Free Flight, Free Speech

I've also figured out how to use OpenVSP to calculate the moments of interia; this will probably be better than Simon's method where you define the positions and mass and let JSBSim figure it out, simply because JSBSim makes assumptions about the size of the masses. VSPAero will take slices through the model and based on the density you assign will calculate the mass and inertias. I've added a section to my writeup (http://chateau-logic.com/content/using- ... flightgear) that covers this.

So here is a new version of my FDM with fixed instability, fixed moments of inertia and revised geometry. http://chateau-logic.com/sites/default/ ... 6-11-25.7z

I shall now go and fly sanhozay's new fdm and see how it compares

Richard wrote:Wing incidence I now have at 0; but adjusted the tail to 1.9 (leading edge down) take this into account for stability purposes however it's not going to be quite right - I did it like this after getting very confused about whether or not I needed to use wing alpha or FRL alpha for the coefficient base of all of the surfaces. I spent about 7 evenings tuning this to get the right pitching moment.

Interesting. I fudged mine with elevator trim, which is not a long term solution but I needed it to make it glide and to avoid the nose pitching down during the takeoff roll. Default Aeromatic ignores wing incidence I believe, and something I was going to try was use alpha-wing-rad in place of alpha-rad.

Your new inertias:

My inertias estimated by Aeromatic++:

What units are you CoG numbers? Feet? And where's the datum?

Interesting that our inertias are close; that's a testament to the new Aeromatic, (and maybe the old one). I've spent a lot of time trying to figure out how to calculate MOI over the last few weeks - I've always understood the calculation of MOI to be a bit of a black art practised by skilled slide rule practitioners in darkened rooms with nary a glimpse at a computer (I'm probably very wrong about this).

All my dimensions are in feet. For VSPAERO the latest run was using AERORP of 6.8,0,0 - however as part of my pitch moment tuning I've had everything from 8.2 to 5.2; from the Beagle figures below 6.8 is close to the aft limit (assuming all my other measurements are right which is unlikely), so the 6.8 is really the result of tuning to get the right sort of stability and basic pitch moment curve..

Using the usual 25% MAC I came up with a CoG range from 6.4981, 6.9211 - using a wing root X of 5.779 and a MAC of 4.23

Beagle themselves refer to an older way of expressing this (SMC); when they state (aft CG limit of 27.5% SMC) (S/b) - which gives us a range of 6.39 to 6.83 (feet);

I've just been flying your revised FDM and it's definitely an improvement. You should be using a kinematic for the control surfaces as they snap to position as they're directly linked to the input device positions - and whilst most of the time with a joystick you won't notice this it is noticeable when using the keyboard (centre all controls NUMPAD 5) or mouse. I think the controls on the craft are connected directly by wire so a 0.2 second kinematic would probably be a good point to get controls that move as though they are connected to things with mass and resistance.

It all goes horribly wrong though if I have maximum deflection (stick back) on the elevators; at 30 degrees of pitch the model becomes numerically unstable. During the takeoff roll I had a lot of trouble controlling the yaw using the rudder but managed to get airborne on my second attempt.

In flight it feels much too responsive; almost massless with no inertia (which I know isn't true because I've seen your inertias and they are the same as mine); so maybe it's the p,q,r damping that isn't right.

This is genuinely so much fun. .....thank you guys for taking part.

Thanks for your advice regarding the kinematics. I never fly with the keyboard so I wouldn't have noticed.

Richard wrote:In flight it feels much too responsive; almost massless with no inertia (which I know isn't true because I've seen your inertias and they are the same as mine); so maybe it's the p,q,r damping that isn't right.

I'm surprised by this. Compared to my previous model it's much less responsive. I thought it felt on the stiff side compared to other Flightgear models but I'm wary of making such feel judgements as I've never even set eyes on a Beagle Pup, let alone flown one.

If you feel like playing with them, there are tuning coefficients under fdm/tuning for pitch, roll and yaw damping. They are set to 1.0 to use the Aeromatic defaults and simply scale the pitch/roll/yaw moment due to pitch/roll/yaw rate elements.

sanhozay wrote:@IAHM-COL: When I've finished tuning, I was hoping you'd test the 11,700ft service ceiling by taking it to Innsbruck.

Getting out would probably be more interesting than getting in.

T'is the track

http://mpserver15.flightgear.org/modules/fgtracker/?FUNCT=FLIGHT&FLIGHTID=6843083

Flies nice

You went that way out of Innsbruck on a 100hp engine!

Thanks for testing. So about 25 minutes to get to 10,000 from 1,900 and 40 minutes to get up to service ceiling. That's in the right ballpark. When I have a half hour to waste, I'll log a climb from s/l to 10,000 (which should be ~31 minutes).

I had good luck.
I had marginal VFR till about 8000, so I could just rollercoasted the mountain side visually. When I reached 8000 feet the weather changed to overcast at low altitudes and I began flyin a haze, so I switched to VFR and kept 11700 for cruising

Update on the way.... hopefully I've not forgotten anything...

I've added an engine system file to reduce the Ve at low throttle and something for the start-up to look a little less like you're throwing the switch on an electric motor.

Also I've included this time my working documents, so you have my spreadsheets and drawings plus url links to web based data.

anytime you ready shoot me the email, and I push