An observation

Ricky Law

Ricky Law

Registered
I have been participating in the official online racing almost every day this winter. With the Tatuus I am competitive for a top 5 position most races. In order to improve, I have watched replays of my races to see what the top tier drivers are doing differently. The first thing that I noticed was the amount of front wheel slip angle showing up on tight corners, enough under-steer to cook rubber. Yet they seem to be paying no price for it in time or tire wear. I could be wrong. What I suspect is that the drivers have cranked up the steering sensitivity in game to high levels( I've tried it and it seems to work ) to improve turn in before the apex. This seems to point to a flaw in tire physics model. Any suggestions?
 
You are correct in your observation. Many top drivers use a short wheel range in their settings (or low steering ratio), almost to the point where it resembles the steering inputs in karting. How aggressive you can go depends a bit on the tyre model and what wheel you have.

It's known that the game's tyre models are not the greatest at punishing high slip usage. I'm not sure if we have an exact answer to what's causing it, but we can have a lot of informed guesses based on data gathering. My own thoughts rely heavily on the lack of dynamic grip loss/gain from the surface temperature. The slip curve in-game might look very similar to IRL, but the game doesn't really punish the extra energy you put into the tyre from the high slip angle. In an IRL scenario you'd cause severe understeer. From what I've seen, iRacing is fairly decent at simulating this.

The high slip angle usage can also come from a tyre construction that's too soft. In the picture below you can see such an example where the peak slip angle migrates a lot with vertical load. Of course, with grip loss from a too high surface temperature you'd be able to see a drop in Mu at the high slip angles, but since it's not really a factor in the game you can drive at those angles just fine. This also explains using a low steering ratio. When the build-up curve is so slow, you'd have to turn the wheel quite a lot with a normal ratio applied. Having a lower ratio helps you reach the peak grip point faster, which is generally better for lap time.

1768945813797.png
 
Thank you Robin for your reply. You have confirmed my suspicion. I remember seeing similar tire flagellation in GTL, and wondering why the drivers paid no price.
 
Obviously as OP suggests there should be something about the tire wear, if not anything about handling physics that could make tires snap and bite more. Maybe races are short, and people can get away with silly things. Maybe it could be realistic up to a degree (abuse of understeer, not super ultra quick steering racks).

Well there is a lot that can be done with tgm realtime section. I am learning to control the later tyre model where you can adjust tread and belt stiffness base + per pressure + per temperature + per rotational speed. So much can be taken under control there. Then obviously the way tyres heat up and cool, and how that results in grip levels. Tyre wear. Tyre micro, macro and adhesive sliding grip multipliers per speed. The damping parameters. Obviously tire rubber sensitivity. Last but not least plain old simple nominal sliding and static friciton parameters. And little bit more.

The whole rest of physics absolutely takes a good role too. High Definition Vehicle file, Ultra Chassis file , Engine file.

I have had all sorts of tires, and I had some tires that were insanely punishing any wrong move. So what capable developers make is by decisions. Not because they can't do better.

Technically I think most cars IRL could be set up to be drivable like that. Just numbed out front end in some ways, mechanically, aerodynamically or both.

I had for a moment hyper sensitive front end on my Nissan R90CK. Cured it by increasing reduction of tread stiffness at high rotational speed. Later on had to readjust how tread stiffness changed due to increase of pressures, as tire suddenly started acting very weird as tires got up to 290-300kpa.

In my Porsche 917K physics project I couldn't get it behaving right at medium to high speed till I increased front end aero sensitivity to ride height, as it got little bit more floaty at front at higher speeds it balanced out well. Interestingly it made it corner better. Smoother.

There has to be other limiations IRL. Steering could become unreasonably heavy and rough, and that would greatly reduce cars controllability on the edge. Simracers can just detune forcefeedback without any compromise.

Again, all the rich magic beauty of race car dynamics gets reduced down to little insignificant non-brainer torque profile and magnitude of the rotating thing on the desk.... the holy device we have to worship after all.
 
Is this an issue that can be solved via correction of the tire or it needs to be adressed at a deeper dev level?

�
 
You must log in or register to reply here.
Back
Top