[Flightgear-devel] Euler rates VS body rates - what do we see in FG?

[Nicol C. <nic...@ho...>]

Hi Guys.

Thanks for the mails. Some of the comments are 100% true (in my modest 
opinion), and some of them are unfortunately ... not 100% mathematically 
correct...

What have we established:

>>you are looking at BODY-AXIS rates and accelerations. If you are in a 
>>turn,
the pitch and roll are going to show values.

No: if you are looking at body rates, then pitch and yaw (q and r) should be 
non-zero.

>> > p,q,r are body rates. In a *steady* turn, I would expect body rates 
>>(p,q,r)
>to be zero. Phi, theta, psi (Eluer angles) are measured with respect to the
>local frame. In a steady turn, the euler *rates* would be non-zero.

Nope. Not 100% true. (p,q,r) zero in steady turn: very wrong!

Jon said it:

>>Therefore q and r should have values and so should psidot, but
thetadot and phidot should be zero or close to zero.

I agree 100%. THANKS FOR CONFIRMATION!

Before we continue... I am sticking to the definitions:
phi, theta, psi / roll, pitch, yaw - Euler angles (earth ref)
p, q, r - body rates (body roll rate, body pitch rate, body yaw rate if you 
have to - but I personnally don't really like it)

Having sorted out what we SHOULD be seeing in FG is:
1) Is what is labeled in the native-fdm packet (i am receiving via UDP) as 
phidot, thetadot and psidot really that, or is it p,q,r - they are NOT the 
same thing (I think you all know it, but if there was any doubt, let's get 
it out the way!)
2) Did I see non-zero / near zero phidot in a constant turn? I am pretty 
sure I was not on drugs!! If so, what is causing this problem?

Just to confirm, I ran the test a little differently this morning: I used a 
different machine, installation, and message interface. Yesterday I received 
the data via the UDP packet. This time I turned to the FG Properties drop 
down menu - Browse Internal Properties - Orientation. What did I find: same 
story...

Pitch (deg): -6
Pitch-rate (deg per sec): -0.9 ... low enough ... did not spend too much 
time getting it perfect
Roll (deg): -43
Roll-rate (degps): 6 ... too high! I check it a number of times... That 
would be 7 sec to get level...
Yaw ... none ?? Don't understand, but not too worried.
Yaw-rate: -6 (that I can believe!)

Pausing in flight is not a great method. Yesterday I was monitoring all of 
these 6 parameters via UDP... and the results were VERY similar. I could 
maintain a contant bank/phi angle, and non zero phidot...

So: if these are body rates (given in the internal browsing and via UDP)... 
then the labels/documentation is WRONG for the Native-FDM UDP packet (also 
on the Mathworks/Matlab site). The labels for the internal browsing is 
debatable if that is the case ... and even if it is body rates, q (body 
pitch rate if you like) should be non zero rather than p/roll rate is...

Something is fishy... Who can help?? Am I missing something? Could it really 
be that no one else has picked it up? Are there not 100s of people using it 
at least in conjunction with Matlab toolboxes?

I must say the idea of running the flight dynamics model as a script without 
need for FG sounds cool to me (an electronic/control/aero nerd in the 
making) .... :-)

Thanks VERY much again for the support/emails, and having the source in 
public domain. AWESOME work... Let's make it even better!

Cheers,
Nicol.

_________________________________________________________________
Share folders without harming wildlife! 
http://www.communicationevolved.com/en-za/