Joe said:
>The current StroutPhysics code is essentially treating the whole
>airplane as a single lifting body; it sounds like to properly model
>these effects, we need to split it into components and model at least
>each wing separately.
I don't think modeling each wing will be necessary, unless you really
want to be uber- accurate in the simulation. I think the game was meant
to be easily played by anyone, so I think we can shave a few corners off.
At least, that's what I plan to do.
For example, there's something called P- Factor that affects all
propeller aircraft strongly. The slower and higher power a situation
you're in, the more you feel P- Factor. Since a propeller is nothing
more than a spinning wing, the down- coming side of the propeller (which
is the ride hand side of most American aircraft when looking out the
windshield as the pilot would do) has greater airspeed and generates more
lift. This tends to yaw the aircraft to the left and needs to be
counteracted by right rudder pressure.
You most commonly see this during take off and slow flight maneuvers. I
don't know how much fun it would be to require our players to have to
press right rudder during take off and gradually decrease it as speed
builds up, but I know how much fun it won't be trying to model it correctly...
Which is a long way of saying, I think the standard lift, drag, turning,
etc. equations will work fine. To tell the truth, I have no idea how you
would find the lifting ability of a Hellcat's fuselage -- we'd also need
to find its drag components and other such things. There are other
factors, like prop wash, we'd have to model, too; not too fun, in my opinion.
>That's easy enough, but here's another thing I don't understand... in
>what direction does the lift vector point? All references I can find
>simply say that it's perpendicular to the local velocity vector, but
>of course there are an infinite number of directions that are all
>perpendicular to that. Which one is the right one?
The lift vector always points perpendicular to the relative wind. The
relative wind is the wind you "feel" as you move through the air -- it is
directly opposite the direction of flight. So, while a breeze may be
coming at you from 9 o'clock (and displacing the aircraft off towards the
right), the relative wind is coming right straight towards you.
Lift is made up of vectors, and it is the horizontal component of lift
that turns an aircraft. So, if you weigh 1000 pounds and you're
generating 1000 pounds of lift, you maintain your altitude. If you bank
30 degrees to the left your weight doesn't change, but your lift does.
You will have 1000 pounds of weight pulling you down. You will have two
lift vectors, neither equalling 1000 pounds. I forget the exact math
involved, but you probably already know it (something with cosines, I
believe). You have a force of 1000 pounds of lift being divided into a
vertical and horizontal element. The horizontal component will apply a
force to the aircraft and turn it; the vertical component will counteract
weight and try to keep the aircraft at altitude.
However, since the lift is now less than the 1000 pounds of weight the
aircraft will descend. If you want to keep your altitude you'll have to
pull back on the stick, increasing your angle of attack, and therefore
generating more lift. However, doing this will cause you to lose
airspeed (increasing lift increases drag, which slows the aircraft), so
unless you also add power in your turn you slow down. This kind of
smooth coordination (rolling into a bank, pulling back on the stick,
applying the appropriate extra power, also applying rudder to counteract
turning forces) is the mark of a competent pilot.
>However, if my analysis is right, then the difference in lift must be
>due to the angle of attack, as Lars suggested. But why the angle of
>attack should depend on the roll is still unclear to me.
Angle of Attack rules the generation of lift. Lift is always produced
perpendicular to the relative wind. Up to now, we've been dealing with
pretty simple assumptions, but aircraft designers rarely build wings that
are parallel to the ground -- and angle is built in to them. Thus, the
leading edge of the wing would be slightly higher than the trailing edge;
this is a sort of automatic angle of attack. Also, engines are not built
aligned along the centerline of the aircraft; they may be pointing up by
a few degrees. This helps generate lift, too. It is not uncommon for
aircraft to actually have to nose- down when they speed up. The
increased power would generate a huge amount of lift, but you want to
translate that increase into speed, and so nose- down slightly to shift
the lift component forward.
It's late here, so I'm off to bed. I'll try to write something up about
roll and angle of attack and dihedral in a little bit. It's kind of
interesting how all this stuff comes together. At least, I think so. :-)
>Sheesh... it's not until you try to simulate something that you
>really find out how poorly you understand it, eh? :)
I know all about that. My college professors liberally colored my papers
with ink... :-)
Tim
_______________________________________________
Unsubscribe or switch delivery mode:
<http://www.realsoftware.com/support/listmanager/>
Search the archives of this list here:
<http://www.realsoftware.com/listarchives/lists.html>
|