. Lift and Drag Coefficients
Measurements of lift and drag were converted into lift and drag coefficients by the following,
1
2
Equation 8
1
2
Equation 9
By definition, an infinite wing would have an infinite aspect ratio because, / , where b is the
infinite wingspan and S is the wing area. The aspect ratio is related to the coefficient of induced drag by
the following equation.
Equation 10
For an infinite wing, where ARÆ∞, it can be seen that Æ0.
20
III. Uncertainty Analysis
A. Fan Angular Velocity
The readout on the Venom Speed Meter has a resolution of ±1 mph. However, during the course of
each trial, the readout would often increase or decrease by 1 mph. The fan speed would then be
manually adjusted back to its initial value by regulating the input voltage. The ±1mph fluctuations of the
readout and the ±1mph resolution combine to give a total error of ±2mph. As discussed in section A.1,
the RPM of the tangential fan is 100 times the mph readout. Therefore the true RPM error is ±200.
B. Air Speed
The airspeed was set using a dedicated LabVIEW program that could be run continuously while the wind
tunnel was brought up to the correct speed for each trial. Because the free‐stream velocity was never
changed during the trials, it was never necessary to calibrate the pitot tube for the data acquisition
program. The result was that the velocity measurements which were automatically collected at each
angle of attack are off by a constant. However, they do provide us with a good indication of the
standard deviation. The standard deviation was estimated using data collected from the control airfoil.
It was calculated for each airspeed and yielded an average result of ΔV= ±0.2 m/s.
C. Lift and Drag
Uncertainty propagation can be performed by Taylor series expansion around the variables in Equation
8 and Equation 9. The uncertainty for the lift and drag coefficients can be derived as follows,
∆ ∆ ∆ ∆ ∆
∆ ∆ ∆ ∆ ∆
Compared with the uncertainties in L, D and V, the uncertainty in ρ and S are negligible. Therefore, the
above equations simplify to,
∆ ∆ ∆
Equation 11
∆ ∆ ∆
Equation 12
21
After calculating the partial derivatives,
2 4
∆ ∆ ∆
Equation 13
2 4
∆ ∆ ∆
Equation 14
where ΔL and ΔD are the uncertainties in the force balance measurements and ΔV is the uncertainty
found in air speed uncertainty analysis.
1
0.8
0.6
0.4
0.2
0
0 0.02 0.04 0.06 0.08 0.1 0.12
‐0.2
‐0.4
‐0.6
‐0.8
Figure 21. Drag polar.
6
Figure 22. Drag Polar
6
http://www.utne.com/uploadedImages/utne/blogs/Environmentalism/Polar%20bear%20drag%20feet%281%29.jpg