interpolated
As an aerospace engineer but an amateur aerodynamicst, I'm always amazed at the mess one gets into when trying to find the profile of an older aircraft. I've tried it for the Messerschmitt Me 163B and the Pander S4 Postjager, and it was messy and confusing to say the least. For the Firebee it was equally bad. It required this web page to sort out the bits and pieces of information that I found.
I had five pieces of information to start with:
Jane's All The World's Aircraft 1960/61 reports for the Q-2A: 'Aerofoil section NACA 63, A014.6. Much later I found a Ryan Q-2A three-view that listed NACA 63 A014.6 too.
Jane's All The World's Aircraft 1970/71 reports for the BQM-34A: 'Wing section from leading edge .264 chord NACA 0009.932; from .264 chord to trailing-edge NACA 63A014.63. Thickness/chord ratio 14%'.
Standard Aircraft Characteristics BQM-34A Firebee (1 February 1968) reports: Incidence: Leading edge to .264 chord -5°51' / .264 chord to trailing edge 0°0'
From factory drawings I knew that the original Q-2A wing had a streamwise chord of 29.00", or a calculated 20.51" perpendicular to quarter chord. The Q-2C / BQM-34A wing had a drooped leading edge extension resulting in a streamwise chord of 33.40", or a calculated 23.62" perpendicular to quarter chord.
a Mark Nankivil photo of a BQM-34 wing tip without the tip cap.
A question mark was whether the profile was defined 'streamwise' or perpendicular to the quarter-chord. Torenbeek, page 436, states that it can be either: A wing section is formed by the external contour of a wing cross section with a plane parallel to the plane of symmetry of the wing, or a plane perpendicular to the quarter chord. From fitting profiles to photos I found out that the Firebee wing uses the second definition. It appears that this is (was) the common method in the American aircraft industry.
I wanted to start with the normal NACA 63-014 profile. Abbott & Von Doenhoff (p119-122) explains the coding system as follows:
6 denotes a '6 series' profile
3 denotes 'the chordwise position of position of minimum pressure in tenths of the chord behind the leading edge for the basic symmetrical section at zero lift'
0 denotes 'the design lift coefficient in tenths'
14 denotes 'the thickness of the wing section in per cent of the chord'
missing: subscript that denotes the low-drag range. One would expect something like NACA 632-014, or using an older notation 63,2-014
Since the design lift coefficient is zero, it seems logical that it's a symmetrical profile, which turned out to be true.
I could not find a listing of NACA 63-014 coordinates: nothing in Abbott & Von Doenhoff, or any NACA / NASA document. I do not know whether that means it does not exist at all. In the end I decided to do a linear interpolation of NACA 631-012 and NACA 632-015 data from Abbott & Von Doenhoff (pages 337 and 338). Since they are symmetrical the interpolation result should be accurate.
| NACA 631-012 | NACA 632-015 | NACA 63-014 interpolated |
| |
| 0 | 0 | 0 | 0 | |
| 0.5 | 0.985 | 1.204 | 1.131 | |
| 0.75 | 1.194 | 1.462 | 1.373 | |
| 1.25 | 1.519 | 1.878 | 1.758 | |
| 2.5 | 2.102 | 2.610 | 2.441 | |
| 5.0 | 2.925 | 3.648 | 3.407 | |
| 7.5 | 3.542 | 4.427 | 4.132 | |
| 10 | 4.039 | 5.055 | 4.716 | |
| 15 | 4.799 | 6.011 | 5.607 | |
| 20 | 5.342 | 6.693 | 6.243 | |
| 25 | 5.712 | 7.155 | 6.674 | |
| 30 | 5.930 | 7.421 | 6.924 | |
| 35 | 6.000 | 7.500 | 7.000 | |
| 40 | 5.920 | 7.386 | 6.897 | |
| 45 | 5.704 | 7.099 | 6.634 | |
| 50 | 5.370 | 6.665 | 6.233 | |
| 55 | 4.935 | 6.108 | 5.717 | |
| 60 | 4.420 | 5.453 | 5.109 | |
| 65 | 3.840 | 4.721 | 4.427 | |
| 70 | 3.210 | 8.934 | 3.693 | |
| 75 | 2.556 | 3.119 | 2.931 | |
| 80 | 1.902 | 2.310 | 2.174 | |
| 85 | 1.274 | 1.541 | 1.452 | |
| 90 | 0.707 | 0.852 | 0.804 | |
| 95 | 0.250 | 0.300 | 0.283 | |
| 100 | 0 | 0 | 0 |
NACA Report 903 says:
'.. the sides of the airfoil sections made straight from apprximately 80 percent chord to trailing edge'
'A special mean line, designated the a=0.8 (modified) mean line, has also been designed to maintain straight sides on the cambered sections'
'A is substituted for the dash'
It does not literally say so, but I read that NACA 63A014 should be a modification of NACA 63-014. Although this is maybe a big assumption, I wanted to find that out first. It would require a rather different thickness distribution, but I haven't anything on that subject. Instead, most of the discussion is about the mean line. Which is probably irrelevant here, since it is probably symmetrical (see below).
Again I could not find a listing of NACA 63A014 coordinates: nothing in Abbott & Von Doenhoff, or any NACA / NASA document. I decided to do a linear interpolation of NACA 631A012 and NACA 632A015 data from Abbott & Von Doenhoff (pages 344 and 345). Since they are symmetrical the interpolation result should be accurate.
NACA Report 903 says: 'Ordinates and theoretical pressure-distribution data for NACA 6A-series basic thickness forms having the position of minimum pressure at 30, 40, and 50 percent chord are presented in figure 2 for airfoil thickness ratios of 6, 8, 10, 12, and 15 percent. If intermediate thickness ratios involving a change in thickness of not more than 1 to 2 percent are desired, the ordinates of the basic thickness forms may be scaled linearly without seriously altering the gradients of the theoretical pressure distribution.'
NACA 63A014 is used on the Canadair CL-41 Tutor wing root profile (see Jane's), and NACA 63A014mod is used on Fokker F27 horizontal tail (see Torenbeek).
| NACA 631A012 | NACA 632A015 | NACA 63A014 interpolated | 
| |
| 0 | 0 | 0 | 0 | |
| 0.5 | 0.973 | 1.203 | 1.126 | |
| 0.75 | 1.173 | 1.448 | 1.356 | |
| 1.25 | 1.492 | 1.844 | 1.727 | |
| 2.5 | 2.078 | 2.579 | 2.412 | |
| 5.0 | 2.895 | 3.618 | 3.377 | |
| 7.5 | 3.504 | 4.382 | 4.089 | |
| 10 | 3.994 | 4.997 | 4.663 | |
| 15 | 4.747 | 5.942 | 5.544 | |
| 20 | 5.287 | 6.619 | 6.175 | |
| 25 | 5.664 | 7.091 | 6.615 | |
| 30 | 5.901 | 7.384 | 6.890 | |
| 35 | 5.995 | 7.496 | 6.996 | |
| 40 | 5.957 | 7.435 | 6.942 | |
| 45 | 5.792 | 7.215 | 6.741 | |
| 50 | 5.517 | 6.858 | 6.411 | |
| 55 | 5.148 | 6.387 | 5.974 | |
| 60 | 4.700 | 5.820 | 5.447 | |
| 65 | 4.186 | 5.173 | 4.844 | |
| 70 | 3.621 | 4.468 | 4.186 | |
| 75 | 3.026 | 3.731 | 3.496 | |
| 80 | 2.426 | 2.991 | 2.803 | |
| 85 | 1.826 | 2.252 | 2.110 | |
| 90 | 1.225 | 1.512 | 1.416 | |
| 95 | 0.625 | 0.772 | 0.723 | |
| 100 | 0.025 | 0.032 | 0.030 |
If NACA 63-014 versus NACA 63A014 are plotted in the same graph, the agreement for the forward 45% is obvious. The cusped trailing edge of NACA 63-014, and the 'straight' trailing edge of NACA 63A014 are clearly visible.
Abbott & Von Doenhoff reports in '6.6 Modified NACA Four- and Five-digit Series Wing Sections' (page 117): 'The suffix -63 indicates sections very nearly but not exactly the same as the sections without the suffix'. Now this probably only applies to four- and five-digit series profiles. Plus, without factory information, we can never find out how Ryan modified the NACA 63A014 profile for the Firebee.
Alternatively, because there's a dash and not a decimal preceding the '63', plus pointing at the Q-2A 'NACA 63, A014.6' airfoil tidbit, it could be that Ryan used a 14.63% thick version instead of the 'standard' 14% thick version. Because of this possibility, I also calculated the coordinates for the 14.63% thickness.
| NACA 63A014 interpolated | NACA 63A014.63 interpolated | 
| |
| 0 | 0.000 | 0.000 | |
| 0.5 | 1.126 | 1.175 | |
| 0.75 | 1.356 | 1.415 | |
| 1.25 | 1.727 | 1.802 | |
| 2.5 | 2.412 | 2.519 | |
| 5.0 | 3.377 | 3.531 | |
| 7.5 | 4.089 | 4.276 | |
| 10 | 4.663 | 4.876 | |
| 15 | 5.544 | 5.798 | |
| 20 | 6.175 | 6.458 | |
| 25 | 6.615 | 6.919 | |
| 30 | 6.890 | 7.205 | |
| 35 | 6.996 | 7.315 | |
| 40 | 6.942 | 7.257 | |
| 45 | 6.741 | 7.043 | |
| 50 | 6.411 | 6.696 | |
| 55 | 5.974 | 6.238 | |
| 60 | 5.447 | 5.685 | |
| 65 | 4.844 | 5.054 | |
| 70 | 4.186 | 4.366 | |
| 75 | 3.496 | 3.646 | |
| 80 | 2.803 | 2.923 | |
| 85 | 2.110 | 2.201 | |
| 90 | 1.416 | 1.477 | |
| 95 | 0.723 | 0.754 | |
| 100 | 0.030 | 0.031 |
What we now have is the wing profile of the Q-2A (USA) / KDA (Navy), the first generation Firebee. The Q-2C / BQM-34A wing was modified with a drooped leading edge extension.
I overlayed the photo with the 14.63% thickness version, and the fit was excellent. Note that the wing skins of the center box are fairly thick at 0.190" / 4.8 mm; therefore the profile drawing is positioned on their outside edges.
The transition between the NACA 0009.932 nose section and NACA 63A014.63 original profile is quoted to be at 26.4% chord. But what chord is that? It's either the original chord or the new extended chord. That extended chord can be calculated to be 23.62" / 20.51" = 115.15% of the original chord. A quick photo analysis made it clear that it was the latter.
The 'Structural Repair Instructions USAF model Q-2C target drone' (T.O.1Q-2C, 1 November 1960) contains a cross-section drawing that shows the positions of the spars webs. The drawing strangely ignores the leading edge droop, nor does the profile represent NACA 63A014.63. I used three dimensions to do another check on the photo, and the agreement is reasonable but not perfect.
NACA 0009.932 is another mysterious profile definition - what does the '.932' mean? A 9.932% profile thickness is possible but unlikely, mainly because three digits behind the decimal point is definitely not standard.
Next I found 'extensions' for NACA profile codes in 'Aerofoil Sections - Results From Wind-Tunnel Investigations - Theoretical Foundations' by Riegels (1961). Riegel explains on page 7 their use to describe variations of the nose radius, using '1.1', '0.825', '0.55', '0.275' extensions. But there is no mention of a '.932' extension.
Time for a very wild guess. The table with coordinates in Abbott & Von Doenhoff page 314 has an open space for the 0.5% chord coordinate. Maybe .932 would fit? It gives a slight V-shape to the extreme leading edge, something that I see in the photo too.
In the end I decided to list the NACA 0009 data, plus the extrapolated 9.932% thick version.
| NACA 0009 | NACA 0009 9.932% thickness | 
| |
| 0 | 0.000 | 0.000 | |
| 0.5 | (0.932) | (1.029) | |
| 1.25 | 1.420 | 1.567 | |
| 2.5 | 1.961 | 2.164 | |
| 5.0 | 2.666 | 2.942 | |
| 7.5 | 3.150 | 3.476 | |
| 10 | 3.512 | 3.876 | |
| 15 | 4.009 | 4.424 | |
| 20 | 4.303 | 4.749 | |
| 25 | 4.456 | 4.917 | |
| 30 | 4.501 | 4.967 | |
| 40 | 4.352 | 4.803 | |
| 50 | 3.971 | 4.382 | |
| 60 | 3.423 | 3.777 | |
| 70 | 2.748 | 3.033 | |
| 80 | 1.967 | 2.171 | |
| 90 | 1.086 | 1.198 | |
| 95 | 0.605 | 0.668 | |
| 100 | 0.095 | 0.105 |
The nose section had two variables: what chord definition was used, and what thickness was used. I made the following drawing to test all four options, ready for test-fitting on the photo.
The smallest option was a bad fit.
Skipping options 2 and 3, it was only the largest option that had a good fit. However, the section was rotated 7.5 degrees instead of the expected 5.85 degrees. However, in the following section, the whole profile is rotated 1.85 degrees nose-up to achieve a 'horizontal' chord line. If we deduct 1.85 from 7.5 we get 5.65 degrees, which is close to the expected 5.85 degrees, so probably that's the explanation.
A close-up view of the fit. If my fit is accurate (quite unlikely), I would expect a kink in the sheet metal of the nose extension on the lower side, so it can connect properly with the 'secondary' flange of the front spar. The 'primary' flange would be the one that connects with the thick center wing skins (see also the picture from the manual, six pictures up). However, photos do not show the kink in the sheet metal.
If the kink is to be avoided, a part of the NACA 0009.932 nose extension must be modified. This is drawn in green, and it fits well with the structural details of the wing. I think this is the best fit that can be achieved within the limitations of this analysis.
The result is the following profile, in two variations. The top one retains the chord line of the original 63A014.63, and I think this is how the real wing relates to the Firebee's reference axis system. The bottom one has a new chord line that has a horizontal chord line between the leading and trailing edges; it is rotated 1.85 degrees nose up. The top one has a thickness of 12.71%, the bottom one is 12.86%. One could also draw the 'streamwise' profiles from them, by reducing the thickness to 70.71%.
A number of assumptions were made in the above analysis. The profile codes were strange but fitted well.
Now that a reasonable approximation of the BQM-34A profile is established, it is also possible to approach the reconstruction from the other side. I plotted the 0009.932 profile rotated 5.85 degrees down around its leading edge (per Ryan data, in red), then added 63A014.63 with its original chord length (blue), and rotated it up around the trailing edge until it was tangent to 0009.932. The required angle was 1.85 degrees, and the tangent point is not far off the 26.4% chord point. I plotted both over the reconstructed BQM-34A wing profile (in black), and the fit is excellent, the black line almost disappears. In this reconstruction, all known Ryan data fits.
About half a year later, I found this presentation slide by Ryan from probably ~1959, showing the profiles side by side, but with some viewing angle. Since we see the chords at a slight angle, therefore they will appear shorter. In other words the profiles will appear to be slightly fatter. I copied the profiles as reconstructed above, and found that they fitted well, but were slightly thinner, as expected. The check is not scientific, but the results are satisfying.
Source: I forgot to note the source, but I think I found it in the SDASM Archives on Flickr. However that photo archive is so enormous that I cannot find it again.
