The online racing simulator
Progress Report December 2011
(551 posts, closed, started )
So Joe came here to disappoint us even more. The simple Pacejka formula took 20+ years to complete. Scawen wants to build a whole real-time system for this game to perfection and he is only 3 years into the project. What can we expect?
Well Pacejka is the Sheldon Cooper of tyre physics (theoretical), and Scawen is the Leonard of tyre physics(applied), so just because it took Pacejka took 20 years to invent/discover new tyre modelling system, does not mean anyone applying/implementing the same existing formula (be it Pacejka or finite model analysis) will take 20 years. This is because I assume that, like in any programming venture, we won't be discovering new theories or inventing theories, but we will be implementing existing theories. I presume that is what is happening in the case of LFS tire-modelling as well.

That said, it is not going to be straight forward either. But the main challenge I think, and as Scawen mentioned is to reuse/implement/apply the existing theories and formulae in such a way that neither speed nor accuracy are compromised too much for each other But will it take 2 years or more? may be yes!
Quote from Flame CZE :With all respect to the devs, I find this a bit weird. I thought that Eric could work on things no matter what Scawen was currently doing and how long it would take him to finish it. It looks like Eric was waiting for Scawen to finish his tyre physics until he could start working. So now, when the tyre physics seem to be nowhere near the release, Eric can make some S1/S2 updates for us. But yes, if he's working on the S3 content and doesn't like to switch from one thing to another, who knows, that might not be the best idea.

Tires are so important as the entire car (suspension, physics, wings and etc)...

To make a real car, you can do first the car then the tires, but to make a virtual car, you need the tires ready to make the car physics...

Also, Scawen said that Scirocco and Rockingham won't be released so quickly, because they found the tire problem, and he can't release a real track without realist physics...


Also... Scawen was thinking what the new tires would take just months to be released, thats why Eric did not made the tracks and cars who everyone wants...



Sorry for my bad english
Quote from Si Mclaren :Tires are so important as the entire car (suspension, physics, wings and etc)...

To make a real car, you can do first the car then the tires, but to make a virtual car, you need the tires ready to make the car physics...

Also, Scawen said that Scirocco and Rockingham won't be released so quickly, because they found the tire problem, and he can't release a real track without realist physics...


Also... Scawen was thinking what the new tires would take just months to be released, thats why Eric did not made the tracks and cars who everyone wants...



Sorry for my bad english

I totally understand this. But by S1 / S2 content updates, I meant things like car interiors or the 3D model updates in general.
Sorry for double posting
Quote from Kristi :So Joe came here to disappoint us even more. The simple Pacejka formula took 20+ years to complete. Scawen wants to build a whole real-time system for this game to perfection and he is only 3 years into the project. What can we expect?

Well, we are in 2011...

20 Years, let me think, its 1991...

In '91, the computers were not as good as at present, there was even much technology to make tires even in real life, so it must have taken so long ...

10 years between 90 and 2000 are the same as three years between 2008 and 2011, no?
Again, as I said in my response to Kristi, we cannot compare time taken to come up with a new thesis and the time taken to implement such thesis for example, if you are using Pacejka's formula, you need not re-do all the work that Pacejka did , because he already did it and we can use his deliverable in the form of his theories and formulae. But we have our own challenges in the programming realm. Let's hope it does n't take too long
Quote from Joe91 :Hi, I'm new to LFS,

Very well put.
I also toy around with simulated tyres in Matlab once in a while and I can see why it takes Scawen so long to come up with something really better than what we have now, but I would've struggled to put it so nicely. Again, well done.

By the way, I didn't realize before that the virtual tyre rig would allow the automated generation of (a large set of) Pacejka-constants. I wonder whether that's the route Scawen is taking.

I still firmly believe Scawen is devoting way too much time to issues we as players mostly don't even notice while he neglects issues as secondary that we very much notice, but that's his decision to make.

Vain
VAIN! Welcome back mate, nice to see you're still around!

greetz

der butz
Quote from maxis123 :
LFS -sim! ,, "" I.R, R.F..." - "sim"

Quote from E.Reiljans :lol iR is 100 times more sim than LFS

At least iRacing do the real racing, while at the same time LFS is up to jerky "cruising" on closed tracks. How sim is that?
A bird in a hand is worth two in a bush. Those two are in fetal stages if I am right
@Butz:
Many greetz to you too.

Vain
#338 - col
@Joe91
Scawen explains a bit about tyre dynamics, and also an outline of his implementation approach here.

I don't think he uses a Pacejka variant as is being suggested in this thread. I think that LFS feels so good when compared with other sims because it has a unique tyre model that is developed specifically to feel good when driven in a sim, rather than one that is tailored towards an academic or industrial context.

Scawen has admitted that if he had known how long it would take, they would have done things differently (here), but unfortunately, when you're breaking new ground, you will discover new unexpected problems - it goes with the territory.

I'm sure that no matter how frustrated any of us are about how long it's taking, Scawen is way more frustrated. Anyhow, I for one am looking forward to driving it when it's finally done.

Col
great news scawen,all the best for the new year
What is the difference between the pic from 2010 and the one from last week?
Attached images
1.jpg
2.jpg
Quote from BlakjeKaas :What is the difference between the pic from 2010 and the one from last week?

Not a very big difference if you talk about visuals. It's more how the tyre shape and the contact patch changes with various forces. But how can Scawen show us what he's been working on, when it's basically lots of code? Physics is nothing which can be easily shown on screenshots. Yes, probably video comparisons but that would be a waste of time, we should better wait for the first test release, so we can experience it in the sim itself.
Quote from BlakjeKaas :What is the difference between the pic from 2010 and the one from last week?

it is mirrored...
I have two theories:
1) code optimization consists in waiting until the hardware is at a level that can handle the calculations
2) in 2008 as a result of an unsuccessful experiment in the DEV's kitchen, was a small singularity (black hole) - as we know with great gravity, time runs slower, but only for people who are in the vicinity

ad. 1) better to use multi-threading and Physics
#344 - col
Quote from BlakjeKaas :What is the difference between the pic from 2010 and the one from last week?

gotta look closer than those tiny thumbnails you posted.

Here's a detail of the rubber mesh from the earlier version. You can see that there is no thickness to the rubber


Now, here's the latest version.

See how the mesh shows that there is thickness to the rubber. This means that the latest model is significantly more complex (and potentially more accurate and realistic) than the earlier one.

cheers

Col
Guys, don't over analyse the one off representative pictures of tyre wire-frame, that is just the tip of the iceberg of what may be going on.
Quote from col :@Joe91
Scawen explains a bit about tyre dynamics, and...
Col

Thanks for the link, I missed it somehow. I did some googleing on this brush tire model approach, it seems so much better then Pacejka! So, thats why LFS feels so realistic. God, I started thinking about this stuff again, downloaded some documentation I found interesting for later, instead of studying for my finals. I have to stay focused somehow!
Quote from repeat83 :u serious?
Do we need say WOW?! ))))

Actually this is a big job. Study and understand the material which the students told in lectures 1 - 2 years + write the code, numerous tests. I don't see anything funny.
-
([Audi TT]) DELETED by [Audi TT] : spam
-
(Tomba(FIN)) DELETED by Flame CZE : spam
-
([Audi TT]) DELETED by [Audi TT] : spam
I would like to know

how about the other things expect the tyre?
This thread is closed

Progress Report December 2011
(551 posts, closed, started )
FGED GREDG RDFGDR GSFDG