AirplaneX - Sinking into ground

[[aps]gnat 28]

Converting my aircraft to airplaneX simulation .....

Rumour has it if I add;

driveOnComponent[] = {"wheel_1_1", "wheel_2_1", "wheel_2_2", "wheel_3_1", "wheel_3_2"};

and have these (basic shaped) wheels in the GEOMETRY LOD (not a Phsyx LOD) it should work .....

Doesn't.

Airplane slowly (~5 secs) sinks into the ground until it rests on the main body.

The LANDCONTACT LOD doesnt seem to do anything .....

Adding copies of the wheels to the Geometry Physx LOD stopped it sinking, but it typically then wont move.

Thoughts?

Thanks!

(need sample from BI !! )

[abs 2]

I got your multiturret demo working as an airplane (haven't tested the multiple turrets, yet). It does not sink in to the ground (it used to) and can drive (it didn't use to). You can find it here: http://ohplz.ca/gnat/

I've added weapons to it, but I'm not sure if I got all turrets as I'm not 100% familiar with the plane.

I've commented the config changes I made, and all I did in the p3d file was modify the Geometry lod.

- Copied your wheels from the res lod to the geo lod

- Added mass of 30

- Named them Component03 AND Wheels

---> If I only named them Component03, they didn't sink, but didn't drive.

---> If I named them only Wheels, it sunk.

You can find the plane under Independant > Air.

Can't wait to see your results with your multi-turret tests....please do report your findings back!

Abs

(PS: If you could do me a small favour, and try this out? http://forums.bistudio.com/showthread.php?168902-buldozer-flashing-at-night I'm wondering if I'm the only one, or if there's some other step I need to take to get it to work. Thanks!)

Edited December 2, 2013 by Abs

[progamer 14]

I may have found a solution to your problem Gnat.

The config is for both airplaneX and helicopterX:

simulation = airplaneX/helicopterX;

Aircraft currently use only the simple wheel definition (see bottom of this post for more details), specify the wheels, the names come from the geometry. Aircraft are special in that they require both the wheel name and the damper name to be defined as a pair. This allows the wheels to rotate independently of one another.

The proper setup for helicopterX and AirplaneX wheels:

wheelComponentSteerL[] = {"wheel_1_1","wheel_1_1_damper"};
wheelComponentSteerR[] = {}; // typically nose wheel + main undercarriage
wheelComponentFixedL[] = {"wheel_2_1","wheel_2_1_damper"};
wheelComponentFixedR[] = {"wheel_2_2","wheel_2_2_damper"};

Specify the size (in m), strength and damping of the vehicles suspension:

(Part of shared vehicle definitions)

damperSize = 0.3;
damperForce = 5;
damperDamping = 20.0;

Specify the strength of the brakes:

wheelBrakeFrictionCoef = 0.4; //values 0.0 -> 3.0 in 0.1 steps

Specify whether or not the vehicle is amphibious:

canFloat = false; // some aircraft can land in water, most can not

Set how much water (m^3) will fill the vehicle per second. Affects how fast the vehicle will sink when driven into water.

waterLeakiness = 25.0;

If the vehicle is amphibious set how much resistance the hull experiences in water, affects acceleration/deceleration and max speed. Higher means more resistance:

waterResistanceCoef = 0.004;

Also:

Mass should be distributed realistically (selecting the entire mesh and applying a uniform mass will cause poor results) and the Center of Mass should be as realistic as possible.

Depending on configuration type used, memory LOD should have single points for axis and bound of each wheel.

Simple wheel definition:

The simple definition MUST be used in conjunction with the shared values above, default values will lead to poor behavior in the majority of cases.

Specify all wheels which steer, names taken from the geometry LOD:

wheelComponentSteerL[] = {"wheel_1_1_damper"};
wheelComponentSteerR[] = {"wheel_1_2_damper"};

Specify all wheels which do not steer, names taken from the geometry LOD:

wheelComponentFixedL[] = {"wheel_2_1_damper"};
wheelComponentFixedR[] = {"wheel_2_2_damper"};

Shared wheel definitions:

Note that these values will be used if their counterparts are not defined in the alternative definition

Specify the size (in m), strength and damping of the vehicles suspension:

damperSize = 0.3;
damperForce = 1.5;
damperDamping = 100.0;

Specify the weight (in Kg) of each wheel and its related suspension parts, heavier leads to more stable vehicles, should be set reasonably realistically:

wheelWeight = 175;

Hope this helps you! :cool: Tell me if my configs are correct and help anyone please!

Edited November 27, 2013 by ProGamer

[[aps]gnat 28]

Thanks guys. Suspected suspension definition after reading the Car tut.

Buried in RL Work, will advise later.

[progamer 14]

Glad we could help!

Edited November 29, 2013 by ProGamer

[[aps]gnat 28]

@Progamer

Where did this stuff come from?

Because its not working and much of it is not found with ArmA3 (or ArmA2)

e.g. wheelWeight

[abs 2]

Gnat;2567908']@Progamer

Where did this stuff come from?

Because its not working and much of it is not found with ArmA3 (or ArmA2)

e.g. wheelWeight

I can 99% guarantee that he grabbed all that from the VBS2 Developer Reference. (http://manuals.bisimulations.com/vbs2/2-00/devref/#Adding_Models/How_to/AMHT_Physx.htm)

Did the model I link not work for you?

Abs

[John Spartan 89]

Ok since I have some experience on this exact subject I'll share my "experiment" results.

First of all I had the same problem as Gnat described and we discussed this last night with him.

So the way I managed to solve this is quite simple and almost as Abs described in posts above.

- GEO lod contains wheels as actual model

- GEO lod, wheels are named not only as component but also it has to hold information about all subsequent/included animated selections [like wheel_1, component01 and suspension_1 are saved as same wheel poly named selections]

- GEO lod, mass is applied to that wheel object

- GEO lod wheel poly must be "Must be closed and convex"

- also don't forget the land contact, points there must have similar names as you use for the wheel in GEO lod

- in cfgvehicles use driveOnComponent[] = {"ngear", "lgear", "rgear"};

- airplaneX or airplane as a simulation does not matter at this moment [depends what BI is bringing down the pipeline latter on]. To be honest Progamer is not really right on that airplaneX realistic simulation [its nothing but a copy of car PhysX, the plane starts behaving as crap in air but drives good on tarmac... my opinion/observation]

It worked for me on FA18 and SU35

Edited December 2, 2013 by John_Spartan

[progamer 14]

I can 99% guarantee that he grabbed all that from the VBS2 Developer Reference. (http://manuals.bisimulations.com/vbs2/2-00/devref/#Adding_Models/How_to/AMHT_Physx.htm)

Did the model I link not work for you?

Abs

I said I "may" have. BI has not given us really anything on the airplaneX simulation or the other physx simulates yet. I saw a bunch of it was similar to Arma 3, so that why I assumed other parts could be as well.

[abs 2]

I said I "may" have. BI has not given us really anything on the airplaneX simulation or the other physx simulates yet. I saw a bunch of it was similar to Arma 3, so that why I assumed other parts could be as well.

I don't get what you're being defensive about. Did you or did you not get the information from the VBS2 DevRef? No need to explain to me what the meanings of the words "may" and "could" are.

Abs

[progamer 14]

I don't get what you're being defensive about. Did you or did you not get the information from the VBS2 DevRef? No need to explain to me what the meanings of the words "may" and "could" are.

Abs

I got it from the VBS3 dev ref.

[[aps]gnat 28]

@Abs

Sorry, tried your link for the first time today, doesn't work.

@John

Thanks, still trying to get it to work. Noticed that if I add extra vertexes under wheel in the LANDCONTACT mod it alters which wheels sink and which don't, but havent found a combination where all dont sink.

I got it from the VBS3 dev ref.

Shit a brick!! If I'd know that I wouldnt have wasted my time ..... Where ever it comes from, don't pretend its yours, provide a link to the info in future.

[abs 2]

Gnat, that's really odd. What exactly didn't work?

Try placing an AI in the driver's seat with you as a commander on the ground. If you tried piloting it yourself, it's just not as responsive and takes a while to get going.

When I get home I'll provide a video of it working if you want.

Abs

[[aps]gnat 28]

No, your download link doesnt work, doesnt download :)

[abs 2]

Gnat;2568921']No' date=' your download link doesnt work, doesnt download :)[/quote']

Oh shit! Sorry man. Just go to http://www.ohplz.ca/gnat/ and download the file there. Hope you can open 7zip files.

Abs

[Edit: So used to using https for work...looks like it snuck in there.]

[odyseus 19]

Hey Gnat, did you get it? if not let me know. I think i was able to make it work on my plane. i can send you a sample if you need it.

[[aps]gnat 28]

thanks Abs, tested, works, but there very limited complexity to the wheel / undercarriage in that model, so its hard to compare ......

looking to sort through the various differences .......

Oh BI, wheres my A3 plane sample ? :(

[[aps]gnat 28]

Ok, some more testing, and some interesting results.

The LANDCONTACT lod is not required, but oddly changes the planes acceleration on ground. Slower takeoff. Not conclusive if it affects flight characteristics.

No longer using a PhysX lod, unclear if that's normal or not.

For the "wheel" elements in the standard Geometry lod seems it MUST rotate. Oddly once that one fact was changed the plane no longer sank into the ground.

This now means the GEO lod wheel has to be VERY round, otherwise it either won't move forward (big flat spot on the bottom of the wheels) or bumps and jumps along the ground like its got square wheels LOL

This probably explain some odd "bumpy" ground behaviour of the A-143 Buzzard.

Variation testing continues. (and suspension still doesn't work)