PART ONE
WING DESIGN By Ray Borst, EAA 1526 ne often hears it said, "Use any airfoil—it doesn't make any difference." In most cases—any airfoil will fly. But the selection of an improper airfoil will decrease the flying qualities of any airplane. An improper airfoil may make a marginal plane out of what would otherwise be a good ship or it will make a mediocre ship out of what would otherwise be an excellent plane. So
O
take heed—especially those who want something "just a little bit better." Of course, wing design is not just airfoil selection. There are several other factors to be considered. They, too, affect the flying qualities of a plane so that an improper choice of these other factors can also drastically affect the plane's flying properties. This paper is intended to help the amateur builder choose good values for all the factors that affect wing design and thus affect the plane's performance. First let us discuss a few things in general concerning airfoils. Excluding the special airfoils, there are only three things which have much effect on an airfoil's characteristics. 1. Thickness 2. Camber 3. Roughness Thickness is primarily thought of as a structural requirement. To be sure, you can't use a thinner wing than the maximum depth of your spar. However, thickness also affects the aerodynamic qualities of a wing — sometimes quite drastically.
Fig. 1 Wing Thickness in percent
Figure 1 shows how the maximum lift coefficient changes as the thickness is varied. Note that the highest
lift is obtained with a 12 percent to 15 percent airfoil. Note, also, how sharply the lift coefficient drops for thicknesses less than 12 percent. Not only does the maximum lift coefficient decrease but these thinner airfoils
usually develop a very sharp stall. Thus, when some of our "hot pilots" decide to use a thin airfoil to decrease their drag, they have two strikes against them— 1) decreased maximum lift 2) sharp stall and to top it off, they don't get much decrease in drag anyway. Note on Figure 1 that reducing the thickness from 12 percent to 9 percent causes a 22 percent loss of lift while decreasing the drag just 5.8 percent. This means that a 12 percent wing could have 22 percent less area and give the same lift and stalling speed as the full size 9 percent wing. Using the 12 percent wing allows a decrease of 22—5.8=16.2 percent of the wing drag. Thus, the amateur builder that uses a thin airfoil is actually penalizing his plane. This loss is even worse for a 6 percent airfoil as is also shown in Figure 1. There are a couple of "hot planes" around—they call them "widow makers" back home—I won't mention any names but the airfoils on those planes wouldn't make a good tail fin on a car, let alone a wing on a plane. I know of no reason for using an airfoil with a thickness less than 12 percent. If you really want to decrease the drag, use a laminar flow 65 series airfoil. But never, never use an airfoil less than 12 percent thick. If you decide to use a cantilever wing, a 15 percent thickness would probably be your best bet and it might even pay to use an 18 percent thickness at the root. Remember drag is primarily caused by the wing surface not by the thickness. A smooth 18 percent wing has much less drag than a dirty 12 or—heaven forbid—9 percent wing. And a smooth 18 percent wing is not any harder to build than a smooth 9 percent one. The next thing which affects an airfoil's aerodynamic characteristics is camber. Figure 2 shows a typical airfoil. If we draw a line half way between the top surface and the bottom surface we get an average line—a median line. Note that this median line curves above a straight line drawn from the leading edge to the trailing edge. The height of this curved median line above the straight line is called camber. Figure 2 shows how the drag coefficient of a symmetric airfoil increases as the lift coefficient increases (either positively or negatively). Now, if we add camber to this airfoil—that is, if we curve its median or average line, we shift this curve to the right. Then, by giving the airfoil the proper amount of camber we can position the minimum drag coefficient at any value of lift coefficient. Of course, the obvious advantage is that we can obtain the minimum drag at the lift coefficient where we do most of our flying-that is, at our cruising lift coefficient. All that glitters is not gold, however. Increasing the camber does not come for free. Increasing the camber also increases the twisting moment that the wing develops. That is, increasing the camber causes the center of pressure to move further aft. Moving the center of pressure further aft moves the lift further away from the main spar and tries to rotate the wing tip about the SPORT
AVIATION
17
-.4-
-.2
«
.2
.4
Fig. 2 Lift Coefficient
spar. This is similar to rotating your wrist when your arm is extended. If your wing has two struts you probably won't have to worry. However, a large twisting moment could cause trouble with a cantilever or single strut wing.
The second curve is for a smooth airfoil. The minimum drag coefficient for a smooth airfoil is about 30 percent less than for a standard airfoil. I guess this bit of information will get some people around to cleaning their wings before they start home. In addition to regular airfoils there are two groups —families—of what I term "special airfoils". Of course, the first of these special airfoils are the laminar flow airfoils. Figure 3 also shows the "bucket" common to laminar flow airfoils. A good laminar flow wing will have about half the drag of standard wings. Thus, you boys who are interested in squeezing that last mph out of the old buggy—take heed—a true laminar flow wing will do wonders to your cross country cruise. Make sure, however, that you select the proper camber to give the necessary lift coefficient at your cruising speed or your efforts will be wasted. The second special family of airfoils is the 230 xx series. This is probably the first time that you have ever heard this family called "special". I have some other names for this family but I can't use them here. Let me say that this family was developed for just one thing—to develop a very low twisting moment. It allows designers to build metal skinned cantilever wings. This airfoil allowed Donald Douglas to build the DC-3. However, while it does have a low twisting moment, it has a very poor stall. Now if you are going to build a stressed skin, cantilever wing, then go ahead and use a 23012. But, if you are going to build a two strut braced wing, you shouldn't even consider it. The only planes that might have to use the 23012 are the cantilever French jobs because all of their wing twist is taken by the spar. But, please, there are so many good airfoils around, don't waste your efforts on this one. For our analysis, we will first specify four things about your "dream plane". These are 1. 2. 3. 4.
Weight . . . . . . . . . . . . . . . . . . . . . . 1200 pounds Cruising speed . . . . . . . . . . . . . . . . . . . . 140 mph Landing speed . . . . . . . . . . . . . . . . . . . . 5 0 mph Type of wing — take your pick a) cantilever i) all metal-stressed skin b) single strut ii) box spar c) dual strut iii) 2 spar & cable bracing
Once we have decided upon these four specifications that our plane is to meet we can proceed with the design of our wing.
L
I
17
* a» O
.2
f
.6
Fig. 3 Lift Coefficient
Third on our list was roughness. Figure 3 shows
how the drag coefficient increases for a rough surface — what the NACA calls standard roughness. This is equivalent to about 1/32 inch grains of sand distributed on the leading edge. Look at the bugs splattered on the leading edges of most of the ships out there now and you will see that this standard roughness is just about standard. 18
JULY
1961
j£
3?
«r
in
a
*
too
0)
tooo
Fig. 4 Weight pounds
Step One
On Figure 4, go along horizontal weight scale and find the weight that you have selected. Next find the line which represents the landing speed that you want. Now go vertically from the weight to this landing speed line — circle the point where the vertical from the weight and landing speed line intersect. Now go horizontally from this circle to the vertical (wing area) scale. The point where this horizontal line crosses the wing area scale gives the required wing area.
c
« «
o
Example
1200* plane and 50 mph Follow dotted line from 1200* on horizontal axis
up to the 50 mph line—circle this point. Now go horizontally from this circle to the wing area scale and a mark the place where you cross the vertical scale. Thus we need a wing area of 137 ft 2 .
Fig. 6 Wing Loading Lb./Sq. Ft.
NOTE: Lift coefficient may be thought of as how hard the wing is working. Thus wing with a high lift coefficient is working harder — producing
more lift at that speed—than a wing with a , smaller lift coefficient. Example
10
o
Wing Ioading = 8.9*/ft2 & VCR=140 mph Follow the dotted line vertically from 8.9 on the horizontal scale to the 140 mph line and circle this intersection. Now move horizontally from this circle to the vertical scale and mark this intersection. Thus this plane will have a lift coefficient of .18 at cruise.
en
c
Step Four
Fig. 5 Weight pounds
Step Two
On Figure 5, go along the weight axis to the weight that you specified. Next find the line representing the wing area we found on chart number 1. NOTE: If the wing area is not exactly the value of
one of the lines, sketch a line representing your wing area between the two closest lines.
Now, go vertically from your weight to the line representing your wing area and circle this point. Once again, go horizontally from this circle to the vertical scale and mark this intersection. This value is the wing loading— the weight supported by each square foot of the wing. Example
1200* plane and 137 ft 2 Follow the dotted line from 1200* on the horizontal scale vertically to the 137 ft 2 line and circle this intersection. Next go horizontally from this circle to the vertical scale and mark the intersection. Thus the wing loading is 8.9 pounds/ft 2 . Step Three
On Figure 6, locate on the horizontal scale the wing loading value that you have just determined. Next, find the line representing the cruising speed that you selected. Now go vertically from your wing loading to the cruising speed line and circle this point. Once again go horizontally from the circle to the vertical scale and mark this intersection. This value is the lift coefficient at cruising speed.
Still using Figure 6, find the line representing your climbing speed. This will be from 1.3 times the landing speed for a 65 Baby Ace to 2 times the landing speed for a big Tailwind. Probably 1.4 to 1.7 times the landing speed are good values. Now extend the vertical line from your wing loading up to this climbing speed line and circle the intersection. Once again, move horizontally to the vertical scale and mark this intersection. This value is the lift coefficient during climb. Example
Wing loading 8.9 pounds per square foot & Vc=1.7 x 50 = 85 mph
Extend the vertical line from the 8.9 pounds per
square foot to the 85 mph line and circle the intersection. Once again, move horizontally to the vertical scale and mark this intersection. Thus the lift coefficient during climb is .48. Note that several airfoils are listed in the blocks
at the right side of Figure 6. They are positioned along the vertical scale approximately where their minimum drag coefficient occurs. Also, each of these blocks has Bn arrow extending above it. These arrows represent the range where the drag is low enough to permit a good climb. Thus to select an airfoil, one picks a block that is bisected by the design cruising lift coefficient and which has an arrow extending up past the climbing lift coefficient. Thus, where we previously determined that we had a cruising lift coefficient of .18 we find that a 2412, a 63-212 or a 63-412 would be good airfoils. Checking these airfoils for climbing lift coefficient of .48 shows that the 2412 and the 63-212, still have low drag coefficients at this value of lift coefficient. Thus, eitheit one of these airfoils could be used. To be continued next month. SPORT AVIATION.
19
PART TWO
WING DESIGN By Ray Borst, EAA 1526 There is another important factor that we must yet consider. This is, what wing span shall we use? The aeronautical engineer uses the term Aspect Ratio. Aspect Ratio is equal to the wing span divided by the wing chord. Since the wing span multiplied by the chord equals the wing area and we have already found our wing area, we don't have much trouble determining what span to use.
Figure 7 shows the relationship between the total drag coefficient—surface drag plus a drag that is caused by the lift and is called induced drag—and the lift co-
To determine the wing span on Figure 8 find the value of wing area that you determined from Figure 4. Then find the line representing the aspect ratio which will give a reasonable value of induced drag as shown on Figure 7. Now move vertically from this wing area to the proper aspect ratio line and circle the intersection. Next draw a horizontal line to the vertical scale of the chart and mark the intersection of this horizontal line on the scale. Now on the left scale read the value of wing span.
-r 3o
____/*:
5 zc a
v>
-4—y-
200
Fig. 8 Wing Area Sq. Ft.
Example
Wing area =137 ft2 Aspect Ratio=6 Move vertically from 137 on the bottom scale to the line representing an aspect ratio of 6 and circle the intersection. Next draw a horizontal line from this point to the vertical scale and read a value of 28.5 feet. Fig. 7
efficient for various values of aspect ratio. Note that the drag increases for a given aspect ratio as the lif' coefficient increases. Also, the drag coefficient increases
as the aspect ratio decreases.
From this graph you can easily see that for low values of lift coefficient the aspect ratio doesn't make a whole lot of difference. Then you expect to find a rather lo^ aspect ratio on fast planes. This is usually true and values down near 5 are common. A slow plane will fly at a higher lift coefficient and thus, to keep the induced drag down to a reasonable value, slow planes will have higher values. This is usually true and values of over 7 are common. Indeed, for thd ultimate in slow planes—gliders—values of 15 are common. 10
AUGUST-1961
To determine the wing chord, on Figure 9 find the value of wing area that you determined on Figure 5. Then find the line representing the aspect ratio that you have selected. Now, move vertically from this value of wing area to the proper aspect ratio line and circle the intersection. Next draw a horizontal line to the vertical scale of the chart and mark the intersection of this horizontal line on the scale and read the value of wing chord. Example
Wing area=137 ft2
Aspect Ratio=6
Move vertically from 137 on the bottom scale to the line representing an aspect ratio of 6 and circle the intersection. Next draw a horizontal line from this point to the vertical scale and read a value of 4.75 ft.
To check the values of the wing span and wing chord that we have obtained, we will multiply them together and they should give {be wing area.
1-l
s/ -** j>/„
k
V2
i'
^ ^^> s \ _— ^
\
\s
\
^^
—V
^ \7 /
-/ f
1 /*
i
4
6
ff
/»
/Z
/+ /L
/9 l
Fig. 9 Wing Area Sq. Ft.
Example
Fig. 11 Angle of attack
Wing span = 28.5 ft. 28.5 X 4.75=136 ft?
Chord = 4.75 ft.
This value compares quite closely with the value of 137 ft 2 that was originally obtained from Figure 5. There is but one more thing that we must determine. This is, at what angle of incidence should the wing be with the fuselage? Figure 10 shows A ^ , the angle of
I urge each of you to get a copy—at the price of $2.95 per copy you can't make a better investment. We have decided to use a 2412 or a 63-212 airfoil. Let's pick the 2412. The NACA reports give data for this airfoil. Figure 11 shows their data for the 2412. Find our design lift coefficient on the vertical scale. Now move horizontally to the average curve and circle the intersection. Next move downward vertically from this circle to the horizontal scale and mark the intersection. This value is the angle of attack that a large aspect ratio wing would have to have to develop the required lift. To correct this angle for the aspect ratio that we are using, we must add the A O^, that we found on Figure 10. Thus our total angle of attack and hence the angle be tween the fuselage horizontal (water lines) and the wing is the sum of our answers from Figure 10 and Figure 11. Example
For our example plane, find a lift coefficient of
0
.2
.4
.6
Fig. 10 Lift Coefficient
attack increase, that is necessary because of these induced effects. At low values of lift coefficient, it may
be ignored. However, for a puddle jumper, this angle can amount to an appreciable amount.
Next, on Figure 11, find the angle of attack required to obtain our lift coefficient from a wing with a large aspect ratio—this curve is the published NACA data.
Test data for most of the modern airfoils was published in NACA Report 824. This Report is out of print now. However, Dover Publications, Inc. Dept. EAA 180 Varick Street New York 14, New York
has published a paperback called Theory of Wing Sections by Abbott and von Doenhoff who wrote Report 824. This paperback is virtually a reprint of Report 824.
.18 on Figure 10. Next find the curve representing an aspect ratio of 6. Now move vertically from the .18 lift coefficient to the curve representing aspect ratio of 6 and circle this intersection. Now move horizontally from this circle to the vertical scale and mark this intersection.
The value read on this vertical scale, +.50 degrees is
the amount that we must add to allow for aspect ratio effects.
Next on Figure 11, find on the vertical scale the value .18. Now move horizontally to the curve representing our 2412 airfoil and circle this intersection. Now move vertically downward from this circle and mark the intersection with the horizontal axis. Read a value of —.45 degrees. Now adding these two values +.50 and —.45 gives +.05 degrees. Thus our wing should have an angle of attack of +.05 degrees at cruise. If the fuselage is to remain level, the wing must be at an angle with the fuselage and hence the wing will point upward .05 degrees. We will neglect this small angle. NOTE: Usually this angle is positive and the wing points upward slightly.
Another item that effects the flying properties is
dihedral. There are planes flying that have no dihedral and have supposedly good stability. It is possible that a continued on next page SPORT AVIATION
11
WING DESIGN . . -
Continued from preceding page high wing plane might have reasonable stability with no dihedral. However, I think that the plane would be more pleasant to fly with 2 degrees of dihedral. It is quite difficult to imagine that a low wing plane could have much stability without using dihedral. A low wing plane will require 3 degrees or so dihedral. The last item to be discussed is, what planform should I use? Should the wing be elliptical, tapered or rectangular? John Thorp says rectangular, and Prof. Raspet two years ago said there is no reason for a plane weighing less than 12,000 pounds for having anything other than a rectangular wing. Let me illustrate with one final graph why these authorities say this. Figure 12 shows several planforms—a rectangular and various taper ratios. I have shaded the region where the initial stalling will occur for these various types of wings. Note that the rectangular wing stalls first at the root and progresses slowly out toward the tip. Each line shows the stalled region for additional increases in the angle of attack. A wing with a slight taper ratio will also stall at the root first but the stall will progress faster toward the tip. As the taper ratio is increased, the point of initial stall will move away from the root as shown in Sketch 3. The extreme position is shown in Sketch 4 where the high taper ratio wing stalls first at the tip. This, of course, is to be avoided unless you like to spin in from one hundred feet. Use a rectangular wing — play it safe and live to enjoy your plane. In conclusion, let me say that the answers you obtain using my procedure are not exact. I think you can come within 5 percent, however. Thus I have not intended to make professional designers out of you. But, if you follow my step-by-step procedure, you will obtain somewhat better flying properties and we will have eliminated the "widow makers". Good luck with your design and let's get them flying. A GYRO-COPTER IN FINLAND . . . Continued irom page 9 Test flights we have made with this gyro-copter
seem very successful. Stability in flight is amazingly good and takeoff is stable too. Fuselage is steel tube construction, rudder and stabilizer are fabric-covered. Rotor blades are of pine, balsa and plywood. Instrument panel: altimeter, air-speed indicator, variometer and tachometer. We are very charmed to fly our gyro-copter and under construction is a new ship with cabin -- which is very comfortable here, near to the polar circle. Kalevi Kokkola lives at Pajalahdentie 21 A 15, Helsinki, Finland and Seppo Kokkola lives at Ulvilantie 3 D
32, Helsinki, Finland.
Specifications KOKKOLA Ko-03 GYRO-COPTER "UPSTART" Rotor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
diameter . . . . . . . . . . . . . . . . . . . . . . . . . . 26 ft. 6 in. type . . . . . . . . . . . . . . . . semirigid with lag hinges section . . . . . . . . . . . . . . . . . . . . . . NACA 23012 chord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 in. rpm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Weight empty . . . . . . . . . . . . . . . . . . . . . . . . . . 270 Ibs. Weight loaded . . . . . . . . . . . . . . . . . . . . . . . . . . 485 Ibs. Min. speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 mph
Cruising speed . . . . . . . . . . . . . . . . . . . . . . . . . . 53 mph Min. sinking speed . . . . . . . . . . . . . . . . . . . . 10 ft./sec. Endurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 h Engine . . . . . . . . . . . . . . . . Poincard 28 hp at 2350 rpm Control . . . . . . . . . . . . . . . . Conventional control stick 12
AUGUST 1961
OF Fig. 12
Civil Aviation Is Being Forced Into 'Pioneer Age,' Jane's Says As printed in the December 21, 1960 New York Times The challenges of cost, noise, air traffic control and airport accessibility are pressing civil aviation into a new "pioneer age," according to the editor of "Jane's All the World's Aircraft." The editor, John W. R. Taylor, writes in a preface to the 1960-61 edition of the authoritative yearbook, to be published tomorrow, that performances in speed, altitude and range "are no longer the sole yardsticks by which the worth of an aeroplane is judged." "In civil aviation," he continues, "the cost of achieving them has been so great in terms of two-mile concrete runways, the noise problem, the nightmare of air traffic control and the inaccessibility of major airports that we are witnessing the first years of a tremendously exciting new pioneer age." The volume contains pictures of a three-horsepower butterfly-like device, produced in the Soviet Union, "that would not look out of place in the 450-year-old sketchbooks of Leonardo da Vinci," and a British man-powered aircraft weighing only seventy-four pounds. "These aircraft represent a reaction against the noise and fury, high cost and remoteness of modern aviation," Mr. Taylor said. "Nor is this the aim only of back-yard, do-it-yourself aviators, for the far more sophisticated and practical Hovercraft, jet-lift and other aircraft have the same broad objective of making air travel more convenient and accessible for everyone, everywhere," he said.
