What I believe to be wrong with the RF2 tire model / grip levels / lack of control

You must distinguish between full gas based slides (at low speeds or at low lateral G forces) and grip lost when G forces are high and tires are in the limit of "lateral grip"...

 

Yes, here it's "powerslides", but it's a good movie

 

And example then:

And your response:

Another example - your comment to a video where you can't tell if driver is actually using any throttle (no sound, slow motion, external camera, low quality) :

You're saying there is throttle or that there is no throttle as it suits you, regardless of what's actually in the video.

Unless there's another hiohaa on this forum, you clearily have trouble recognising simple facts in video clips. That's why I said you can't expect people to take your arguments seriously - you're making a poor job putting them through.

 

That video does not show what you think it shows, IMO. To me, it shows someone who has gotten much smoother in 2 years. There's a lot more progressive throttle on the 2011 video, and not nearly as much sawing at the wheel, either. If the setup is the same in both videos, then I'd say if anything, the NTM induces a bit more understeer in that particular car. As RedBull or any other F1 team around from the Bridgestone to Pirelli (or even from Michelin to Bridgestone) can tell you, a difference in tires can drastically change the balance in a car (given the same setup). Which is essentially what appears to be happening here when going from the OTM to NTM.

My own experience with the Miata going from OTM to NTM was that the NTM seemed to break away much quicker and just kept sliding. You could REALLY chuck that Miata with the old model and remain in control. When they changed to the NTM, I definitely had to be a lot easier on the steering--which is exactly what your video shows as well. And that sounds very much to be the opposite point to what you and James have been trying to say this whole time.

 

Lol, he probably tried both cars at the same time. 2009 it's just the year of the car, not the date of the video. And trust me, it's not the driver skill, it's just the way to drive both tire models.

 

Anyway, It seems like blakboks said.

 

Well, they only list one Nationwide car on their website. When I did iracing, I primarily stuck to the road cars, though. So maybe they have two spec cars, but going by their website and that video, it certainly seemed like same car, different years. If you have iracing id be curious to see a screenshot showing both spec of cars. In either case, I agree that you have to drive the NTM differently, and even said so in that post as well. My point was that you have to drive the NTM more like rF2 than the OTM--which was the opposite of what hiohaa had intended to show, if I understood correctly.

 

sorry for bring back this again but i cannot get tired of seing your video in youtube. You must put this video in your signature.
I just can´t get this light in my settings and i think its because my monitor settings. Is there a site teaching how to ajudst brightness, contrast, and other thing? Change my view about graphics in rf2 in a good way after your video

 

James Im basing my views on real driving of a few different types of racecars, it seems you are basing yours on NKP and all old sims which always felt much more like a game because of such progressive slip angles. Every car and tire is different in real life and so you cant expect a rally car to be the same as a race prepared gt1 car. RF2 is heading in the right direction and the one whos mind will be blown most likely is yours as sims head away from the NKP and gpl type physics.

You have asked for videos and they have been posted, would you be able to put one up with you actually doing a good pace. You did ask for this as some form of validation for others to be able to question your views. Yet I havent seen anything posted by you that on tbe same basis for you asking others justifies your views.

RF2 isnt perfect but its finally going away from that gpl or nkp flawed idea of progressive control on racing slicks, this is easily done on snow or ice. Even on slicks if you have a very skilled driver but at well below race speeds. There are some cars which are under tired or have different tires where you will need to drive at pace with a bit of angle but dont expect this in all.

Iracing started a NTM because they bad real drivers using the sim that had no connection to the old model brought along since gpl. They are heading in the right direction and if AC wants to be considered a sim they will also need to go away from their old model.

You wonder where ISI gets the data? Try all the pro teams they work with in RFpro they cant give us the full physics of the red bull f1 car but can base it off that and other cars they have worked with.

I wish some people will realize that not all cars are driven the same way and different rubber gives different results. Race slick that arent fully shredded are the least progressive and can be quite snappy at the point it regains traction. To a road tire that is much more progressive and will not stick as well as the race slick so at pace will still have plenty of sideways angle.

 

"You have asked for videos and they have been posted, would you be able to put one up with you actually doing a good pace. "

Looks like you don't bother reading other peoples comments

"You wonder where ISI gets the data? Try all the pro teams they work with in RFpro they cant give us the full physics of the red bull f1 car but can base it off that and other cars they have worked with. "

F1 teams develop there own tire model and ISI's engin is more about driver in the loop testing rather than raw car feel I'm sure RFPRO and RF2 would also be allot better with a $15,000-$30,000 FFB wheel and a team of software enganears employed to develop a single car.

"I wish some people will realize that not all cars are driven the same way and different rubber gives different results. Race slick that arent fully shredded are the least progressive and can be quite snappy at the point it regains traction. "

Have you considered that other people actually do get it but your totally oblivious to the crux of the conversation?

Evan Feels 3 thinks there is something up with the GT car and he loves RF2 physics. ISI are still working on the tire model. I also agree with Feels 3 in that Allot of it is going to be car specific but unlike feels 3 i would say there is also a general issue with the tire model in general.


from what I understand ISI use Pacejka model in RF2 , the big issue with Pacejka is the following


Stefano Casillo: Pacejka is a monster of something like 100+ parameters, every parameter can potentially screw something up, and realistically, you never have the raw data available to even try to get those 100+ parameters.

To fix this I tried to build a tyre model that has a physical based theory in the formulas and that can be tweaked with very few parameters that cannot create anything “destructive.” This involved moving away from that particular pacejka model. The basic pacejka “magic formula” is still valid to create basic curves, but a good tyre model needs to be more dynamic. The model we see in netKar Pro V1.3, Ferrari Virtual Academy (FVA), and AC are all based on the idea of generating a “base” slip curve that is then adapted to the behaviour of the tyre. I call it “slip profile,” in netKar Pro and FVA this adaptation is done in steady state analysis, in AC I added the dynamic of the carcass and thread to the equations.

This not only makes the tyre more dynamic when put in extreme situations, like riding kerbs and so on, but also makes a big difference on flat roads, in the way it reacts to load transfer when you turn the steering wheel. I makes it feel… Rubbery.

Maybe its simply the case that ISI don't have enough data for tires in the first place ? that would explain why the physics in RF2 go crazy when apreaching the limits of the tires.

NKP moved away from Pacejka in the last build and nether FVA or AC use Pacejka simply because its very limited when it comes to the more exstreem tire situations IE when driving on the limmit.

The only exception is if you have a huge amount tire data available and you have correctly captured that tire data.

Now I assume that for F1 teams that hire a a team of people to simply get tire data and properly implement it into the simulator along with everything else and probably have relationships with the tire companies this is less of an issue

 

In short


1) An empirical model only works if you have the correct data to start with.

2) Theoretical models will produce a more realistic tire when you don't have all the data to hand for tires


Evan with a solid empirical model It will be limited by the resolution of its implementation and from that inconsistencies might arise where as a Theoretical model will always remain consistent . Granted a floored theoretical model will be consistently floored

You can see why ISI have to use an empirical model for there simulator its because for there whole "driver in the loop testing" you need an empirical model in order to give back specific tire temperature engine ware and specific data for teams to act apon. However again this will all fall back on the quality of the teams tire model and specific car model.

 

Pacejka was in rF1, I thought they abandoned Pacejka in rF2.

 

From what I understand they use Pacejka but pre render all the given values with an advance system that models a virtual tire this then bakes the data into a file that can be looked up in real time.

I think the specific reason they made the changes they did with RF2 and the way the tires data is gathered is to try and bridge the gap between the fact that its near imposable to get the required tire data from manufacturers of the tires.

I could be wrong but that's the impression I get.

I would like to here it from the horses mouth

 

Thats fine As I said as far as I understood they used Pacejka If they don't I still believe what I said before is the case Pacejka or not ISI's engine is built on An empirical model ( lol maybe I'm wrong about that as well ? I wish there was more information if anyone has it please link! )

 

An excerpt from the Rf2 TGM Tyre Tool Manual

"The TGM is in essence a brush model, using a 6-DOF rigid ring to simulate frequency modes. A highly sophisticated model is tested over a wide range of conditions to supply adequate information to generate a detailed lookup table. This resultant data is able to be processed quickly enough for the real time model to sufficiently run in a real time environment at high sampling rate, providing accurate behavioural results."

 

If I'm not reading into it wrong . That seems to confirm what I was saying in that the TGM tool is basically a replacement for the lack of real world tire data. But then even if you are getting high resolution data from that into a lookup table , how do you know that the TGM tire behaves realistically and what is the TGM itself based on ?

ISI should have a month off from RF2 and just make a documentary about tire models and simulators and give away all there trade secrets