Hi, I'm new to LFS, got a wheel just a few days ago, and after some time I spent with the demo, I went for the S2. Then I read a lot of topics here on the forum, and read the old progress reports by the devs, and now I have a basic idea about what is going on in the community.
I am a vehicle engineer student (haven't learned about this kinda stuff yet officially), and from the age of 14, I have a returning desire not to just play, but also to develope a racing simulator. After I took up some programming skills, I tried it several times, made some actually driveable, but not too interesting applications. I did some research on this topic, and I would like to share some info with you.
Actually it doesn't really take a lot of time, and doesn't really require any special skills to write a program, where you can drive around in some environment. There are graphics engines and physics engines to use, you just have to set them up correctly. The collision detection and general rigid body physics in those engines are pretty realistic, and physically correct, they give satisfactory results. It's not a really big deal to add some kinda suspension and wheels to a rigid box, better physics engines even have some built in functionality to do this kinda stuff, and then you can drive around in your own game happily.
Problems come, when you want some realism. It's kinda straight-forward to create realistic suspension and drivetrain models. The problem, as you can possibly guess, are tires. As you know, tires are not rigid objects, the surface contacting the road changes according to internal pressure, the actual load on the tire, the tire's physical parameters, like size, rigidity of the sides, the friction depends on road surface, tire threading, etc. etc. I can easily cause myself a headache just by thinking about it's complexity! It's really, god damn hard to describe them mathematically.
One of the actual solutions of this problem is the so-called Pacejka magic formula. Hans B. Pacejka is a professor on some university, and his way to model tire forces is really awesome at the firs sight. He collected a large amount of data about actual tires, registered the forces produced by the tires, and then, fitted some functions on these datas. His formulas usually contain some input parametes, like slip angle, and load, and some „magical constants”, and they return the forces produced by the tire. It's easy to code, and fast to execute. As far as I know, it's pretty much an industry standard in simulation. The problem is, that these constants have no physical meaning at all, the model is empirical, so you need a load of real life data if you want the car in your sim to act accurately. If you want to change something, for example, want to change the tire pressure in the sim, you have to get a tire, set it's pressure, and test it to get the correct value of the magic constants. Other downside is, that below some speed, they are not accurate, because in the formulas, you divide with the speed, and when it decreases, the resulted force gets unrealistically big. There are solutions to work around this problem, but after some reading, none of them looked very realistic for me. If you want to see Pacejka in work, check http://racer.nl
The other approach would be some finite element analysis stuff. The problem is, that when we compare it to Pacejka, it's ridiculously complicated, and process intensive. They require a large amount of computation power, and a lot of knowledge about tire structure to do it. They are mostly used in non-real time applications. I guess the non-real time model what Scawen is usually talking about works this way.
What I think Scawen and the sim developers of these days are possibly doing (not sure, I found available , detailed stuff only about Pacejka) is finding some crossover method between the two models described above. Pacejka published the first version of his formulas in the late 80s, and the last came out, if I remember correctly, in 2006. I think Scawen is actually quite fast, LFS was published in..2003? It's only 8-9 years of developement, and he started the new tire model about 3 years ago...
I really hope that this made any sense, and I didn't write too much stupidity. Oh, and I hope I didn't make a large amount of mistakes, I rarely use my English for anything else then reading and listening.