Re: Flight Models for newbee's

Curtis L. Olson wrote:
> It's true that with all the FlightGear dynamics engines, you need to 
> have some idea or reasonable guesses as to how the aircraft flies in 
> order to model it.  With X-Plane you don't need to know anything about 
> performance, you just plug in the mass and geometry and airfoil and the 
> airplane flies.
>   

Let me add one more comment here to help compare/contrast the different 
dynamics engines.

With X-Plane you input the aircraft geometry, mass, and airfoil.  
X-Plane takes this data and comes up with it's best approximation of the 
aircraft's performance.  Again, this is good if you are wondering how 
your design might perform.

FlightGear's YASim dynamics engine takes an approach that is not unlike 
X-Plane's approach (dividing the aircraft up into sections, computing 
the forces on the individual sections, and then summing them up.)  
However, YASim has one critical difference.  With YASim, you input the 
same sort of aircraft mass and geometry data as you would for X-Plane, 
but instead of an airfoil, you input actual performance numbers for 
cruise and approach.  The beauty of YASim is that it has a built in 
solver that creates a wing/airfoil with just the right properties so 
that you nail your performance numbers dead on for the aircraft mass and 
power you have specified.

It is interesting to see the YASim solver in action.  If you decrease 
the raw power of your engine without changing anything else, you will 
get a lower drag, higher lift airfoil.  Lower power means you will have 
a lower rate of climb, but the reduced drag means you still hit your 
cruise numbers dead on, even if it takes longer to get up to your 
altitude. And the lift is increased to balance out the reduced power.  
If you find that your aircraft doesn't slow down quick enough at idle 
throttle, perhaps you haven't provided enough engine power, so the 
solver has given you an extremely low drag wing.  You can play around 
with engine power to tune your climb rate and your deceleration rates, 
while still nailing your specified cruise numbers dead on.

All of YASim's computations are based on physical reality so as long as 
your model inputs are in a realistic range (in terms of weight and 
power) the performance will also be in a realistic range.

For as simple and clean as Andy's YASim code looks on the surface, it 
really produces remarkable results.  But again, like X-Plane, YASim is 
only as good as it's built in assumptions, which advance you rapidly up 
to some threshold, and beyond that it becomes difficult to get the model 
any closer to reality.

I'm not a trained aerospace engineer, so I've been learning about 
dynamics modeling through the FlightGear project.  I find it all 
incredibly fascinating.  Each of these tools (JSBSim, X-Plane, YASim, 
etc.) do very interesting things in very interesting ways.  They each 
have a certain beauty in their design and function.  And they are all 
very different with different strengths and weaknesses.

Curt.

-- 
Curtis Olson        http://www.flightgear.org/~curt
HumanFIRST Program  http://www.humanfirst.umn.edu/
FlightGear Project  http://www.flightgear.org
Unique text:        2f585eeea02e2c79d7b1d8c4963bae2d