Re: "How Things Work" Website

[William Beaty <billb@eskimo.com>]

On Sat, 29 Mar 1997, John Mallinckrodt wrote:

> I do not happen to be a particular fan of the Bernoulli-based explanation
> for lift and much prefer transfer of momentum arguments.  On the other
> hand, I thought the explanation provided at the "How Things Work" website
> was rather good.

I thought that most of the other material in the website is very good.  It
could be improved by adding a webpage index, since the file was up near
300K.  I'd have liked to browse the file, rather than having to page down
through the whole thing. 

> I'm curious about why you call the "airstreams must join at the trailing
> edge" argument "false."  In the steady state, laminar flow regime in which
> the Bernoulli principle holds (and which is, admittedly, questionable for
> airplane wings) mustn't this be the case?

Nope.  I'm not sure, but I suspect that in some 2D simplifications, the
parcels of air may indeed be required to rejoin at the trailing edge.  If
so, then this is an artifact of the simplifying assumptions, not a feature
of real airfoils. In real 3D wings there is a sort of singularity trailing
behind the wing, called a "vortex sheet"  by aerodynamics textbooks.  When
generating lift, the parcel underneath the airfoil is slowed and the
parcel above speeds up. The parcels which were once adjacent are
permanently separated from each other.  As a consequence, an overall
vortex motion appears with the vortex core centered near the tip of the
wing.  The vortex core acts as a boundary for the edge of the vortex
sheet. 

While this does screw up the elementary-school aerodynamics explanation,
it does clear up confusion about thin airfoils.  In a thin airfoil the
path lengths are equal, so how can the "Bernoulli Effect" apply?  When an
uncambered thin airfoil (e.g. sheet of plywood) generates lift, the air
above is sped up and the air below slows down, and the parcels do not
recombine at the trailing edge, nor do they need to.  There is a relative
difference in velocity, and so a differential pressure pointing upwards. 
"Bernoulli" therefor applies to tilted sheets of plywood, not just to
cambered thick wings having unequal path lengths, and those who defend the
"Bernoulli" explanation can breath a big sigh of relief.  ...and then
search for a better explanation for why the air above the airfoil *really*
moves faster.

Jan-Olov Newborg in Sweden, an associate of Dr. K. Weltner of the infamous
"anit-bernoulli" papers in AJP and TPT, sent me a couple of GIFs which I
should strip down and install on my webpage.  They depict the path of a
sheet of smoke, a "smoke pulse", as a wing flys perpendiculary through it
in a wind tunnel.  The GIF is pretty small, I've attached it to this
message as a MIME file.  It's also at: 

  http://www.eskimo.com/~billb/wing/airgif.html
  http://www.eskimo.com/~billb/wing/airfoil.html

He mentions that this "phase shift" effect is proportional to lift.  If
any airfoil attack angle is adjusted in order to eliminate this phase
shift, lift is eliminated as well.

So, since the parcels of air do not rejoin, and in fact it seems that this
"phase shift" is a necessary component of the lift-generating mechanism,
it is very wrong to base the "Bernoulli" airfoil explanation on a
requirement that the parcels rejoin at the trailing edge.  I doubt that
this fact will do little to counter the spread of this misconception.
After all, it's in all the textbooks, and how can that many textbooks be
wrong?    ;)

......................uuuu / oo \ uuuu........,.............................
William Beaty  voice:206-781-3320   bbs:206-789-0775    cserv:71241,3623
EE/Programmer/Science exhibit designer        http://www.eskimo.com/~billb/
Seattle, WA 98117  billb@eskimo.com           SCIENCE HOBBYIST web page