Underbody points digging in? Overhanging nose

Discussion in 'Car Modding' started by MileSeven, Apr 19, 2022.

  1. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    652
    Likes Received:
    256
    We’ve discovered that under some circumstances, one of the vehicles I’m working on (a closed-wheel gt/sedan with a long overhanging nose) will come to a sudden stop if you go off the track and ride across bumpy grass etc.

    It’s as if the car has buried its nose in soft mud!

    The vehicle has three underbody points defined on the leading edge of the nose. To facilitate analysis of whether the nose was grounding, I defined two of those points using the four ‘special’ underbody definitions (as these ones are responsible for ride height data?). Typical ground clearance can be of the order of 50mm or so at times, reportedly.

    No diffuser defined - so I thought this usage would be ok.

    I’m guessing that the special underbody points might be at fault and I should swap them around to a more conventional location?

    …or are all underbody points likely to cause this issue if they strike non-track surfaces?
     
  2. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,388
    Likes Received:
    6,602
    Undertray points (HDV points) shouldn't do this, you'll only get the defined friction from them.

    You should check your collision body - it should be a 'blob' with no projections or holes. Acute angles will catch surfaces.
     
    MileSeven likes this.
  3. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    652
    Likes Received:
    256
    Thanks - I thought it was a bit strange if it was an underbody thing.

    Now, where did I leave Google?o_O
     
  4. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    652
    Likes Received:
    256
    As far as I can tell (exporting it from 3DSimEd in da0 to Blender), the collision mesh doesn’t have any obvious gaps or protrusions but does have a rather ‘bluff’ nose (where do you have to host images to attach them here?!?:().

    The reference plane looks to extend to the front of the vehicle and then has a vertical plane representing the bumper/nose. This looks to give it a 90-degree leading edge where the nose would impact the scenery - would it benefit/does it need a more rounded (or angled/raised) lip instead?
     
  5. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    652
    Likes Received:
    256
    AC8AA595-4D84-4EEC-B99B-C60664E334D5.jpeg

    Here you go - this is the as-is collision gmt visualisation for the nose.
     
  6. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,388
    Likes Received:
    6,602
    What's with the shading? Makes it a bit confusing to look at. Is that at all indicative of some errant verts or unusual construction?

    Tell you what, rather than a fair bit of back and forth trying to identify anything, just comment out the coll body in the .gen file. The game will make its own very rough collision body. If that cures your problem, you know where the issue is.
     
  7. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    652
    Likes Received:
    256
    Apologies @Lazza - I’ve only learnt overnight how to get the vertex view in 3DSimed and don’t have that view on this phone…

    I have tried adjusting the lower leading edge up and it shows an improvement but needs testing more widely. I’ll also try your comment method.
     
  8. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    652
    Likes Received:
    256
    So...

    It appears that:
    a) The undertrayparams co efficient of friction entry for the original mod was set to 2.5 (all other vehicles I've checked so far are <1.0)
    b) The track I was doing most of the 'excursion' testing has a CollFrict TDF setting roughly twice that of other randomly-sampled tracks according to the modder and it has been reported that the stuck-in-the-mud effect is much reduced when this is 'corrected'.

    I've also massaged the coll.gmt to better visually match the skin. This raised a question though - how does rf2 align the collision mesh and HDV and is there any way to compare the bodywork and collision GMT in 3DSimEd to make sure they match sensibly enough?

    Colmeshnew2.jpg
     
  9. lonestar29

    lonestar29 Member

    Joined:
    Apr 30, 2013
    Messages:
    87
    Likes Received:
    83
    MileSeven likes this.
  10. MileSeven

    MileSeven Registered

    Joined:
    Apr 1, 2014
    Messages:
    652
    Likes Received:
    256

Share This Page