Re: Physics of Flight

[John Denker <jsd@MONMOUTH.COM>]

Please disregard previous post;  my mailer sent it before I was ready.

At 12:48 PM 8/17/99 -0400, Michael Edmiston wrote:
> I am just looking for an overall picture of why the air is deflected
> downward, and why the overall airfoil shape is important.

Cool.  Let's concentrate on that.

> I have never
> accepted the picture that the airfoil shape provides greater path
> length above than below, hence forcing the air to go faster over the
> top.  Apparently my refusal to accept that picture was correct.

Yes, good for you.

> I have
> always assumed that the goal was to direct air downward, and that the
> airfoil design allows us to do a better job of that.

Yes.

> The goal for me
> has always been to explain why the proper airfoil design does that.

Hmmmm.  Figuring out which bit of the airfoil contributes to which bit of
lift is quite tricky.

Hint:  Circulation is important.  The *trailing* edge is critical in
determining the amount of circulation, via the Kutta condition.  Remember
they control the airplane using ailerons on the trailing edge.

> Concerning boundary layers:
> Sometimes we might be saying different things because John is looking
> at the boundary layer and I am looking farther out.

Actually, I try to avoid looking at the boundary layer.  To understand the
basic lift-producing process, you can ignore the boundary layer.  You can
treat it as a thin layer of "lubricant" that makes an insignificant change
in the shape of the airfoil, while allowing the air to slide past without
sticking (i.e. without the local velocity going to zero).

> I assume it is
> very important in terms of analysis and comparisons of wings to try to
> understand as much as possible about boundary layers.

Not really, not for the basic "overall" picture you called for.

> And since
> changes in the boundary layer affect the motion of the air further out,
> we need to be quite concerned about what is happening in that boundary
> layer.  But the typical streamline photos/drawing we see are pictures
> of gross air movement, and, as John says, we cannot see what is
> happening in the boundary layer in those pictures.

The only reason I discussed the boundary layer is because you asked about
turbulence.  Why don't we just drop all discussion of turbulence and
separation?  They aren't relevant to the basic lift-producing process.

> There isn't enough air mass in the boundary layer to
> give us the needed lift.

That's for sure.

> I am talking about a grossly stalled wing.  I'm way
> past a little separation of the boundary layer to the point that
> streamlines above and behind the wing are all screwed up from the nice
> laminar-appearing streamlines we saw when the wing was working the way
> it is supposed to work.

If you want to understand the grossly stalled wing, you've got a lifelong
research project ahead of you.  There is *no* snapshot that will show you
what's going on, because the flow pattern is unsteady.  The problem is
grossly nonlinear, too.

> The reason I use a gross picture of a stall is because I assume the
> gross picture is eventually what happened when a plane crashed because
> of a stall.  If John is correct that nothing drastic happens at the
> critical angle of attack (and I assume he is correct) then I presume
> all is not lost when a plane slightly exceeds the critical angle.

Definitely all is not lost.  Stall practice is a big part of pilot training.

> If
> the pilot makes the correct maneuvers, recovery ought to be possible.

The recovery is counterintuitive if you've never thought about it and
practiced it, but routine otherwise.

> But if the pilot makes the wrong maneuvers, then the situation gets
> worse and worse and a gross stall can occur, including major separation
> with turbulence above and behind the wing (including far out from the
> boundary layer) with major loss of lift and major loss of air speed.

Actually in typical general-aviation airplanes if you don't get into a
spin, it's rather hard to get the airplane deeply stalled.  The airplane
starts a stall recovery on its own, and you don't have enough control
authority to deepen the stall.

> At that point I assume the pilot needs both talent and altitude if
> s/he is going to recover.

Typically 100 feet of altitude and ordinary amounts of skill suffice.  In
most cases 50 feet suffices if you're good.

> Because of the tendency for a
> fluid to follow the surface of an object moving through it, a wing
> traveling in horizontal flight with positive angle of attack will
> direct the air downward.

1) Actually there is no particulary fundamental reason why a fluid should
follow the surface.  Detached flow is just as plausible as attached flow,
and....

2) Even attached flow and an angle of attack is not quite sufficient. A
wing with an angle of attack moving through a *superfluid* will have nice
attached laminar flow everywhere, but no lift, because of no circulation.
(Zero viscosity means the Kutta condition isn't enforced.)

Remember the Kutta-Zhukovsky theorem:
  lift equals circulation times airspeed times span times density.
Understanding circulation is indispensible to a physical understanding of
lift.  Try the fluttering-card demonstration,
  http://www.monmouth.com/~jsd/how/htm/airfoils.html#fig_flutter

It may change your intuition about what aspects of airfoil shape are important!

Cheers --- jsd