CRJ700-family sim crashes

[lissalissa]

Hello, all,
I'm having an issue with certain members of the CRJ700-family. The CRJ700 loads in fgfs just fine. But as soon as I try to load ANY of the other models in that family, fgfs crashes shortly after loading the hydraulic system. When loading the CRJ700, the germane console output is:

Code: Select all

Hydraulic done.
warning: The VAD has been replaced by a hack pending a complete rewrite
This aircraft does not use the latest yasim configuration version.
CRJ700 aircraft systems ... initialized
Pilot dual control ... initialized
Sound on.
Takeoff mode: taxi-and-takeoff
Starting takeoff (taxi-and-takeoff) announcer
Takeoff mode: takeoff
Starting landing (takeoff) announcer


This compared to the console output when the CRJ900, or any of the other non-700 CRJ's, crashes on load:

Code: Select all

Hydraulic done.
warning: The VAD has been replaced by a hack pending a complete rewrite
This aircraft does not use the latest yasim configuration version.
YASim SOLUTION FAILURE:
Insufficient elevator to trim for approach
Segmentation fault

My assumption is that the problem lies in a script file somewhere in the CRJ700-family, and if I knew which script was the likely culprit I could compare the two files and figure out where the issue likely lies, but I've not done development with flightgear, so I'm not sure where to look. Can someone either 1) point me in the correct direction, or 2) propose an alternate cause/solution?

I'm using the CRJ700-family downloaded via the Aircraft center last week, and I'm running fgfs v2017.3.1 on Ubuntu 17.10

[wkitty42]

that solution failure thing is where YASim is running the calculations on the craft's design and weight numbers and not finding a positive solution within some number of iterations of the formulas... that's generally 10000 interations, IIRC... looks like it doesn't like something about the broken crafts' FDMs... the FDMs are where the craft is described in numbers... like the wing sweep, wing shape, center of gravity, stabilizer information, aircraft weight, etc...

[sp-lcz]

The CRJ700 loads in fgfs just fine. But as soon as I try to load ANY of the other models in that family, fgfs crashes

Any other? I rember that version 900 and 1000 had the same problems but ER and LR versions of 900 and 1000 worked fine for me.

BR
Lukasz

[lissalissa]

The CRJ700 loads in fgfs just fine. But as soon as I try to load ANY of the other models in that family, fgfs crashes

Any other? I rember that version 900 and 1000 had the same problems but ER and LR versions of 900 and 1000 worked fine for me.

BR
Lukasz

You are quite correct; my apologies. It looks like it's just the 900 and 1000. I take it this is a known issue and thus not one fixable by me?

[D-ECHO]

The problem is a bad CoG (Center of Gravity). A quick fix/hack would be to lower the number for x in this block

Code: Select all

<ballast x="15.3" y="0.0" z="-1.9" mass="10000.0" />

(line 184) in CRJ900.xml from 15.3 to something around 5-9, you can experiment which one gives you the most realistic results (in terms of nose/tail heaviness).

[jsb]

*upps* That should not have happend. I am sorry to head that a broken version made it into the official repository. Unfortunately my free time is somewhat limited but I am quite sure I have a development branch on my PC where this problem is fixed. In the mean time (before I can test and push an update) you may want to experiment with the <ballast> and change the mass.
Unfortunately the wing of the CRJ700-family is modeled in two segments and I did not implement the MAC calculation for this case in yasim yet, anyhow, you may try:

<installation path>\FlightGear 2017.4.0\bin\yasim.exe <FGDATA path>\Aircraft\CRJ700-family\CRJ1000.xml

Code: Select all

SOLUTION FAILURE: Solution failed to converge after 10000 iterations
==========================
= YASim solution results =
==========================
       Iterations: 10002
 Drag Coefficient: 16.749
       Lift Ratio: 140.189
       Cruise AoA: 0.38 deg
   Tail Incidence: 2.07 deg
Approach Elevator: 0.000

               CG: x:-0.015, y:0.000, z:-1.228
    Wing MAC (*1): x:-1.44, y:8.00, length:2.4
    CG-x rel. MAC: -0.599
    CG-x  desired: -2.155 < -0.015 < -2.036

Inertia tensor [kg*m^2], origo at CG:

  206764.344,   0.000, 159921.313
    0.000, 2738790.000,   0.000
  159921.313,   0.000, 2896078.250

(*1) MAC calculation works on <wing> only! Numbers will be wrong for segmented wings, e.g. <wing>+<mstab>.


Compare the CG x-value of a working variant and your modification to a non-working variant to get a hint in which direction you should tune the CG / ballast.
I hope I find some time to fix this and push an update.