(This image was made by late Aerodynamicist Alexander Lippish, The German Me-163 creator, at Collins Radio windtunnels, USA, in 1953. It was published in his book. Alexander Lippish was an expert on smoke in windtunnels and his smoketrails were world-famous.)
(This image is made by Aerodynamist Martin-Ingelman Sundberg at KTH windtunnels in 1992. Sundberg, who first saw syncronized smokepulses when visiting a windtunnel maker in USA 1962, made this smokepulse video to show how ICAO pilot education was wrong in explaining winglift with "airflow longer path over wing".)
Many authors point out that asymmetrical airfoils give positive lift even
if the angle of attack is zero. They offer this in order to prove that
"wing shape", and not "attack angle" should be the explanation of choice.
But there is a problem here. To determine if an airfoil is tilted, we
cannot rely on construction of the geometrical attack angle. Geometrical
attack angle is very sensitive to tiny bumps on the wing's leading edge,
since tiny bumps can change where we draw the main 'chord.' Yet tiny bumps
on the leading edge can have little effect on deflection of air, while the
tilting of the airfoil shown in the fourth section can have an enormous
effect upon the deflection of air and upon lifting force. SMALL FEATURES
ON THE LEADING EDGE CAN CAUSE US TO TILT THE ENTIRE WING, WHILE WE DENY
THAT WE HAVE DONE SO.
To determine the effective attack angle, we cannot trust the simple
geometrical rules. To determine whether an asymmetrical wing is REALLY
set to zero angle of attack, we instead must inspect the trailing edge of
the airfoil to see if it directs the air downwards.
(This image is a Navier-Stokes 2D airflow calculation around a SAAB
340 wing made by Aerodynamist Krister Karling,SAAB Aerospace)
Note that the asymmetrical (cambered) wing at the top of the diagram has
been adjusted to produce zero lifting force. There is no "slip" or "phase
delay" between upper and lower airflows. In the middle and bottom
diagrams, the angle of attack is progressively increased, which also
creates an increasing lifting force. The increasing angle of attack also
increases the phase delay between upper and lower air flows.
So not only is the common "wingshape / Bernoulli" explanation wrong, but
it even covers up some of the most interesting phenomena in airfoil
physics: the fact that the time delay between upper and lower airflows is
proportional to the attack angle and the lifting force!