As some of you know, I have been thinking of developing an application that takes into account factors such as tyre temperatures, fuel load, tyre wear and so on to accurately predict the potential lap time for lap "n" in a race. From there it would be easy to add a feature to calculate the optimum strategy (when to pit, how much fuel to put in, weather to change tyres and so on). Tyre temperatures is one of the biggest problems and creating a projection of when they will be at what temperature is not so difficult but calculating how this affects laptime would be an immense project. If I could calculate the lateral and longitudinal grip of a specific type of tyre at given temperatures and I would be able to project the tyre temperatures for lap "n", then given the amount of fuel left in the car, calculate how that affects acceleration in a straight line (BobSmith's torque curves for cars and the calculation for drag/lift could be used to calculate the acceleration on the straights) and I would be able to use the track maps to create a laptime prediction assuming (constant speed turning). This laptime prediction would not be at the full potential of the car but then it can be used to calculate it as a percentage of the potential laptime of the car on lap 1. Then you would multiply that percentage by a user entered laptime for lap 1 and you should get a pretty accurate laptime prediction. From there it shouldn't be too difficult. Does anyone know where I can get information regarding this subject?

EDIT: If it is necessary to calculate the camber of the tyre to the road at a given cornering speed and g-force (and this is getting complicated here, because we now have live camber), and also calculate how the tyre has worn in each part (Inner, Middle and Outer) and therefore calculate its potential grip I would be willing to. It would actually be a smaller project than trying to estimate laptimes out of thin air. This isn't getting any prettier though, because I would basically have to have most of the physics in the game taken into account in order to do this and I won't be able to without Scawen's equations.

It's probably changing as S2 gets updated, and unless the developers are publishing the formulas they use (not likely), your best bet is to get on the skid pad and note the g forces as the tires heat up while turning.

oooo something for Bob to answer .. i bet he's lurking around here somewhere....

If you check GRC you'll see I've already added acceleration testing... so yes it is possible with my data.

I also want to add weight transfer and tyre load sensitivity, neither are particularly difficult. Working out how temperatures affect grip is really simple, just requires some effort:

Pick one car, and use this throughout. Go to the skidpad/autocross arena and see what G you can pull with the tyres immediately. Do a quick burnout, try and get both front and rears up by 10 degrees. Repeat. Then another 10 degrees. Repeat. Then repeat the whole process. You could also just let the car sit there and the tyres cools, and get the colder temperatures.

Just put all this data into a table in Excel and plot the results on a graph. Fit an equation to the graph, and voila.

Yeah but the problem is that there are Inner, Middle and Outer to think about and its kinda difficult to get each part up as much as you want it to go up. I want this thing to be pretty accurate.

If you want to work with I/M/O temperatures, then you're gonna need to calculate tyre loads on those three sections as well, which means you'll need to know camber as well. Good luck.

Start simple, then add increasing complexity. It's the only way you can sensibly work (unless you're going something really easy in the first place), LFS is done the same way. So start with an average tyre temp since the loads are easy to work out, and ignore camber FOR NOW.

I guess that's what I'll do, but IF (quite a big if) Scawen was kind enough to give me the equations it wouldn't be that difficult to do it the complex way I guess.

Isn't it better to take the car out for 20 minutes testing rather than do some weird calculations, which can't predict human mistakes and wind strenght/direction?

Just my opinion, but I doubt anyone who would sign up for 1 to 3 hour race would calculate his grip and strategy like this, without any real testing.

There's a huge difference between how people race. Some racers can drive their set for 30 - 40 minutes, other people using the same set can kill their tyres in few laps.

That's what has to be calculated - how quickly your performance drops off so that you can calculate estimate laptimes and therefore the strategy...

You propably have to make your own data with your own setups. I think everyone has personal driving style. So how fast you wear out your tyres and what is your average fuel consumption depends on your driving style and your setups.

Instead of calculating the lap times based on the results. It would probably be best to calculate how much off your normal lap time you should expect to achieve. e.g. if your average lap time is 1:40:50 then it would say +03:40 if you have bad tyres and high fuel etc instead of saying 1:43:50 because everyone has their own driving style and other people are "better" or "worse" than the average.

That's the way im thinking of doing it :
[potential lap time of the car currently]/[potential lap time of the car on average] * [racer's average laptime]