Solid rear axle trouble

Discussion in 'Car Modding' started by mantasisg, Feb 8, 2021.

  1. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,425
    Likes Received:
    3,284
    Hi, I am here with hope to get some help and advice.

    I have quite a lot cars froms 50s and some cars from 60s with solid axle geometries and I face the issues of compliance in all of them. Basically the issue is that solid axle seems to significantly reduce roll and pitch of the chassis comparing to normal both axles independent geometry. I suppose that fully independent geometry should be more compliant, as wheels will be less restricted (thus independent), but I experience solid rear axle limiting compliance way more than I would think is believable. Besides solid axle, this issue is also massively affectign dedion geometries that also has to be simulated by using solid rear axle geometry.

    I thought I had found the solution for a while and was very happy about it. I could achieve proper amount of body roll and pitch by altering stiffness of rear wheel balljoints, I thought it is good workaround. But recently a massive drawback was discovered, which is that weakening these joints mess up how rideheights gets initiated in game and also wheels themsleves gain some forward shift during accelerations. Now that I know these things, I can not be happy with this particular solution anymore, although in terms of car handling it is alright as long as you don't mind that rideheights gets messed up and mismatch with AI rideheights happens, and you also can't feel the wheels shifting forward during accelerations, can only be seen realtime external looking at player controlled car because otherwise rF2 displays .pm geometry.

    I have looked at Boxmaster, it generally doesn't have these issues because it is simply stiff. The other vehicle is interceptor quad bike which is totally broken, it also displays massive wheels shifting forward during acceleration and is all broken in many ways. Gasoline Alley mod seems not to be using ultrachassis, so thats not a good sign, I suppose they have chosen that as a workaround, which I would not like to do...

    Is there anyone who has good experience with solid rear axle geometry for rF2 ? Do I not understand something, or rF2 solid rear axle has underlying issues ?

    Would be awesome ifthere is something I just don't understand, because frankly I hardly expect S397 doing anything for clutch and drivetrains, let alone even thinking about solid rear axles...
     
    Last edited: Feb 8, 2021
  2. Rastas

    Rastas Registered

    Joined:
    Feb 17, 2017
    Messages:
    529
    Likes Received:
    1,136
    In my AE86 mod i use a solid rear axle,and is great actualy,altough is still not perfect,i have used the online tools by Chris to convert the pm from RF1 to ultra chassis,and after a few tweaks is working great :)
    Here is a video in a japanese touge :)
     
    Corti likes this.
  3. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,425
    Likes Received:
    3,284
    Do you pay attention to body roll, pitch ? When you need to achieve proper chassis kinematics it becomes more challenging, especially in case of very compliant car. Your car looks very stiff, perhaps as it should, therefore you are not getting significantly exposed to the issue that I have. Also have checked how geometry looks in real time ?
     
    Last edited: Feb 8, 2021
  4. McFlex

    McFlex Registered

    Joined:
    Feb 23, 2012
    Messages:
    863
    Likes Received:
    249
    How do you guys modell the solid axle? Do you use the physics spreadsheet?
     
  5. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,425
    Likes Received:
    3,284
    To begin with I used Boxmaster for three link and Interceptor Quad for four link geometry as a refference and then tweak stuff with ptool. Haven't used spreadsheet.
     
  6. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    9,606
    Likes Received:
    4,618
    I would say make an equivalent .pm version in gasoline alley style, check that it seems to behave the way you want, then fully stiffen your flex chassis and compare the behaviour.

    I haven't tried this myself, and admittedly haven't carefully checked how my solid axles are behaving - but they're also very different cars to your 50s and 60s wallowers :)
     
  7. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,425
    Likes Received:
    3,284
    I'll keep it in mind, it sound like good method to evaluate if there is an issue, but also what if there is same issue going on with PM file, anyway, will test some time.

    I was also thinking on some other method. That I actually have pretty much set, but haven't done properly. I think pretty good demonstration would be to have everything in cars physics identical, just change rear axle geometry - one with independent rear geometry, other with live axle.

    Anyway to my limited understanding live axle should only restrict rear wheels position so they are always in parallel, I can't get it how it should restrict body roll and pitch. De Dion tube should be same like live axle just little more flex in the tube and less unsprung weight.
     
  8. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,425
    Likes Received:
    3,284
    Hi there again.... I don't suppose there is any significance about the issue if some modder experiences it. So I have been thinking what could make devs to be interested in it ? For a while I haven't got an idea. But recently I just realized that the go-karts could be it.

    Correct simulation of solid axles is important for kartsim, right ? I heard stiffness of rear axle is important part of kart setup. Kartsim guys should understand it very well.
     
  9. Raintyre

    Raintyre Registered

    Joined:
    Jul 28, 2013
    Messages:
    200
    Likes Received:
    155
    You can use a rear solid axle, but I think the difficult thing is to add flexibility to that chassis. I personally didn't achieve it and still didn't see one working in Rfactor 2.
    I am not saying it doesn't exist ..
     
  10. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,425
    Likes Received:
    3,284
    I found one way to add complience to solid rear axle chassis, and was happy for a while unitill some unwanted sideefects were discovered. It kinda shot down mine 1954 cars physics update on which I worked a lot. What I did to get desired chassis kinematics, was essentially reducing stiffness of rear wheel hinges. The bad effects ofthat was that wheels would gain shift forwards with acceleration, and even worse - would loose default rideheight, because apparently game initiates car in game by using pm geometry, and then shifts to ultrachassis, and some mismatches appear due to chassis flex parameters...
     
  11. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    9,606
    Likes Received:
    4,618
    Dragging this up as @mantasisg is making this a theme in recent posts so I'd like to get to the bottom of it.

    Calculating roll from basic vehicle properties isn't a massively complex task, and personally I've had what appeared to be correct results (certainly within an error of margin as good as the subjective observations above) so if anyone can offer some concrete examples I'd like to know about it.

    Perhaps we should make a simple test vehicle to investigate?
     
    Emery likes this.
  12. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,425
    Likes Received:
    3,284
    I've had over 30 cars which needed solid axle simulation. I couldn't achieve what I needed in most if not all of them. To name an issues basically, couldn't achieve compliance such as observed in real life evidences, as well as not achieving right handling (which obviously goes together with chassis compliance).

    I am not sure how that connects to calculations which Lazza did, which he says had correct results. I would certainly be happy if my troubles were all related to some particular lack of knowledge or a mistake, which I could simply improve myself, I could never find it yet.

    @Lazza That is nice idea to have a test vehicle to investigate. I'd perhaps even suggest having two or three different cars. For example one with small tracks and wheelbase and smaller size overall, other with large tracks and wheelbase and bigger overall. I could offer few of cars that I have worked on from 1954 or/and 1967 endurance packs by created by woochoo. For example big car could be 1967 Mustang, small car could be a 1967 Austin Healey.

    [​IMG]
    [​IMG]

    Or just simply take them and try to achieve realistic solid axle levels of roll and pitch, just to see that can work like here:
    [​IMG]
    or like here:
    [​IMG]
    or achieve it to move like this:
    [​IMG]

    To be honest more difficult to find pictures with excessive dive for solid axle, but I sure know some examples for cars with De Dion:
    [​IMG]
    D50 by the way was spectacular:
    [​IMG]
     
  13. davehenrie

    davehenrie Registered

    Joined:
    Jul 6, 2016
    Messages:
    4,757
    Likes Received:
    2,662
    How about a Hinged De Dion?
    [​IMG]
     
  14. Emery

    Emery Registered

    Joined:
    Oct 24, 2010
    Messages:
    2,826
    Likes Received:
    1,422
    Or the flexi De Dion of the Maserati Tipo 63, 64, & 65?
     
  15. Emery

    Emery Registered

    Joined:
    Oct 24, 2010
    Messages:
    2,826
    Likes Received:
    1,422
    In real life, they have shockingly soft springs and poor front wishbone angles, just enough front swaybar to keep the inside rear wheel on the ground. But I haven't tinkered with it in rF.
     
  16. doddynco

    doddynco Registered

    Joined:
    Dec 13, 2017
    Messages:
    277
    Likes Received:
    466
    Could you please share some ptool screenshots of the attempted solid rear axle to see what you're dealing with?
     
  17. Jokeri

    Jokeri Registered

    Joined:
    May 17, 2014
    Messages:
    195
    Likes Received:
    31
  18. mantasisg

    mantasisg Registered

    Joined:
    Aug 17, 2016
    Messages:
    2,425
    Likes Received:
    3,284
    I tried everything I could think of. Including very soft springs, dampers, anti roll bars all at once. I also even tried unrealistic high CG values, and unrealistically low instantaneous centers just to see what will happen. As far as I remember nothing worked to achieve what I needed till I made wheel hinges to be very soft, but thats of course not the way to go, I am almost sure issue comes from ultrachassis, and probably only is noticeable or in effect when cars are soft and has (should have) large ranges of motion. I don't remember everything in fine detail, I hope to come back to working on physics of these cars soon, and I'll report better.

    Yes, I can. I'll do it when I'll come back to modding, hopefully soon. I am happy for your interest guys, perhaps it is solvable, although it is very likely an rF2 issue.
     
    Emery likes this.
  19. John R Denman

    John R Denman Registered

    Joined:
    Mar 18, 2018
    Messages:
    443
    Likes Received:
    497
    Geometry is everything when forces and vectors are at play.
    Anti-Dive, Anti-Squat are just one plane (longitude) where geometry that transfers horizontal axis into vertical axis loads.
    Anti-Roll geometry is just as effective at shifting axis's.
    Look at the back view.
    If the outer pickup point at the upright on the lateral plane is lower than the inner pickup point of the chassis, the strut will oppose jounce at a rate equal to the sin of the angle relative to the horizon. Anti-Roll.

    The lower strut will have the most profound control of Anti-Roll. Upper strut geometry is designed to control the arc motion after identifying the lower strut range needs.
     
  20. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    9,606
    Likes Received:
    4,618
    To help place this in an rF2 context, can you explain this a little more? Diagrams, references? I think your use of "strut" is a bit unusual.
     

Share This Page