Sunteți pe pagina 1din 5

CFD Project

Nicholas Elkins
06/06/2014
MAE 150B
Spring 2014
Professor Colonno


Analysis
To calculate my values for C
l
, C
d
, C
m,c/4
, and x
cp
I ran the data through a program in
Matlab. First off, I calculated the midpoint, the slope and the normal vector for each panel. To
calculate the normal vector I used the relationship between it and the slope of the panel. I found
that the angle of the panel is the inverse tangent of the slope, and this angle is also the same as
the normal vector angle. I then broke the airfoil into four parts: slope>0 and y>0, slope<0 and
y>0, slope>0 and y<0, and slope<0 and y<0. This allowed me to find the relationship between
the angle and normal vector for each panel. I ensured that each vector pointed outward using
proper if statements. I also calculated a second normal vector for the skin friction coefficient. It
was also found using the panel angle. I ensured that the x portion of the vector was always
positive, and the y portion was determined by breaking the airfoil into four sections again. In the
same for loop I interpolated C
p
and C
f
at each panel midpoint.
I used equations 1.15, 1.16, and 1.17 as the basis for my numeric integration of the
surface of the airfoil. It was easier to calculate C
n
and C
a
and then use equations 1.18 and 1.19 to
find C
l
and C
d
. I estimated all integrals as a sum. When a dx was needed I estimated it as the
absolute value of the change in x from one end of the panel to another. Dy was estimated in the
same manner. In the case of finding the value for c
n
I multiplied C
p
by its negative normal in the
x-direction, I multiplied C
f
by its normal in the x-direction, and then I added the two together and
multiplied by dx, summing in each iteration of the loop. I used the same process to calculate C
a

except that dy replaced dx, and the normals in the y-direction replaced the normals in the x-
direction. The process for calculating C
m,c/4
was similar. I had to calculate the moment arms in
the y and x-directions. This was done by subtracting the midpoint of the panel from the quarter
chord point, (1/4,0). I ensured that I multiplied the x portions of C
p
and C
f
by the arm in the y
and the y portions by the arm in the x. To get the x and y portions I multiplied by the normal in
that direction. The center of pressure was calculated by using equation 1.22 to relate the C
l
, the
C
m,c/4
, and x
cp
.
I used the same integration method to calculate my values using thin airfoil theory. The
only difference is that my dtheta was a constant instead of being different for each panel. I also
used the transformation using equation 4.21 to transform from x to theta. To calculate the
necessary values I used the equations 4.51, 4.61, 4.60, and 4.64. Each integral was approximated
in Matlab. I used a very, very small dtheta to ensure accuracy of my integral. I also checked each
one in Maple to make sure.




Results
Case 1 Inviscid
alpha Cl cd C
m,c/4
X
cp
/c
-5 -0.355998 0.011592 -0.060604 0.079763
-2.5 -0.065438 0.005417 -0.052246 -0.548413
0 0.221235 0.007351 -0.050110 0.476502
2.5 0.505132 0.016839 -0.053950 0.356804
5 0.786676 0.033677 -0.063590 0.330834
7.5 1.065753 0.058041 -0.078828 0.323965

Case 2 Inviscid
alpha cl cd C
m,c/4
X
cp
/c
-5 -0.349766 0.013289 -0.057043 0.086912
-2.5 -0.048230 0.005037 -0.047018 -0.724874
0 0.248721 0.007623 -0.044663 0.429571
2.5 0.543372 0.019217 -0.048913 0.340018
5 0.834086 0.039658 -0.059347 0.321152
7.5 1.111692 0.071012 -0.075761 0.318149

Case 1 Viscous
alpha cl cd C
m,c/4
X
cp
/c
-5 -0.344222 0.017954 -0.054449 0.090893
-2.5 -0.056734 0.012516 -0.045732 -0.556086
0 0.228229 0.014907 -0.043265 0.439567
2.5 0.510978 0.024802 -0.046911 0.341806
5 0.789684 0.042004 -0.056558 0.321621
7.5 1.061853 0.066392 -0.072047 0.317850

Case 2 Viscous
alpha cl cd C
m,c/4
X
cp
/c
-5 -0.335128 0.017606 -0.055213 0.085247
-2.5 -0.053935 0.011547 -0.045818 -0.599496
0 0.230062 0.014374 -0.043051 0.437128
2.5 0.512430 0.025520 -0.046583 0.340907
5 0.789481 0.044683 -0.056273 0.321279
7.5 1.052381 0.071999 -0.071784 0.318211

Thin Airfoil Theory
alpha cl C
m,c/4
X
cp
/c
-5 -0.3205 -0.0531 0.0843
-2.5 -0.0464 -0.0531 -0.8959
0 0.2278 -0.0531 0.4832
2.5 0.5020 -0.0531 0.3558
5 0.7761 -0.0531 0.3184
7.5 1.0503 -0.0531 0.3006

Thin airfoil theory seems to match up best with the inviscid case 1. For both inviscid
cases the x
cp
matches pretty well. Case 1 seems to match a bit better. The C
l
values for cases 1
and 2 match better around zero angle of attack. It seems that the higher the angle the farther
values are off. For both cases the C
m,c/4
values match up well. They hover above or below the
value that thin airfoil theory gives. The viscous cases also match up well with the thin airfoil
theory approximations, just not as well as the inviscid cases. The values of C
m,c/4
are quite close
to the approximations. The C
l
values also follow a similar trend and are close to the values. The
values of x
cp
match up better at higher angles of attack and are farther off at around zero angle of
attack, or normal flight conditions. It seems that thin airfoil theory is a good approximation, but
it is not exact. I think that thin airfoil theory matches up better with Case 1 because these are
closer to normal flight conditions. It is around atmospheric pressure and temperature. In Case 2
the flight conditions are simulating a much higher altitude. One thing about the tables that does
not look right is that the x
cp
at the angle of attack of -2.5 degrees is negative in all cases,
including thin airfoil theory. This is because the lift is close to zero at this angle of attack. When
the lift is close to zero the center of pressure moves outside of the body of the wing.
Case
L=0
Inviscid 1 -1.9293
Inviscid 2 -2.0940
Viscous 1 -2.0223
Viscous 2 -2.0250
Thin Airfoil -2.0773

As you can see, the alpha where zero lift occurs is close to that predicted by thin airfoil
theory for each case. However, Case 1 for inviscid is farther off than the other cases. I dont
know exactly why this is. Its still not off by a large amount. Case 2 for inviscid is very close to
the predicted value. The viscous cases are closer than Case 1 for inviscid but not quite as close as
Case 2. Thin airfoil theory makes some general assumptions about the geometry of the airfoil.
First off it assumes that the airfoil does not have large slopes or angles. This assumption is used
to approximate dz/dx as about theta and sin of theta as about theta. This means that if there are
large slopes in the airfoil, such as around the leading edge, thin airfoil theory will be farther off.

S-ar putea să vă placă și