In-flight science

We could all follow Daedelus and soar like birds - if we get the science right, says Battle of the Geeks' Ian Johnston

By: Dr Ian Johnston (Department of Engineering and Innovation)

  • Duration 10 mins
  • Updated Tuesday 26th September 2006
  • Introductory level
  • Posted under Engineering
Share on Google Plus Share on LinkedIn Share on Reddit View article Comments

Flying is easy.

Well, generating lift is easy – any fairly flattish thing will work as a basic wing, as anyone who has tried to carry a sheet of plywood or a paddling pool on a windy day will know. And as long as the object isn't too heavy, getting off the ground is pretty easy too.

So what is it that makes flying so difficult? The problem is not creating lift, but controlling and using it.

A sheet of plywood disappearing downwind is undoubtedly flying, but it is not a practical basis for transport or communication. Nature, of course, makes it look ridiculously simple and has evolved a huge range of flying machines, from the straightforward low-performance glider (the flying squirrel) to the ultra-long endurance sailplane (albatross or condor) or the highly manoeuverable fighter (hawk, eagle, or bat).

Birds can fly partly because of their shape, partly because of their beautifully lightweight construction, partly because they have power available and most of all because of the amazing control they have over the flying surfaces – wings and tail.

For almost all types of flight and types of flying machine, control is the key. Controllability is almost always linked the idea of stability, which in engineering is defined thus: "A stable system is one which will tend to return its previous state if given a slight disturbance."

Imagine a ball bearing. First, picture it sitting in a soup bowl. If it is pushed slightly to one side – by a finger, a spoon or perhaps even a small current of air – the ball bearing will tend to return to where it started as soon as the disturbing force is removed. This is called stable equilibrium (the "equilibrium" part means that the forces are in balance).

If it is sitting on a flat surface, it will stay where it is moved to – this is neutral equilibrium.

And if it is sitting on top of a curved surface, a small disturbance will become larger and larger – this is unstable equilibrium.

Almost all wings are naturally unstable. As they move through the air they generate lift – a force at right angles to the oncoming airflow. The lift can be thought of as acting at a point, which is generally about one-third of the way back from the leading edge, and for stable flight this will balance the downwards force of gravity on the plane: with the lift force aligned with the gravity force the plane is in equilibrium.

However, if the angle of attack of the wing increases slightly, the lift force not only changes in size (it normally increases slightly) but also moves forward.

The combination of lift and gravity now produces a twisting torque on the wing, and since the upwards force is in front of the downwards one, this torque tends to increase the angle of attack even more.

Which means a further movement of the lift force, even more torque, more rotation and so on.

The wing is unstable: the slightest increase in angle of attack when in equilibrium will inevitably lead to an upwards zoom and stall, and the slightest reduction will result in a vertical dive.

The tail is the key to resolving this problem. In normal flight the tail is generally aligned with the airflow (zero angle of attack) and producing no lift. If the aircraft pitches up slightly, the tail, which is really a small wing, starts to produce some lift and tries to force the nose of the plane down. If it is mounted far enough behind the centre of gravity, this overcomes the tendency of the wing to pitch up and instead returns it to the original position.

Similarly a nose down pitch of the aircraft gives rise to a downwards force at the tail which in turn brings the nose of the aircraft up again. The aircraft is now stable.

It's a common misconception that to cause a dive, the tail of a plane produces an upward force to push the nose down. In fact, a dive is produced by changing the shape of the tail (using the elevator) and reducing the amount of lift it produces for a given angle of attack.

This changes the equilibrium state of the aircraft to a nose down one, since a bigger angle of attack is needed at the tail to balance the offset wing force.

Most people are surprised to learn (I certainly was!) that the tail produces a downwards force during a dive, and the steeper the dive the greater this force has to be.

Most pioneers of aviation (Jean-Marie Le Bris, Otto Lilienthal, Percy Pilcher) – wanted to produce flying machines which would be stable in the air. Although this sounds logical, stability is not nearly as good an idea as it might sound.

A car with jammed steering is completely stable, but quite uncontrollable, and for flying controllability is the key. By building stable aircraft, the pioneers made machines which would crash in a variety of beautifully stable ways. Most of them paid with their lives for this mistake.

As we saw above, controlling a wing is essentially a matter of using the tail to harness the essential instability of the lift force. Trying to make a stable wing which can be forced into a dive is not nearly as effective as simply restraining one which will, given half a chance, dive by itself.

To make an aircraft controllable, therefore, it has to have an element of instability.

It was the Wright brothers - Wilbur and Orville - who first came to this conclusion, after many hours of experiment with tethered and free flying gliders.

Their flier was light, adequately powered and,above all, the first properly controllable aeroplane ever built. As a result, if the nose started rising - pitching - or a wing started going down - rolling - in flight, the pilot was able to do something about the situation instead of simply hanging on and hoping to emerge relatively unscathed from the wreckage.

Today the use of instability to achieve controllability has moved even further with the advent of computerised cockpit systems. Many modern fighters are quite unstable in flight and rely on constant automatic corrections, made far more quickly than a human pilot could ever manage.

When a sudden manoeuvre is requested by the pilot the computer simply indulges the aircraft's constant desire to perform spectacular aerobatics.

Of all the requirements which have to be considered when designing an aircraft, controllability is the most important, the least obvious – and by far the most difficult.

More like this