Now here's an idea almost as good as underground airports*:



http://www.newscientist.com/article/...-airports.html
A CIRCULAR runway for airports is being considered by the US navy, and
the idea was recently tested in principle at General Motors' proving
ground. It is thought to promise some advantages but to be more
expensive than providing similar operational capacity in the normal
way. One particular advantage is that it would save one-third of the
space occupied by a conventional airport of equal capacity.

The idea is to match the circumference to the landing speeds of the
aircraft that would use it, assuming that each aeroplane would need to
be sure of one sixth of the circumference. This would mean that for
big jet aircraft, a circular runway of rather more than 60,000 feet in
circumference would be required.

Taking the usual first-class runway width of 300 feet, an aeroplane,
touching down on the outside edge and aiming just to miss the inner
edge, would have a run of 4860 feet before it approached the outer
edge again. In that distance, thrust reversers and brakes would have
been applied and the speed reduced perhaps sufficiently to use the
nose wheel gently to steer the craft back onto another tangential
course for the next mile of its run.

As an alternative, the runway could be made wider. In that event cost
would rise steeply. A 10,000-foot runway 300 feet wide costs little
under £1 million. The need to devise new landing techniques and to
retrain pilots is acknowledged in official references to this study,
but emphasis is laid on the "unlimited runway" and on the
"minimisation of crosswind factor" by enabling take-offs and landings
to be made in any direction.

This might reduce the number of aircraft that could use the runway at
the same time. Only when crosswind was not of serious strength could
the runway be used by six aircraft simultaneously and the claim that
it would conduce to high traffic density be justified.




*

http://groups.google.com/group/rec.a...a?dmode=source