09.09.2024 Category: Site news
There is probably very little understanding why tires are such hard things to get the feeling spot on. So, try to absorb the below mini 'lecture' on R/C car tires and a physics tire model.
HISTORY
Tires are the most complex part to understand, be it for real cars, and in a sim it is even more complex! A lot of research has been done by tire manufacturers and universities to understand how tires work and what affects the tire characteristics. Now consider trying to come up with a good 'tire model' for r/c cars using foam tires. Zero reference to go by as foam tires are not used in the real car world. So we had to do all the research ourselves. 'We' meaning our physics expert Todd Wasson who had not even seen an r/c car race when he took on the job of physics engine developer for VRC...
CREATING A STARTING POINT
The only thing we had available in 2000 when Todd started working on this, was data from the Serpent DRX-2000 datalogger, especially of our home track in Heemstede. With that data we could extract cornering speeds and knowing the corner radia we could calculate the kind of lateral forces the tires could develop in low, mid and high-speed corners. We could also approximate acceleration and braking performance using this data. That was our starting point. Not a lot to go by...
NOTHING LINEAR
Tires are so complex because nothing is linear. The overall force a tire can produce in X (longitudinal or forwards and backwards) and Y (lateral or sideways) changes with overall tire load, with camber and caster angles, and with acceleration and braking forces. Add to this the fact that foam tires don't flex like real tires do and maybe you can now grasp how complex it is to define how a foam tire should perform.
TIRE TEST RIG
Upon Todd's request we designed and built a tire testing rig. Such machines exist for real cars but not for r/c cars of course. And of course we had to take a conservative approach, not wanting to measure everything. Else we would never get to testing stage. Basically, we placed the tire on a 'grinder belt' to drive the wheel at a fixed speed (the tire should have driven the belt but that was too complex to achieve), and the wheel could be loaded vertically while changing the camber and caster angles. With load sensors Todd was able to measure the forces the tire could produce under all the different angles.
THE BEST HELP WE COULD GET
Unless you are into theoretical vehicle dynamics you probably have never heard of Doug Milliken. He is an absolute authority on vehicle dynamics and has published many books about vehicle dynamics. Just look up that name and you will find out! Todd was an avid follower of Doug's publications. And as a sheer coincidence, Doug moved from the Mid-West to the same city where Todd lived, St. Paul in Minnesota. And they got in touch... And Doug got involved in testing these little foam tires. And in analyzing the results. How lucky can you get. This was a tremendous boost in understanding the foam tires characteristics. I doubt if we would ever have achieved the performance characteristics quality, we have in VRC without Doug's expert help...
BUILDING A TIRE PHYSICS MODEL
Having useful tire characteristics data is one thing, creating a tire physics model for it is an entirely different matter. In any sim the tire physics model determines the actual racing performance, the 'feel'. It took Todd another year or so to come up with a tire physics model that produced that feel (and it has been refined further ever since!). And of course hours and hours of finetuning the parameters with the help of Michael Salven, Julius Kolff, Rene Cornella and other experienced r/c racers (Beta testers) to come up with something that felt like driving a 1:8 nitro on-road car on foam tires!
The tire test rig was also used to test 1:10 ISTC rubber tires which also gave Todd a good lead on the off-road tires 8 years later!
PARAMETERS
I can't emphasize enough how complex a tire model is, and how complex the actual tire parameter configuration is. Without the data from these tests it would have been a non-starter, just a huge guess. I have attached an image of the tire parameters we use to configure the 1:8 on-road foam tires, the values don't mean anything for you anyway. I assume and hope this will forever answer the question: why is this all so complex to get spot on? I am convinced (for almost 100%...) that none of the r/c car tire manufacturers, be it for on-road or off-road, have ever made any effort to understand tire characteristics as what we had to do for VRC Pro.
TIRE PHYSICS MODEL
This is the part of the physics model that calculates 350 or 500 times per second how much longitudinal and lateral force any of the 4 tires produce at any given moment. There is plenty of specialized literature on tire physics models, and they have one thing in common: they are all extremely complex. Just imagine trying to capture all these variables in a series of inter-dependent formulas! And of course, there is nothing ever made for simulating the behavior of r/c tires! And it has to be a super-efficient model not stressing your pc to the max while doing 350 or 500 calculations per second (500 Hertz used for the 1:12 car only because of the very small suspension movements...).
IN CONCLUSION
For real tires it is 'trial and error' style of development. You try a different compound or a different insert, or change the profile and see if it works any better.
For a sim you have to understand what parameters to use to 'define' a tire and what values to apply, else you simply can't create a 'virtual' tire.
Hope this gives you a better insight in the complexity of the tire model and makes you appreciate VRC Pro even more. And this is just one part of the VRC Pro physics model!
But remember, no physics model is perfect, a sim is, and will always remain, an approximation of reality!
Comments
Only active members can post comments
(Total posts: 5)
|
Frans B
|
Thanks for Posting, Reading up on this fantastic Detail :)
|
|
Pieter BFounder
|
Posted by: Nutty Fox (GB) on 9.9.2024 21:54:08 (UTC)
Very interesting too say the least.
what time i have spent with then only about 50laps or so.
I feel that they will be faster, its only a matter of time to understand the way the act differently.
I like the first turn in, seems sharper, like Jimmy says too much use of the power and you may spin
On short speed way i was about 0.3 faster but very inconsistent.
Thanks for all the hard work to every one who has spent a lot of time on this.
when are spec comming out?
And i am happy to see your timming this time has been good....plenty of time before Worlds
I expect that you will probably have to adjust your driving style to optimize the use of the new tires. More like modern 1:8 onroad cars behave, faster into the corner and on-throttle half way into the corner, rather then 'old school' ideal lines. Anyway, time will tell...
|
|
Nutty F
|
Very interesting too say the least.
what time i have spent with then only about 50laps or so.
I feel that they will be faster, its only a matter of time to understand the way the act differently.
I like the first turn in, seems sharper, like Jimmy says too much use of the power and you may spin
On short speed way i was about 0.3 faster but very inconsistent.
Thanks for all the hard work to every one who has spent a lot of time on this.
when are spec comming out?
And i am happy to see your timming this time has been good....plenty of time before Worlds
|
|
Jose M MModerator
|
The presentation and explanation is like a work of art, and a very extensive piece of work that deserves thanks.
My experience shows that a lot of progress has been made in this simulator, and that the work done with the tires is very close to reality.
Thank you for everything.
|
|
Pieter BFounder
|
The VRC Pro physics engine is a 'black box' carrying Todd's signature. We have no access to the actual formulaes and structures, and if even if we would, we would not be able to understand any of the inter-dependebilities. The VRC Pro physics engine has been developed from the ground up, and has been worked on for 14 years.
The VRC physics engine reflects Todd's ideas and theories of how to approach all the vehicle dynamics and other physical aspects of an r/c car, in an asphalt or carpet on-road, or an off-road racing environment, and build it in such a way that it is capable of calculating the total behavior and exact position of the car 350/500 times per second!
Edited by author: 9.9.2024 15:14:15 GMT
|
