Aircraft ocassionally teleports to the underground

Hi, We have an issue with aircraft abruptly teleporting to below-ground level.
So far, we’ve validated that this happens even with no extensions on the
affected machines. Also, this happens to other aircraft, not just ours (albeit
with much lower frequency). On top of that, I’ve validated that we’re not
setting the altitude ourselves via SimConnect. See examples:

@FlyingRaccoon fyi

Just to add to this: - It’s not reliably repeatable in any scenario. Some
machines seem to have it happen more frequently, but we couldn’t find
correlation (system specs etc). - It doesn’t seem scenery related as we
couldn’t find any correlation to specific approaches/airports having it happen
more or less (shooting the same approach 30 times might only yeald this
behaviour once). - Even if slewed back onto the tarmac/air, the aircraft will
simply continue to spin/flip in place as if the flight model is completely
corrupt. - This issue first cropped up around SU10 IIRC, not correlated with
any particular update from our side, (we verified this by sending affected
users various older builds of our A320 and there was no change in behaviour).

Hello @AlexxV [@FNX_Dave](/users/2078/fnx-
dave.html) Hard to tell what’s going wrong without a reliable way to reproduce
the issue to investigate it. What airplane should I test to have the maximum
chance to reproduce the issue? Does it happen with the clear skies weather
preset enabled? Regards, Sylvain

Hey Sylvain, Thank you very much for responding, indeed this is a tricky one
which is why we’ve not reported it until doing as much testing on our end as
possible, but we’re very much at a dead end after several months. Honestly it
seems the best aircraft to reproduce the issue in order are the Fenix A320,
followed by the A310, and very occasionally the PMDG 737. Indeed users have
been able to reproduce it with clear skies. We have one user who can very
reliably experience this issue (e.g. with 50% chance on approach), but users
with similar/identical specs do not get it. To keep things clean we’ve had
affected users remove/uninstall all other add-ons and test purely with vanilla
MSFS + Fenix, and the result is the same. If it’s not something you can
reproduce within a reasonable time, we’d really value it being shown to some
of the engine techs to see if they’ve ever seen anything like this or have any
ideas. We’re a bit stumped at this point! Thanks again, Dave

Hello @FNX_Dave, Our best chance is to gather
data and info from one of your user who has the issue reliably. Any chance you
could redirect the user who encounters the problem 50% of the time to this
discussion? The fact it happens with airliners during approach makes me think
about issue with flaps or spoilers. We would need to know if the user touches
any specific input before it occurs. Also, recording a video with DevMode
aircraft debug panels enabled would probably be useful (Drag, Tracking,
Weight, Sim Tuning) Regards, Sylvain

Hey @FlyingRaccoon, We have a potential lead, and a video. See here: In this one the most interesting item to us is
the Load factor Ny value showing -nan(ind) just as it occurs, suggesting
something is being divided by zero perhaps? I have a couple more videos below,
showing other panels and instances if it’s of use: As a follow up
to your other question, we could not find any consistent user inputs that
occur before the event (either flight control or otherwise). Additionally, our
flight modeler is struggling to debug much without more documentation/detail
about the flap model, e.g. we saw that roll does affect CL somehow but it’s
not documented either. Thanks for the help, Dave

Hello @FNX_Dave Thanks for the videos. Yes a division by 0 is what I’m
suspecting as well. The local factor nY is a consequence of the issue rather
than a cause I think. Same for really high lift and drag values which I assume
are caused by a really high velocity but don’t give us the cause. I’ve
forwarded those videos to our core sim engineers. In the meantime, any chance
you can record a setup similar to the first video but where debug views are
fully displayed? (we can only see a subset in yours) Regards, Sylvain

@FlyingRaccoon thank you very much for the quick response. I will ask the
customer in question to record further videos with more of the debug window
open/visible. Unfortunately he’s quite limited in resolution so it’s
challenging to get all visible at the same time.

Not necessarily more windows. The ones displayed in the first video are fine
but they need to be entirely visible (No scrollbars).