By George G. Spratt (EAA 17426) Spratt and Company, Inc. P. 0. Box 351

Media, PA 19063

Photo 1 — Chanute double deck glider flying in the Lake Michigan sand dunes. Summer of 1896.

J.HE DECADE SURROUNDING 1900 was the most
active in flight development and experimentation of any other similar time period . . . and much of this development was done with what we would now call hang gliders. Visiting a present day ultralight meet one is quickly impressed by the similarity: airframe weights are all but identical, much of the control is by the pilot shifting his weight, airfoils are usually single surface, air speed is the same and foot launching is common. This early hang glider period was terminated by the development of the gasoline engine and cheap fuel. Then the thirst for power and speed, accelerated by the on40 MAY 1980

coming world war, obliterated nearly all that had been done during this period: the open fuselage was replaced by an enclosed fuselage, wing warping by the aileron, the single surface by thickened airfoils, weight shifting by dynamic controls and wheels replaced foot or track launching. About the only vestige still remaining is the vertical rudder used by the Wrights to overcome adverse yaw, but not even this is used universally. Almost none, if any, of the present hang glider enthusiasts were present during that time so it is understandable that much of the work is being redone and, unfortunately, many of the mishaps repeated. It would be truly worthwhile if someone would write a full history of this pioneer stage in aircraft develop-

Photo 2 — Weight double deck glider. Kitty Hawk 1901.

ment before too much of it is lost, or distorted by nonaircraft historians. It is my hope that the following brief review of but one of the control and stability problems encountered almost 80 years ago may be of help to someone now experimenting with single surface curved airfoils. Photo 1 is a Chanute glider flying in the dunes on the shore of Lake Michigan. This was the first glider built in this country to be flown consistently, without mishap, by

many people, some of them novices.
It was flown in winds from 10 to 31 miles per hour, at air speeds of 20 to 40 miles per hour. This glider weighed only 23 pounds but had a wing span of 16 feet,

chord of 4 feet 3 inches and area of 135 square feet. The
airfoils were single surface, circular arc with a camber of 1 in 10.
Of it Chanute said: "It was found steady, easy to
handle before starting, and under good control when

heavier structure required. This meant putting skids below and the pilot above the lower wing. With this greater weight, the pilot moving his body would have little effect, so dynamic control would be required. For bank and turn, twisting the wings had been used by some previous experimenters so was adopted. For pitch control a small horizontal movable vane ahead of the wings should give greater moment than the 2 inch pilot movement Chanute claimed as ample. By the fall of 1900 the Wrights had their glider ready to take to Kitty Hawk for testing. The wing area was 165 square feet, the weight 52 pounds and the wing curvature an arc of 1-22. Wilbur took it to Kitty Hawk to assemble and test — a lot of work for one man, so

Orville joined him three weeks later. Some tests as a
kite showed the lift far less than expected so it was

taken 4 miles south to the sand dunes to attempt to glide down the windward side as Chanute had done near
Lake Michigan. They tried the elevator in the front and in the back, but still could not prevent a dive or stop one once started. Results were in no way comparable to those claimed by Chanute and they told him so. Chanute had never encountered this condition, even with no control vane, so was at a loss to know what was happening. Chanute asked my father, Dr. George A. Spratt, if he knew why the action was so different and what could be done about it. Although it was to be several years before my father would discover the reversal of center of pressure travel in the circular arc airfoil, he knew from his own experience as well as that of others that there was
something erratic about this airfoil. Sometimes it would

under way — a motion of the operator's body of not

over 2 inches proving as effective as 5 or more inches in the Lilienthal machine." Not easily visible in the photo is a tail having horizontal and vertical airfoils. These were connected to the

wings by an automatic device, designed by Mr. Herring for the purpose of securing stability. Mr. A. M. Herring
was then in the employ of Mr. Chanute, having built this and several other aircraft for him.

During this time he developed and applied for a patent on this method of spring loading the tail so as to tend to keep the wings at a constant attack. There is some question about the effectiveness of this device as Chanute said later, "A few hidden defects were gradually evolved, such as lack of adjustment in the automatic device." There have also been questions about just how much of Herring's thinking, as well as mechanical skill, went into other details of this aircraft. Later, when the Wrights decided to build a glider it was logical to start with this design. Even as homebuilders today, they wanted to make some changes.
Foot launching was out as they were thinking eventually of adding a 200 pound motor plus the much

fly stably and at other times become a diving demon. On the other hand he knew that a fiat plate, while very poor in load carrying ability and therefore unsuitable for an airfoil, was stable under all conditions and docile throughout the entire flight range. He suggested that the curvature be made only at the forward portion of the airfoil then fair into a flatter after portion . . . thus
compromising some of the lifting ability of the curve for

the stability of the fiat. The Wrights decided to continue


Photo 3 — Chanute triple deck glider flying at Kitty Hawk in 1902.

their tests in 1901 so used this curvature in a much larger machine, so glides could be made in lower wind
velocities. The wing area was an unheard of 308 square

feet. (290 considering cut cuts.) The span was 22 feet with a parabolic curve of 1-12. After the Wrights took this machine to Kitty Hawk my father went down to visit them and see how it was working. He was met at the pier with the emphatic statement: "It won't work." Back at camp they found,
after some experimenting, the flimsy ribs were flexing

A flat surface at 90° angle of attack has the center of pressure at 50% chord. Then as the angle is decreased it moves steadily forward, over the entire flight range, to 0° angle of attack. This makes the airfoil stable in pitch throughout the flight range because should the airfoil tilt down, as in a dive, the center of pressure moves forward tending to return the airfoil to the original angle. Likewise an upturn moves the center of pressure aft
tending to bring it back down.

under air load into an arc, destroying the desired effect.
When the ribs were braced, as can be seen in the photograph of this machine (Photo 2), it could be controlled easily by the small forward elevator. Only now that the aircraft could be controlled and flown was it found that the roll control did not work at all as expected — but that is another story. Chanute found this airfoil so interesting that he had a t r i p l a n e b u i l t and taken to Kitty Hawk for the Wrights to fly and compare. In Photo 3 it can be seen flying from one of the dunes. The extreme forward position of the camber is apparent. My father, thinking of this as only a temporary fix and in no way a true solution to pitch stability, continued to work on the problem . . . work that led later to his discovery of the reversal of center of pressure travel on the circular arc.
42 MAY 1980

In the circular arc at 90" attack the center of pressure is at 50% chord, exactly as with the flat surface.
With decreasing angle the center of pressure continues to move forward in a manner identical to the flat surface until an angle of about 15° is reached. So far the circular arc is just as stable as the flat plate and as long as the aircraft continues to fly in this region is completely stable. But with the circular arc this changes abruptly at this angle. Now the center of pressure turns

and starts to move aft with decreasing angle of attack at
an increasing rate. In some cases it moves to or beyond the trailing edge. Thus if an aircraft flying in the stable range accidentally decreases the angle, due to a gust or

increase in speed, to less than 15° the center of pressure
moves rapidly aft causing more decrease in angle and increase in speed. In other words, an uncontrolled dive.


nil; l i l > I H ' \ \ i - | . UK T i l l , in: \ \ | i \ y l \ T l u \




3O M p H










0 02



——— Hurfe*. I'Unt————— Kuiter. Curvrd. ('Mtitwr ^

————— Surfor




Plate I — Comparison made by Eiffel of C.P. travel on several single surface circular arc airfoils to a flat surface.

Plate III — Airfoil characteristics of single surface circular arc having spars at leading and trailing edge.



Plate II — Vector diagram of single surface circular arc air-

foil having spars at leading and trailing edge.

Later, M. G. Eiffel made accurate measurements of this characteristic with his wind tunnel in Paris. The results of this work are shown in Plate 1 comparing several different cambers to a flat plate. As my father continued to work with the circular arc he realized that the center of pressure location on the surface is only part of the picture. You must think of the resultant of all forces acting on an airfoil as a vector. The center of pressure is only the point at which this vector passes through the surface. More important is the direction or slope of this vector. Obviously any aerodynamic force acting on the segment of a circle, neglecting friction, must be normal to this surface and thus pass through the center of radius. Sketch 1 is a typical single surface circular arc airfoil with the vectors corresponding to several angles of attack between 0° and 20°. Clearly, if the center of gravity of the aircraft was at the center of radius there could be no pitching moments and the aircraft would have neutral stability. No correction would be required when flying through turbulence, even with the air striking the airfoil at various angles. It now should be clear why the Chanute glider worked so well and was so stable while the Wright copy was uncontrollable. Looking at Photo 1, the pilot is well below the wings, his center of gravity nearly coincides with the center of curvature. The approximate position marked "C. G. — Chanute" in Sketch 1. Compare this to the Wright glider in Photo 2 where the pilot is actually above the lower wing, in the area marked "C. G. — Wright" in the sketch. Any moments encountered in the flight range of Chanute's machine could easily be corrected by the 2 inch body movement Chanute claimed while those in the Wright machine are more than four times as high. No surface is entirely frictionless and some structure in the form of spars is required. How badly would this distort the theoretical vector diagrams we have been using?

To find out I built an airfoil of 40 inch span and 10

inch chord with a spar at the leading and trailing edge.
Compression members were between these spars to resist tension of the single surface. This airfoil was tested in NYU wind tunnel with the results shown in Plates II and III. Vector diagram Plate II was a pleasant surprise. Structural drag did not seriously distort the focal point but tended to bring it closer to the wing. This is desirable because it is usually this distance that limits the possible aircraft size. For the ultralight builder who is not concerned with speed but enjoys flying with a minimum of sophistication and a maximum of ease and safety, the single surface wing should be considered . . . provided he thoroughly understands the unusual stability characteristics of this wing. Plate III shows that even including structure this wing compares favorably in lift to any modern day airfoil. (Data was taken at 35 miles per hour air speed.) True, the drag is higher than with many thickened airfoils but to offset this one deficiency there are many advantages. Using a spar at the leading and trailing edges, with a compression member between makes the lightest and simplest construction possible. The ribs need only be rigid enough to maintain camber. With a little mechanical ingenuity the wing could be made foldable. Plate IV, also prepared by Eiffel, shows in polar form the lift and drag of several different cambers and a flat surface. Coefficients given are metric. To convert them to our more familiar units multiply Ky by 15.95 to get GI and K x by 15.95 to get Cj. For instance, the surface having a camber of 1 in 7 develops maximum lift at about 20° angle of attack. At this angle Ky is .094. Multiplying this by 15.95 gives 1.499 C j which is high indeed for an airfoil operating at low Reynolds number.

Kx 0 !0

0 OS









iirliwr, I'liim- ^
urface. Curved. CiUlllRT .,

Plate IV — Comparison made by Eiffel of lift and drag of several single surface circular arc airfoils to a flat surface.

Sketch 1 — Typical vector diagram of single surface circular arc airfoil showing approximate C.G. locations of Chanute and Wright gliders.


Latest version of the Spratt land plane.


44 MAY 1980