The .tgm, and +ttool

Discussion in 'Car Modding' started by Tim Bennett, Jan 17, 2012.

  1. McFlex

    McFlex Registered

    Joined:
    Feb 23, 2012
    Messages:
    1,031
    Likes Received:
    317
    Today i wanted to test the tire tool. I added "+ttool", saved it and clicked the developer icon. Then i get 2 black "loading screens" and it crashes to desktop. The TireTool "icon" is still in the taskbar and i see a running rf2 mod mode in the task manager but there is no tire tool. Any ideas?
     
  2. Knut Omdal Tveito

    Knut Omdal Tveito Registered

    Joined:
    Mar 16, 2012
    Messages:
    47
    Likes Received:
    3
    Just wanted to share a little Excel tool for building tire geometries in a quick fashion:
    Download here: https://dl.dropboxusercontent.com/u/14052158/TireToolSheet.rar

    I included a few examples which I did some quick validation against FEA analysis results (both Excel sheet and simple .tgm included):
    -175SR14 (1996)
    -235/45ZR17 (2000)
    -295/75 R22.5 Truck tire (2006)

    Some comments:
    -So far it's only 31 nodes and the number of nodes on the sidewall/tread etc... is fixed. Also the number of plies is limited to 4. But so far this hasn't been a limitation.
    -Note that these were made quickly and the material data can be improved. But the calculated tires were adjusted to get close to FEA results and experimental results.
    -Realtime section of the examples TGM only has a few test scenarios and only slidingBaseCoefficient to simplify the validation againts experiments. If you want to use them in rf2 you need the whole WLF model (Adhesion/Micro/Macro curves) and more test scenarios in the TGM file.

    View attachment 14146
    View attachment 14147
    View attachment 14148
    View attachment 14149

    I also have a simple brush model which is useful for the realtime section. When I have time, I will finish it and post it here.
     
  3. Led566

    Led566 Registered

    Joined:
    Jan 25, 2012
    Messages:
    358
    Likes Received:
    24
    Very well done.
    Thank you!
     
  4. Knut Omdal Tveito

    Knut Omdal Tveito Registered

    Joined:
    Mar 16, 2012
    Messages:
    47
    Likes Received:
    3
    Here's the brush model.

    Note!!! When I input the WLF friction curves into the tgm and ttool the results didn't make sense. Suddenly the macro peak was switched with the adhesion peak and they got scaled up and down. However I managed to get the same WLF transform and the correct location of the peaks. So this version is only to demonstrate how contact patch size, tire stiffness, temperature, velocity and static/sliding friction coefficient works. Will post a new one once I have time to dig into rf2 telemetry.

    Download here: https://dl.dropboxusercontent.com/u/14052158/TireBrushModel_TGM_V1.xlsx

    View attachment 14163
    View attachment 14164
     
  5. Knut Omdal Tveito

    Knut Omdal Tveito Registered

    Joined:
    Mar 16, 2012
    Messages:
    47
    Likes Received:
    3
  6. Dhurara

    Dhurara Registered

    Joined:
    May 12, 2012
    Messages:
    16
    Likes Received:
    0
    Thank you for sharing this great work:D
    Honestly speaking, I can understand your work partially because of my limited knowledge.
    But I think your work will bring us breakthrough for rF2 tyre modding.
    Keep up the good work.
     
  7. lordpantsington

    lordpantsington Registered

    Joined:
    Oct 5, 2010
    Messages:
    849
    Likes Received:
    79
    So where are people putting their ptool/ttool folders (RE 798)? I had previously put them in Program Files, but since there is data creation and files moving in and out of that, it seems wrong.
     
  8. WiZPER

    WiZPER Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,517
    Likes Received:
    186
    HELP! I simply can't make ttool show deflection/compressions - neither as points nor "3D". Everything else seems to work correctly, and yes I HAVE raised the "ground" and altered the camber to test... frustrating! Happens (well not) in both B930 and older B798

    Hmm_noflex_ttool2.jpg
     
  9. Michael Borda

    Michael Borda Car Team

    Joined:
    Oct 21, 2011
    Messages:
    187
    Likes Received:
    143
    Did you remember to set iterations per frame to something non-0? (like 20 or so). Also many of the tyres require a DT multiplier of less than 1.0 to work (or the tyre may 'explode'). That looks like a GT tyre to me, so I would suggest a DT mult of ~0.8.
     
  10. WiZPER

    WiZPER Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,517
    Likes Received:
    186
    It runs tests just fine, but doesn't visualize it other than by heating, contact patch and bristle temps - ZERO visual deformation is my issue.

    I've previously been able to see the deformation and compression after loading a working .TGM and doing RealTime testing...

    I've rendered tyred without expolding mesh using DT=0.92, on auto it seems to self adjust in B930 though, usually goes down to 0.88 or there about. (using all cores to the max, which is cool btw)

    Rendered tyres DO work, it just annoys me that I can't use the graphics :(
     
    Last edited by a moderator: Feb 23, 2015
  11. WiZPER

    WiZPER Registered

    Joined:
    Oct 5, 2010
    Messages:
    1,517
    Likes Received:
    186
    Hmm, NOW it works... Not exactly sure why :p

    TyreRenderTest2.jpg
     
  12. Raintyre

    Raintyre Registered

    Joined:
    Jul 28, 2013
    Messages:
    226
    Likes Received:
    187
    The tyre tool is very interesting and sensitive. Every small change in the shape has a big effect on the contact surface.

    However, when doing deflection tests at QST section I miss something like a clock timer in seconds (or miliseconds). As you probably know, when the test starts and the ground surface is touching the tyre, the tyre mass goes from a initial "unstable" state (big loads, bounce up and down) to a final balanced situation where the deformed tyre rests on the surface applied. But some tyres take few time for the transformation, while other take much more time, even when they start with similar loads.

    It is difficult for me to know whether this "transformation speed" difference is due to tyre tool CPU processing requirements, or because it will be the true in-game behavior of the tyre. If tyre tool is slower transforming the tyre, does it mean this tyre has slower reactions on the track? How can I measure how quick is the tyre absorving loads? Is there a way to know how much 'real' time is passing by when running QST tests?
     

Share This Page