RE: [Jsbsim-devel] Aitfoils at high alpha

[Bill G. <bi...@ao...>]

> Given what you know about simulation, let me state these questions to=20
> you (and anyone else that wants to answer):
>=20
> 1) What do you think is the value of providing some modeling of high=20
> alpha lift/drag characteristics - even if the data is sort of=20
> generic,=20
> providing perhaps (at best) only "plausible" dynamics?
>=20
> 2) Do you have any insight or suggestions about modeling high alpha=20
> lift/drag characteristics, given the use of the JSBSim coefficient=20
> buildup method?


I'll throw out some words on the subject, since I know a little bit =
(maybe
not as much as I think, though). This is just intended to be
thought-provoking, not the start of any flame-war.

One of the places that we'd see high AOA is in a flat spin, a maneuver
prohibited in a lot of aircraft. Normal straight and level flight, with =
the
minor variances from that state, that we generally encounter use a =
linear
set of equations, and we assume the airflow over the wings to be the =
same,
and we generally don't model the airflow over each wing. In spin =
modeling,
your modeling has to switch from linear to angular equations to fully =
and
properly describe the airflow and dynamics of the aircraft. The wings =
are
not seeing the same airflow and thus have to be modeled independently.

Spin testing is done on a new aircraft, just to see how easily it goes =
into
a spin, and how to recover from that spin. If the aircraft doesn't enter =
the
spin nicely, or can't be recovered predictably, the manufacturer might =
slap
some more surfaces on, such as dorsail fins, strakes, etc. Predicting
behavior in a spin is well beyond the scope of the JSBSim, and most of =
the
people that will use it. (I wouldn't do it myself).

Okay, what about high AOA without the rotational aspects? Well, in all =
but
the most acrobatic aircraft, you probably won't see high AOA. The
manufacturer is unlikely to provide data much past stall, and you =
probably
won't see data anywhere near 90 degrees.  Having the model handle AOAs
higher that stall is a good idea, and preventing the model from =
exploding if
a high AOA is encountered is a good idea, but maybe that protection is
merely upper limiting the AOA to 15 degrees over stall, or some other =
value
determined by the roll of the dice.

So, I guess I should ask the question, to what ends are you modeling =
high
AOA?

Bill Galbraith