John Dickenson

The inventor of the Modern Hang Glider in 1963




Author: John W. Dickenson

14th Feb 2011



David Cook, designer of the remarkable ultralight sports aircraft, the “SHADOW”, asked me to explain the Shadow’s unusual low-speed flight characteristics at high angles of attack that would cause other aircraft to stall and/or spin.  At what could be regarded as ‘stall speed’, the Shadow is able to maintain a controllable steady-state ‘mush’ and shows no tendency to drop a wing or spin.  On hearing of the Shadow’s low-speed flight credentials, I just had to know what airfoil was responsible for the Shadow’s most desirable low-speed performance.  Further, I was most surprised to discover that David Cook had designed the airfoil himself.  Airfoils have been a life-long interest for me and I was determined to find out more about this mysterious airfoil.  What I learned from this airfoil has challenged all that I know about airfoils and caused me to doubt popular airfoil theory.  As I searched more into the matter, I found I was not alone with my doubts.  NASA have released a series of papers also challenging the accepted view on airfoils.  Follow their online “Theories of Lift” Guided Tour that examines three popular, but incorrect, theories.

It turned out that David Cook’s airfoil (I shall name it the “DC ONE”) is a popular helicopter rotor blade airfoil for the first 33% of its chord (i.e. front third of wing from leading edge back towards the trailing edge).  I regard this as an astounding choice for a low-speed aircraft, as the tips of helicopter blades can exceed the speed of sound.  The next shock was that the upper and lower surfaces proceed as a straight line to the trailing edge, giving the airfoil an almost symmetrical cross section.  The DC ONE is ‘thick’ as airfoils go, so the straight line from the upper curve of the airfoil to the trailing edge is ‘steep’ - this is the ‘master clue’ to its performance and to how airfoils really work.



There are a number of possible reasons for the Shadow’s ability to hold a high angle-of-attack and still remain stable.  It could be the result of the design and layout of the wing and elevator system.  The first possibility is that the tailplane and elevator do not have enough power to pull the aircraft into a fully-developed stall.  The relationship between the wing and the distance of the tailplane from the centre of effort (CE) and or (CG) may contribute to this.

The angle formed by the wing in relation to the setting of the tailplane may also be a factor.  Normal practice is to set the tailplane to provide the best angle-of-attack for the wing, typically 8 degrees, giving the highest L/D ratio at cruise.  With the Shadow airfoil, the relationship may be different (the angle of ‘no lift’ is not known).  As a consequence, the angle-of-attack at cruise may be less than normal and the wing may not be able to reach a high enough angle-of-attack to establish a full stall. Typical of all well designed modern aircraft the wing of the SHADOW is washed  out at the tips, this can delay stall, and to prevent a wing dropping at stall  and initiating a spin, but washout is not the reason for the Shadows non stall behaviour. Certainly the Shadow’s fin area assists to keep directional control in the ‘mush’ condition.  If the layout of the wing and tailplane is exactly the same as the Volmer Jensen VJ-23, it could explain some of the aircraft's characteristics.  This would relate entirely to wing area, aspect ratio, the distance of the tailplane from the wing’s centre of pressure.  I can only imagine that Jensen would have striven for a low speed, soft stall such as the Shadow has inherited.  For a foot-launched aircraft to enter a vicious stall 30 or 40 feet off the ground would have to be the worst possible situation for the pilot. 

The only other major difference between the VJ-23 and the Shadow is the airfoil used.

When I heard that the flight characteristics of the Shadow at stall are those desired by every designer of low-speed aircraft, my attention was turned to the airfoil David Cook had developed for his aircraft.  I undertook an intensive review of this unusual airfoil and the way it worked.  The result of that review has caused me to change my view of the way airfoils work, and has lead me to a belief that conventional wisdom about airfoil performance, based on wind tunnel data, is not providing a theory that fits with practical application.  I searched energetically through reams of data about airfoils, testing in wind tunnels and the relationship between Bernoulli’s theories (as applied to the upper surface of wings) and Isaac Newton’s theories about active forces (as applied to an inclined plane moving against the air).  I have decided that for airfoils, Mr Newton’s theories are more relevant than those of Mr Bernoulli’s.

The difficulty, as I see it, is that when an airfoil is tested in a wind tunnel (connected from one side of the tunnel to the other), the upper surface forms a type of ‘venturi’.  The moving air is being accelerated between the wing upper surface and the roof of the tunnel.  When a positive angle-of-attack is applied, the air passing under the wing is forced into a narrowing situation at the trailing edge of the wing, increasing velocity and decreasing pressure, where the opposite should be happening.    I reason that this pollutes the data obtained in a wind tunnel. It is not the same as if the airfoil were placed in free air.

There is one other important factor - the air flows around the wing in a wind tunnel, so the air in fact has energy and the airfoil is inert.  In normal flight, the wing of an aircraft moves through the air and has energy, while the air is inert.  The air does not flow around or across the wing - it is merely parted in the same way as a boat hull parts the water through which the boat moves.  If you were to place two floating objects in front of a moving boat, so that the boat passed between them, they would appear to move along each side of the boat (as viewed from the boat), when in fact they are perfectly still in the water.  So it is with a wing as it parts the air. 

It is easy to see how the lower surface of the wing produces lift, by comparing it to a water ski.  When the water ski planes across the water at a positive angle-of-attack it can support its skier.  When the lower surface of the wing proceeds through the air at a positive angle-of-attack it is ‘planing’ on the air.  Clearly the re-direction of forces acts to support both the skier and the aircraft, all according to Mr Newton. 

I believe the upper surface of the wing acts in two ways. First, at the same distance from the leading edge of the wing, the air is forced upwards a further distance than the air moves downwards on the underside.  Because the distance along the upper surface is longer than the distance along the lower surface, the air on the upper surface is ‘stretched’, particularly in the first 30% of the wing’s upper surface.  This is the reason for reduced pressure on that part of the upper surface.  The air is not accelerated over the wing, as occurs in a wind tunnel. Second, I also consider that the trailing 70% of the upper surface has more contribution to lift than is conventionally believed, as the air moves downwards on the rear upper side of the wing in accordance with Coanda effect.  This downward-moving air from the upper surface air joins the air that has been deflected downwards underneath the moving wing, and thus adds to the lift generated by the lower surface.  In the case of the DC ONE airfoil, the very steep descent of air down the upper surface of the airfoil to its trailing edge enhances the lift on the underside of the wing to a greater extent than is usual.  This may be the reason for its superior performance.

Is there any evidence that the underside of the wing generates more lift than the upper side?   Well, yes there is!  One example is that most aircraft can fly happily upside down with the curved side of the wing down and the flat side up.  Another is the ‘model gliders’ we used to make as children.  Their wings were made of 1/16 inch thick sheet balsa.  The flat wing models flew just as well as those that we fitted with ‘curved’ airfoils.     

During the Second World War, the first German jet fighters could not get off the ground with the wing parallel to the ground.  It did not matter how fast they roared down the runway, the curved upper surface of the wing was just not going to suck them up into the air.  It took some fancy and dangerous tricks to get them airborne.  To obtain lift-off, all aircraft must ‘rotate’ to get a positive angle-of-attack on the underside of the wing.

For a good diagrammatic demonstration, please view the “Incorrect Theory #1” page on NASA’s website.  On this page are two diagrams.  The top diagram looks like an airfoil in a wind tunnel, showing the upper and lower ‘streamlines’.  Where the two streamlines leave the trailing edge of the airfoil, they flatten out to horizontal, which would happen in a wind tunnel.  The animated diagram near the middle of the web page shows an airfoil (which incidentally looks very like the DC ONE airfoil) with air flowing around it.  Use your imagination to see it as an airfoil slicing through the air.  Scroll through the instructions in the grey text box to observe the effects of adjusting the airfoil’s angle-of-attack.  It demonstrates effectively the downward defection of air, above and below the airfoil, after the air has passed the trailing edge.  This reinforces my contention that the downward pressure of air leaving the upper surface of the air foil contributes to the lift efficiency of the lower surface.

John W. Dickenson OAM. BA psych. FAI DIP.
The Inventor of the Modern Hang Glider in 1963
© Copyright 2011


Wake turbulence forms behind and below the Boeing 747 as it passes through the clouds when descending for landing. London Heathrow Airport

Photo source

Talk about things you cant duplicate in a wind tunnel! The above photo is of a large jet decending, which the visable wake turbulence is BELOW  the decending air craft. Fancy that! This demonstrates dramatically the huge amount of air driven downwards, sort of turbo charged by the air decending down the upper side of the wing, note how far below the wing it progresses, the mass of the air displaced must match the mass of the aircraft. Sure there must be some lift generated over the the top of the wing, but it doesn't look to be much according to this picture.  

John Dickenson 2011