• Which the release of FS2020 we see an explosition of activity on the forun and of course we are very happy to see this. But having all questions about FS2020 in one forum becomes a bit messy. So therefore we would like to ask you all to use the following guidelines when posting your questions:

    • Tag FS2020 specific questions with the MSFS2020 tag.
    • Questions about making 3D assets can be posted in the 3D asset design forum. Either post them in the subforum of the modelling tool you use or in the general forum if they are general.
    • Questions about aircraft design can be posted in the Aircraft design forum
    • Questions about airport design can be posted in the FS2020 airport design forum. Once airport development tools have been updated for FS2020 you can post tool speciifc questions in the subforums of those tools as well of course.
    • Questions about terrain design can be posted in the FS2020 terrain design forum.
    • Questions about SimConnect can be posted in the SimConnect forum.

    Any other question that is not specific to an aspect of development or tool can be posted in the General chat forum.

    By following these guidelines we make sure that the forums remain easy to read for everybody and also that the right people can find your post to answer it.

Auto Platform causing aircraft crashes?

L_Rob

L_Rob

Messages
310
Country
us-washington
Hi guys,

Maybe this one is for Arno, or maybe one of you knows what's going on.

I used Arno's Auto Platform to harden the ground surface of my quite large 3D model and thought it worked great. (It added ~1000 platforms.) The cars would follow the streets in the model, and even use a freeway overpass while vehicles went under it. I can walk the area with the P3D Avatar, etc.

But it FSX it turns out that even with crash detection turned off, the moment the aircraft is over the model the crash effects (smoke, sparks, etc) display. The aircraft keeps flying however. (This does not happen in P3D at least from my checking so far.)

[Crash boxes are turned off in MCX, so it's not that.]

With crash detection enabled in the sim, the aircraft crashes in FSX, P3D v 2 and v3 as soon as it is over the model. I've double checked; no stray platforms floating in the air, nothing else unusual that I can see.

Removing all platforms eliminates this problem.

Any ideas of what's causing this behavior? (I didn't catch this because I use P3D as my working sim and I leave crash detection off.)
 
Hi,

Did you place the model at scale 1.0 in the sim or with another scale? The platforms don't scale with the model, so they might end up somewhere else then you thought (although you should see that with the cars as well).

Also 1000 platforms is quite a lot, maybe that is part of the issue. Are they all nicely facing upwards or also at weird angles?
 
Hi Larry:

That's an interesting result, and it compels a few questions. :coffee:

IIUC, what you are seeing is related to the "GroundRoll" display Effect (*.Fx file) attributes listed in the Aircraft.Cfg file [EFFECTS] sub-section, normally seen when a user aircraft contacts the ground (terrain surface) ex: when the landing gear are not properly deployed, or a wing scrapes the ground and/or contacts a scenery object (such as a building or other airfield scenery object ...or its crashbox extending outwards away from the object.)


Are you describing run time user aircraft crashes during user-controlled, Slew Mode, or AI mode SimObject navigation ? :scratchch

GaryGB
 
Last edited:
Thanks so much for the replies Arno and Gary. Beta testing is stressful enough for me, but worse when I don't have a clue!

Arno, the scale is exactly 1.0. I used a normal setting of .95 with the Auto Platform. Not all the ground surface was hardened with this setting, but enough to make it practical. All the platforms seemed to be appropriate. Yes, 1000 triangular platforms is a lot. The model is about 1 km square and I've already simplified the ground surface polys where I could yet still retain the look I wanted. Maybe fewer would work, but it would be nice to know the underlying cause.

Gary, I'm describing what happens in normal mode, flying an aircraft over the model. I have not confirmed all of this myself, but in P3D with crash detection turned off, there is no problem. If you turn crash detection on in P3D, the aircraft will crash and the sim will reset. (I have confirmed this last bit, and removing the platforms solves the problem.)

In FSX, with crash detection off, the effects that go with a ground crash display, but the plane keeps flying. I'm assuming that with crash detection on in FSX, the aircraft will crash when over the model.

For now I've removed the Auto Platforms, but I really like what they added. It sure would be nice to figure out how to solve this issue, but if not, oh well.
 
Auto generated platforms do not normally cause this condition, I have hard decked several aircraft carriers and LHD type models and the only thing even close I ever encountered was when I tried to auto platform a "ski jump" type deck, I did not understand or use the normal filter so I positioned several platforms manually. Airplanes could take off normally, pretty sure crash detection would have been off, but when a helicopter flew at low altitude over the edge of the sloped deck it would emit the crash effects.
 
 
Last edited:
Thanks Tejal and Gary,
As mentioned above, I have the 'nocrash' flag set in MCX options, and when the .bgl is compiled the info window included a line saying that no crashbox was generated. I'll confirm again. Never hurts to return to basics when there is a problem! However even it was a crashbox, it would have to be one with a vertical dimension of several 300 or 400 meters, far above the highest part of the model.

Rick, so it sounds like you have encountered a similar problem (crash effects but no crash) when the heli was not in contact with the platform. Right? Do you remember how far away the heli was from the nearest platform when the crash effects were displayed?

I've been manually hardening the decks of my ferries and not had this problem. One significant difference is that the platforms on the ferries are all exactly horizontal. I wonder if that makes any difference.

[edited to add] I have confirmed that the model in question has no crashbox(s) by using the 'display crash boxes' option in MCX.

In trying to think of alternate explanations the following came to mind. It's not an adequate explanation but maybe it will trigger another thought. In effect, each platform has it's own crashbox. The aircraft should crash if you try to go through the platform. If one has a very large area covered by a number of platforms, from the standpoint of the aircraft perhaps these are all calculated as one large rectangle. Since Auto Platform creates tilted surfaces, perhaps this would have the effect of tilting the entire horizontal area thus giving it a vertical dimension. Super-imposing all the different tilts from the different individual platforms might end up creating a large box that is not a crashbox, but somewhat functions as one. OK, this does not make sense. So i'm still at a loss.
 
Last edited:
Here are a few images, you can see it is very close. At the time I thought it was an anomaly, which is why I recorded it. I researched and experimented fairly intensively with things like crash box density, the idea being that you can have one box representing the entire model, in this case mast head to stem to stern - not a good plan for an aircraft carrier; or you can have thousands of boxes that amount to a crash box filled representation of your exact model, with the corresponding model complexity and size. I chose something in the hundreds or dozens that left the flight deck clear, the helicopter's condition would have to be the result of poorly tilted platforms, imo.
crasheffect3.jpg

crasheffect2.jpg

crasheffect.jpg
 
Hi Larry:

I'm thinking that the CrashBox attribute may not be the basis for triggering the above Effect (*.Fx) in the user aircraft, since IIUC, the aircraft does not truly "crash", and instead continues navigating; thus one might wonder whether contacting the BoundingBox (aka "B-Box") of a scenery object and/or a platform placed above ground level may instead be involved.

I have not yet tested to see if the above cited "GroundRoll" display Effect (*.Fx file) attributes listed in the Aircraft.Cfg file [EFFECTS] sub-section is also triggered by user aircraft contact with the BoundingBox (aka "B-Box") of a scenery object and/or a platform placed above ground level.


One possible obscure 'trick' to try that may "fool" the FS run time rendering engine, is to place the large ground surface object with its base RefPoint(s) slightly below the local terrain mesh ground surface (at a negative or 'minus' elevation) by at least -0.01 Feet AGL. :idea:

Or are you already doing this 'below ground object base RefPoint' (negative or 'minus' elevation AGL) placement for this 3D ground surface BGLComp-XML-type scenery library object ...with the workflow you described in another recent thread ? :scratchch

GaryGB
 
Last edited:
Rick, this must be similar to my problem. I'm assuming that you did not have crashbox(s) enabled. One would think that when an aircraft is well away from the nearest platform as is shown in your clips, that it should not either crash or display crash effects. This does not happen on carriers where the flight deck is always horizontal, right? So it must have something to do with the tilt as you mention.

Arno, could the fact that individual platforms are tilted account for this? If so, why? This might be strictly coincidence, but from Rick's last clip, it looks like if one drew a line parallel to the deck at the point nearest the heli and then translated this line to intersect the stern of the boat it would then intersect the heli.

Gary, putting the model datum slightly underground (under water in this case) is worth a shot! I checked the bounding box; the top of it is far below altitudes where the aircraft run into trouble --- trouble meaning that if crash detection in the sim is turned off, you only see the effects; if it is turned on, the aircraft crashes.

[Edited to add] A few more checks in P3D v3; the aircraft will still crash when passing over the model at over 4000 feet! Although the model seems intact, as Rick has mentioned sometimes with repeated saving in MCX, corruption is possible. I should recompile directly from the dae, add the auto platforms and confirm or disconfirm the issue.
 
Last edited:
I've had the exact same problem with ski jumps.
You will notice the problem is worse with longer aircraft. What seems to happen is the plane pivots near it's centre as it climbs the ski jump. Therefore the nose-gear ploughs through the surface and clips the crashboxes, causing sparks.
The solution is to duplicate the platforms under the ski jump and therefore prevent the crashboxes from being generated.
So, for Gary's HMS Invincible, merge in duplicate ski jump platforms 125cm down and export with crashbox granularity of 3.
skiJc.jpg

The ideal solution would be to tag the ski jump only as 'nocrash'.
 
Hi Essex:

Interesting suggestion ! :)

BTW: Any apparent "HMS Invincible" images above were posted by Rick Keller as, IIUC, author of the described 'platform' modifications to that model in a 'worked example' that also triggers the SimObject "ground contact Effect (*.Fx) display anomaly" under discussion in this thread. ;)

GaryGB
 
Last edited:
I did the Vikramaditya, a vessel purchased from Russia in a very similar arrangement to China's Liaoning; both of them developed from Kuznetsov class carriers. On this page at Sim Outhouse, user gp183601 claims authorship for Invincible and elsewhere for all Kitty Hawk Class CV's modeled for the sim. I did not find a name for this user.
 
Rick,
Just if you feel curious: I'm wondering how high the heli can be above your carrier deck and still crash or trigger crash effects. I was really surprised to find that I got a crash at 4000 ft alt. One problem would be to make sure the test involved flying directly over the deck as seen far below. I really hate to cry 'bug' in the sim code when this may be operator error (me), but I'm starting to get suspicious.

If I can find the time, it might be worth playing with just one tilted platform, or maybe one horizontal one and one tilted one to see if I can cause a crash that way.
 
The elevation limit of the condition is expressed in the images. In fact it was barely noticeable for the MiG's negotiating the ramp. The reality is that it is an extreme transition for burdened jets and there must be very specific weight restrictions to allow it. The slight spray of sparks seems almost typical for that. I did extensive tests of the model before releasing it and I would have to say the condition is consistent with the description provided by Essex, for example a jet could perform a low pass without triggering the condition; I just happened to notice it with the KA-27 and "explored" it.
 
If you're putting platforms on several different materials, you could try just putting platforms on only individual materials to isolate the one causing the problem.
 
Guys, I'm reasonably confident I've found a solution to my original problem; aircraft crashing when flying over a large model with the ground surface hardened with Arno's auto platform.

Originally I had used a normal setting of .95. Based on a few brief experiments, I think this represents the cosine of the maximum angle a triangular platform can be tilted from the horizontal, ~18 degrees in this case. I created platforms on 3 or 4 different versions of the model so I somewhat doubt that the cause of the issue was an accidental error in application. In adding platforms to a couple of remaining seasonal model variants, I tried using a setting of .97, which I think, corresponds to ~14 degrees from the horizontal. Now no crash, not in FSX, P3D v 2 or in P3D v3.

Why this is now working I have no idea!

Turns out there was an odd side effect; the hardening was preventing a problem wherein one could see through the ground surface of the model if it was behind a railing or a fence. The sim seems to treat the hardened surface as if it were a ground surface. Can't see through that. But the solution to that was in adjusting the material settings for the railings and fences.
 
You must log in or register to reply here.
Back
Top