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

I said

"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'm trying to understand what the benefit of TGM is , I assumed it was to make up for a lack of real world data and the inability to test for more extreme situations. But like you are saying TGM obviously has to be based on something empirical otherwise you might as well use a theoretical model from the getgo.

I don't really understand how ISI's method all fits together ! lol

 

So your saying ISI has no input or knowledge from this testing regarding tires?

Race tires arent secrets kept by race teams as their cars engineering is, mostly tires these days are fixed spec for many categories and Im sure using the knowledge learned in developing realistic tire performance for professional simulations would be of great benefit for their own sim. I wouldn't be surprised if they had a database of all the tires they have simulated in the past and at least have lots of real world data and tire composition to work with. Its not like they just sell the software with an instruction manual and then completely abandon the customers, they will work with them if they need any help. They build the simulation and the customer just needs to input the data from their cars, I wouldn't be surprised if ISI made all the tires for the customers because its the most critical part. At least they have built the TGM tool which when imputing the real world tire data will give you the tires simulation and lookup tables used by the sim.

If they dont use real world data then what on earth are they meant to do seriously, they built the TGM tool and it will work in much the same way as fluid simulations do. They wont be 100% perfect but its a hell of a lot better way then getting some guy to look at an aero part and make an assumption of what it will do in terms of fluid dynamics.

Race teams dont use simulators to simulate tire performance they use them for car performance and setup. The teams arent involved in tire construction or development so they have to deal with what they are given.

ISI provides the tools so they can quite easily enter all the physical specifications of the cars and these will be modeled within the sim much like what they are doing with RF2.

This is your topic is it not "What I believe to be wrong with the RF2 tire model / grip levels / lack of control" This is an issue your having and not everyone else sorry to say you seem to be in a very loud minority here. Its early beta and its not perfect but its much better then your stated perfect example of NKP which treats all tires in pretty much the same way as what GPL and Nascar racing did back in the day. If this was so good why are all the new sims upgrading their tire models, iRacing, AC, RF2 I think you like the old model because you found it easier. Thats fine then go and load it up and drive but dont expect progressive simulation developers to take a step backwards and go back to what was the best solution in the 90s.

What are you basing your view on that the GTR tire is so wrong, and all the tires in RF2 to be honest and yet NKP is so right? is it reality or other sims you may have tried and liked. The fact that you like it or find it easier doesn't mean its realistic just that you like it, and going by your thread topic you find it has better grip levels and control as your struggling with this. Im just quoting your topic so that is your main point is it not. If that is not your point as I am obviously not getting it could you actually state what your point is. There is no point hiding it and I am not the only one here that apparently doesn't get your point.

What trade secrets? in your opinion they have no clue so who would want such lousy information?

TGM tool uses real world tire data in a detailed physical simulation, you give it the data and it will output the lookup tables based on these simulations. Maybe you should look up the word simulation to find out what it means, by the way sim is short for simulation.

Why did Red Bull and many other professional teams choose ISI for their software simulation? Obviously they have faith that ISI can produce an accurate simulation, they dont do this for fun like you and me its actually big business for them. They are spending millions of dollars and so will get the best option available to them. I must say I have more faith in the opinion of professional racing teams then any point you may raise here.

 

"Race tires arent secrets kept by race teams as their cars engineering is"

Everything I have read tells me that specific tire data is very hard to come by and evan when they have the tire data due to limitations in how you can test tires there are still holes in that data. ( for race car usage that is not basic road usage)

"Why did Red Bull and many other professional teams choose ISI for their software simulation? "

Becuse ISI's simulation allows them to plug in there own specific data and get acuret results from it within a specifc usage envalope. If that makes the car reolistc on the limit or not is separate to weather or not you can get meaningful test data for components before you test them on a real car.


"If they dont use real world data then what on earth are they meant to do seriously"

that was my question it seems for a consumer simulator ( intended for users to drive and race against each other , not for testing the effects of car components) it simply makes more sense to use a predominately theoretical tire modal than an empirical one.

This is because its more important to have a car that moves around realistically and consistently though its full spectrum of movement than one that gives out solid data or only moves realistically within a limited spectrum of movement.


I'm sorry but RF1 and RF2 are going to be drasticaly different to RFPRO after a Race team has pluged there data into it .

A race team will likely have a team working on one car and sorting out tire data and car physics for there specific model of there specific car and tire they use.

Thats why RFpro is so good because it allows teams to put there own data in and get usable results from it.

As to if RF2 is good as a consumer simulator is a separate issue to the results that can be achieved with RFpro.

 

GTR and F3.5 to me are less wanting to spin and less likely for back end to fly out under power when at moderate speed. The cars also seem more stable on curbing or if you have 1 wheel on grass. Cars seem to have more consistent grip across the track rather than just on the racing line.

Mind you the Cars feel like they have a lack of overall grip still ( evan though its more consistent as a whole) and we still have the low speed issue of not feeling the back end come out at all if you are already mid turn on slower corners.

Separate to all the physics the sounds on the GT , FISI and F3.5 are terrible the sound in RF2 is so bad in general its depressing, Maybe for me sound is more important than it is for others ?

probably placebo or some random set-up change lol

 

I tried Clio on slicks on wet tarmac... it is completely undriveable. I don't know if it changed with current build or was present in previous one.

 

Yes, slicks don't work well on wet road - you need to put on wet tyres.
When road starts to dry, you can feel the change in grip as you go from wet to dry part of the road.

I tried Spark F2 and it also has more grip on dry part of the road so it seems like this has changed since last build and wet weather makes much more sense now.

 

+1, this thread is getting insane even for me lolol

 

lol

 

OMG I'm choking...lol

 

Lol love the slicks on wet roads what a surprise, I havent tried the new build but in 101 the road was still drying too fast and with a full ai field you could already get on the slicks pretty easily on a damp track. Its actually quite fun as you can almost do dry laptimes but if you run one wheel onto the very wet stuff you will have no chance. I was crashing at about 20kmh as the track got soaked mid lap at mountain forest. The ai really sucked in the wet and were much slower then i could go maybe they were always on the wrong tires but i see them pit come out and still slow. Even on the wets they did struggle pretty bad being maybe 7 seconds or more slower then i could do.

Probably they change to wets very late and as the track dries so quick half a lap in they could easily be on slicks already and stay out burning up wets.

 

James I am pretty sure that the Pirelli F1 tire would have the most data available for any tire ever made, when Pirelli moved into F1 the FIA gave them specific design orders for tire performance. The way I see that they would have most likely done this as with most engineering these days is used a simulation model first and try and get this to those specifications later with the real world tests trying to match up to their simulation models. Seeing as they produced a tire almost exactly to within the set design parameters It shouldnt be all that hard to simulate this tire.

How much of a secret is this though thats the big question, do Pirelli share all the tire data with the f1 teams? would they share this with simulation developers? They would have run 1000's of test laps while developing the tires and I think they had 2 full time F1 drivers testing these tires whilst in development. They would have testes both real world and simulations and I think they would have got as close as possible between the 2. This would have taught them a great deal about tire construction, even though a tire manufacturer would know a lot they dont always get so much real world data to work with. They would then be forced to use simulations for most of their internal testing with quite limited real world testing.

Most categories around the world now do things in a similar way if they have a set tire for the category. The tire manufacturer and race teams would always be sharing lots of data to help out future tire development to keep them improving within the specs set by the series directors.

All this work between tire manufacturers and race teams directly sharing data will give them a much better understanding of tire construction and likely performance. This data unfortunatly will be very hard for most mod teams to get their hands on but it looks to me like ISI is using this large amount of physical data on tires in their TGM simulation. If this provides accurate results then it surely must be correct, however like I said very early in the post the more accurate the real world simulation model will be the more likely you will feel limitations due to your hardware setup. This is where a 100k motion sim and a 30K FFB wheel would make a huge difference.

The big question is how close to reality do they want to go and how much do they fudge the data in order to make the cars drivable with our limited hardware. I think going a very true simulation model is the way to go but then this means that not only do the tires need to be simulated correctly but every part of the car and also the track surface must have realistic properties.

The best tire model in the world will be flawed by the weakest link in the chain, and no matter how good a simulation is it will never account for all the variables in real world situations. I think as modders will find it very hard to get data on tire construction which is needed for a good TGM model. There should be a point where ISI stop making cars and just focus on tires for various categories if they have access to any of the real world data. Even if they make example tires of most of the general construction materials to give the modders some good bases to start with.

I think modders as do the race teams with RFpro will have a much easier job entering the physical data of their cars, but they wont have access to all the tire data either and so would need to work with the tire manufacturers instead. I really cant see the RF pro customers needing to develop the tire model and that would in my eyes be provided to them by ISI. This is the absolute core of the simulation and if this is wrong then it makes no difference about how accurate the car simulation is.

This will also mean that there will be much more recent tire data say within the last 5 years and anything beyond that if constructed quite differently to current tires will not have enough data around possibly.

Here is a quick little video showing some very basic info on F1 tires by Pirelli, Im sure more like this could be found on the net.

 

You should try it rather than laughing.
I wanted to say, that this effect is overdone. Maybe I should add, 'undriveable even at low speeds'

 

I'm afraid there is a fundamental flaw in the physics of the universe:

http://hyperphysics.phy-astr.gsu.edu/hbase/frict2.html

Abrupt transition from static friction to kinetic (and vice versa) prevents true racing simulations from having smooth car handling transition from stable to unstable state and back. If ISI doesn't immediately start to develop a FTM (Fake Tire Model) I'm forced to ask for refund!

 

Why not? Probably they can use ISI base if they want, but with the money they invest on engineers (F1 teams for example), I don't see the problem of developing their own tire models.

 

Sorry I was not laughing at you...but the reply was nice!

I've already asked Tim what's is changed on tires...so I know something feels different than before.

 

Actually I was just confirming that you cannot drive on the wet with slicks
I believe it was more less ok to do it in previous builds. Some people were even reporting better lap times on wet road

How can you tell it's overdone? Have you ever driven a car in the rain on slick tyres?

I think you're getting some weird impressions here. I took a Clio for a short ride yesterday and I haven't spinned once on my outlap in heavy rain. I was driving on slicks and my cornering speed was around 60 km/h in VIP corner and around 90 km/h in Parabolica. On dry track these speeds are 85 and 130.

Maybe lack of speed perception lead you into believing you were driving much slower than you actually were?

It reminds me of people in our leagues complaining that cars break too easily, even when they "just barely bump the wall", while in fact they were hitting a wall at 40 km/h and 45 degrees. In reality that would wreck your car.

 

Yet the tires were still all over the place with teams not knowing if they would last or not. So evan if the teams were given data for the tires beforhand it was not enough for them to be able to know what would happen until they got the car onto the track and tested it in reality.

Obviously there are so many verables in real life it makes it near imposable to know for sure what's going to happen on the track day with varouse temperatures car set-ups and everything else.


The piont I'm making is that the ways to test race tires and real race loading and situations of tires are limmited.

To the best of my understanding you can get very usable results for a good spectrum of operation of a tire but the models they use become less and less reliable for predictions of what will happen with the more exstreem usages of the tire.


Another point separate to the issues I perceive with RF2s simulation, if you leve that aside the sound in RF2 is terrable its so bad it makes me not want to drive.

Physics is what I am after from a simulator but in terms of fooling my fleshy brain into believing I am sat in a real car you might have the best physics in the world but if it sounds terrible then its going to totally brake the immersion.


I agree home hardware is totally limited and that is a massive limmitation for driving simulators


Like I said before I think Rf2 will become more progressive with its grip and car stability as the game moves through development. ( by that i dont mean the cars are less twitchy they should still move around the same its just how much control/ability to change that movement the user has that's important)

I do worry though that even if RF2 does end up with the best physics, its awful sound , bad UI , clumsy mod system and general early 2000s attitude to software development will probably push more people over to other simulators like AC

Maybe ISI is simply better at making professional simulators ?