Page 1 of 6

Matthew Rausch

MNE 490-01: Fundamentals of Flight

Project 1 Report: Due Nov. 7th, 2024

Systematic Geometry Study

In this geometry study, three 4-digit NACA airfoils were compared to each other. These
airfoils are NACA 1412, 2412, 3412. The difference between these three airfoils is their camber. The
1412 airfoil has a mean camber that is 1% of the chord, the 2412 has a mean camber that is 2% of
its chord, and similarly, the 3412’s camber is 3% of its chord. The purpose of this study was to
understand how geometry affects properties of lift and drag for a given airfoil. Each airfoil was
tested at a Reynolds number of 2.5e6, Mach number of 0, and a range of angle of attacks from -5° to
24°.

Figure 1: Xfoil polar of NACA 3412, 2412, and 1412.

 Page 2 of 6

Figure 2: Xfoil polar with adjusted drag coefficient axis

Figure 1 above shows a polar generated from Xfoil comparing the three NACA airfoils.
Looking first at the lift curve, the slope of the linear portion of the curve’s are very similar.
However, the NACA 3412 has the highest lift coefficient for a given angle of attack, while the
1412 has the lowest. This also means that the 3412 has the highest clmax, and similarly the 1412
has the lowest clmax, occurring at roughly the same angle of attack. Given that the 3412 airfoil
has a greater camber, we would expect to see higher lift coefficients for a given angle of attack
compared to the 2412 and 1412 airfoils which have less camber.
The effect of increasing camber also changes the zero-lift angle of attack. The 3412
airfoil has a lift coefficient of 0 at an angle of attack of -3°, while the 2412 and 1412 have zero
lift angle of attack equal to -2° and -1° respectively. Again, these are results we would expect to
see since an airfoil with a greater camber naturally generates lift at a zero angle of attack due to
 Page 3 of 6

its curvature. Thus, a greater cambered airfoil must achieve a more negative angle of attack to
counter its natural tendency to generate lift.
While the NACA 3412 can generate more lift for a given angle of attack, the tradeoff is
that its stall is more abrupt. The 3412 airfoil has a steeper slope after its clmax meaning there is a
more drastic loss in lift with increasing angle of attack. The 1412 airfoil shows the best stall
characteristics given that its slope after clmax isn’t as steep as the 2412 or the 3412. This trend of
poor stall characteristics with increasing camber is because flow separation, which causes stall,
tends to occur earlier on a more cambered airfoil. We can also see this by looking at the
boundary layer plot from the Xfoil polar in figure 1. The 1412 airfoil transitions to turbulence
very close to the leading edge, meaning almost all the flow over the top surface is turbulent,
which helps to delay flow separation. The 2412 transitions at roughly 25% of the chord, while
the 2412 transitions closer to 50% of the chord, which allows flow separation to occur more
easily leading to a more abrupt stall.
Figure 2 shows the CL vs CD graph from the Xfoil polar, with the CD axis adjusted to
more clearly show the lift coefficient at minimum drag. The figure shows the minimum drag
coefficient for each airfoil at roughly 50e4. At this drag coefficient, the 3412 has the highest lift
coefficient of roughly 0.65. The 1412 has the lowest lift coefficient for minimum drag at less
than 0.5, while the 2412 was at 0.5. As we have seen so far, a more cambered airfoil produces
more lift, so it makes sense that the 3412 airfoil has the highest lift coefficient for minimum drag.
Finally, the best option for reduced drag on an airplane that cruises at a lift coefficient of
0.3 would be the NACA 1412 airfoil. From the polar shown in figure 2, you can see at a lift
coefficient of 0.3, the 3412 airfoil has a higher associated drag coefficient. The 2412 and 1412
curves almost intersect each other at that cruise lift coefficient, so it is hard to tell which one will
produce less drag. To get a more accurate answer Xfoil was used to report the total drag for each
airfoil at a lift coefficient of 0.3. Xfoil reported the total drag on the 3412 foil to be 59.3 counts,
53.1 counts for the 2412 foil, and 51.3 counts for the 1412 airfoil. Based on this data, the NACA
1412 airfoil would be the best option for minimizing drag at cruise lift coefficient of 0.3.

Optimization Study
Part two of this project was an optimization study with the goal being to reduce drag as
much as possible. A NACA 2412 airfoil operating at a Reynolds number of 3e6, a Mach number
of 0, and a lift coefficient of 0.3 was used as a baseline operating condition. Different geometric
parameters of the foil were altered in an attempt to reduce the drag on the foil at the specific
operating point. Additionally, as a structural constraint, the minimum thickness of the foil must
be greater than 12%.
 Page 4 of 6

Figure 3: Baseline NACA 2412 airfoil compared to final design

Table 1: Comparison between baseline and final design

VARIABLE Baseline Design19

leading edge radius 0.013 0.0095
position of max thickness upper 0.314 0.5
max thickness upper 0.079 0.079
curvature upper -0.251 0.5
position of max thickness lower 0.2 0.5
thickness lower 0.041 0.041
curvature lower -0.479 -0.7
trailing edge gap [DO NOT
CHANGE] 0.001 0.001
delta yte [DO NOT CHANGE] 0 0
trailing edge angle 3.185 2
trailing edge internal angle 14.269 14.269
skin friction drag: Cdf 0.00489 0.00288
pressure drag: Cdp 0.00085 0.00063
total drag: Cdt 0.00574 0.00351

A total of 22 designs were made for this optimization study, however design 19 had the
best combination of drag coefficients, with values shown in the table above. The skin friction
drag coefficient was reduced by 0.00201, the pressure drag coefficient was reduced by 0.00022,
and the total drag coefficient was reduced by 0.00223. While all drag coefficients were reduced
by a significant amount, it is worth noting that while some designs were able to achieve a lower
pressure drag coefficient, design 19 had the lowest skin friction and total drag coefficients.
As you can see in the table above, many of the variables were changed in design 19. The
only variables that did not change were max upper thickness, thickness lower, trailing edge gap,
delta y, and trailing edge internal angle. To create design 19, I used a systematic approach,
changing only one variable at a time until that single variable was optimized. For example, I
started by changing the position of max upper thickness and found that by pushing this position
backwards, the drag coefficients would decrease. I continued to alter this variable until there
 Page 5 of 6

were no more improvements in drag, then moved onto the next variable and repeated the same
process. Throughout this process I found that some variables needed to increase, while others
needed to decrease to reduce drag, and some variables did not need to change at all.
What is interesting about my final design is that its shape closely resembles that of a
laminar airfoil design. You can see this in figure 3 above. Laminar airfoils are thinner in the front
and have a max thickness closer to the center of the airfoil. This creates a favorable pressure
gradient just after the leading edge, which allows for more laminar flow over the foil. This can be
seen in figure 4 below showing the Cp plot of design 19. Just after the leading edge there is a
slightly favorable pressure gradient until 50% of the cord. You can also see that transition on
both the upper surface and lower surface occurs very close to the trailing edge, meaning most of
the flow over the airfoil is laminar. This significantly reduces skin friction drag since laminar
boundary layers move slower and create less shear stress, leading to less skin friction.

Figure 4: Cp plot of Design 19.

Additionally, shown below in figure 5 is a polar comparing the baseline airfoil and design 19.
Focusing on the lift curve of the two airfoils, an interesting result is that design 19 has a slightly
steeper slope than the baseline but have the same zero lift angle of attack of -3°. This means that at
low angles of attack, the baseline and design 19 produce the same lift coefficient. However, at
higher angles of attack the baseline has a higher lift coefficient and CLmax. Furthermore, design 19
has a very abrupt and drastic stall compared to the baseline. It is the type of flow over the airfoil that
is giving rise to these differences. On design 19, flow over the top and bottom surface is almost
 Page 6 of 6

entirely laminar, as explained earlier. While this has benefits in decreasing drag, its tradeoff is that
laminar flow separates much more easily than turbulent flow, which is the dominant type in the
baseline foil. When flow separates there is a dramatic loss in lift and stall occurs which is what we
are seeing in the lift curve comparison.

Another result of the polar comparison is that design 19 has a lower lift coefficient for
minimum drag than the baseline at around 0.3. Given that design 19 was specifically designed to
have lower drag at that specific lift coefficient, we should expect to see this from the polar
comparison. Finally, you can also see from the polar that the boundary layers on design 19 is
almost entirely laminar giving rise to flow separation at relatively low angles of attack. Conversely
transition occurs earlier on the baseline foil leading to more turbulent flow, which can help delay
flow separation.

Figure 5: Polar comparison of Baseline and Design 19.

After comparing polars and Cp plots of both the baseline and design 19, it is clear each
airfoil has a specific designed function. The baseline tends to perform better at a higher lift
coefficient given that its drag is at a minimum at a higher lift coefficient. Additionally, it has better
stall characteristics and can handle high angles of attack. On the other hand, design 19 operates
best at a low lift coefficient and angle of attack, since its drag is minimized at a lower lift coefficient
and has poor stall characteristics. This idea is supported by the boundary layer plot on the polar
showing that there is mostly laminar flow on design 19, while the baseline transitions too turbulent
earlier. A more laminar boundary layer significantly decreases skin friction drag, but separates
more easily.