Tire data

[peaty]

hi
As I've been saying in other posts, I'm not a chassis guy but I'm trying to learn. I've been reading a lot these days on the topic and I've learn many things while reading or talking with some of you guys. My today's question is regarding tire. I've found some data in Avon' website. I would like to know how useful is having that kind of information and what can I do with it, in other words, what calculations can I do. So could you please help me to make a list of all the things I can calculate and/or the things for which tire data is useful!?

Thanks

[Jersey Tom]

Good question, one I'm sure that will yield a variety of opinions. Here's mine:

If you know what you're doing, and the data isn't junk, tire data can be useful for many things. "Tire data" is a pretty broad term though. Loaded radius and spring rate data can be very important on downforce cars to accurately predict ride heights. Force and moment data can be useful in figuring out what the balance and handling of the car will be. As an easy example you can calculate your understeer gradient and limit under- or over-steer.

With that said, I pretty much wouldn't trust any tire data I wasn't directly involved in testing.. or at least very familiar with the rig, procedures, tires, etc. It's really easy to generate junk tire data. Or as a corollary it's really hard to generate consistently good data. You can have two engineers test the same tire with slightly different procedures and come up with wildly different data sets. You can then have two other engineers take the same set of data and come up with wildly different models. Even then, you can have two analysts with the same model that come up with vastly different performance recommendations.

So it's far from clear cut. The field of vehicle dynamics is small enough as it is. A small subset of that have involvement at some level of the tire data game. An even smaller subset of that do it with some real level of competency. I would say there are only a handful of gurus who truly "get it" from the standpoint of tire physics, to testing, modeling, and analyst work.

Ultimately I would say it's a double edged sword, like many things in motorsports. To give an analog that maybe more people are familiar with... it's like an ADAMS model of a vehicle. Full multi-body representation of every control arm and piece part with geometry and masses and inertias and compliances and all these things. To an expert user you can get a lot of insight out of such a tool and answer some very specific questions. To a non-expert it's probably a distraction and waste of time - or at least you can say there are better uses of your time.

Likewise I would say with tire data, if you've really got some expertise and experience it can be a very powerful tool for handling analysis and car setup. If you're new to everything it can be a distraction, inefficient use of time, and lead you down the road to the "wrong" answer.

Real opinion which I'm sure some would (and should) debate: Do you need tire data to do good suspension design? No..

[Greg Locock]

One of the big problems with Flattrac data is that it is measured at slow speed on a belt sander. Then when you get into the real world you get to apply 'corrections' to that carefully measured data to make it match what you measure on the road.

Now, if these corrections were mostly in the non linear range, and say 10%, you might think there is some integrity in the process. Sadly we often have to correct the linear range, by as much as 30%, to get good correlation. Now when I say good correlation I mean anything from overlays of parking effort forces in the tie rods, like this (on concrete and a slip plate) used for designing the PAS

http://www.mediafire.com/view/itsv3imir9g0xu3/PE.JPG


which admittedly doesn't need an ADAMS model, so long as you can predict the slip plate curve (the small one) in your spreadsheet

-to being able to predict the steering wheel angle at which a SUV will start to rollover, at a given speed, and correctly model the post-wheel liftoff behaviour - ie does it roll or does the wheel touch down again? Or in crosswinds, if the car is hit by a blast of wind when passing a truck is it going to be stable?

All of these things can be estimated in simpler models than ADAMS, but perhaps not as accurately, and almost certainly with more effort, once you have the ADAMS model built. And sadly the tire is probably 50% or more of the correlation work.

But for a race car, where frankly if the coil spring needs to be made 20mm bigger in diameter you can just do it, is there as much value in this?

...well, yes there is. On an open wheeler it is crucial that you have linear range understeer in a straight line, or at least, your instability speed is greater than the max speed of the car. Your tire data is the only way you can work this out in advance, especially if you have aero.

You don't need a fancy ADAMS model (the equations are in Bundorf's SAE paper, or probably RCVD), but you do need to know how the cornering stiffness varies with vertical load for each tire. I have also run into problems on a production car where a tire did not have the same sensitivity as the one it replaced - we were using that to get limit understeer, the new tire didn't saturate, so the car would oversteer at the limit, which is not acceptable for a road car these days.

[Jersey Tom]

One of the big problems with Flattrac data is that it is measured at slow speed

Sounds like you need a newer Flat Track.

[Greg Locock]

Good point.

Actually our current machine is deficient in several ways, so I assume a zillion dollars has been set aside for a new facility. I guess a high speed test trashes the tire even more quickly.

[gixxer_drew]

Curious, how are you dealing with tire heating on the test rigs?

[DaveW]

Curious, how are you dealing with tire heating on the test rigs?

That is a very good question. My guess would be that Ta would be recorded & the test procedure would be written in stone.

Greg, JT, what do you think?

[Jersey Tom]

Greg, JT, what do you think?

I think details on tire testing is where I don't give away freebies

[peaty]

Thank you all for sharing part of your knowledge with me. These days I've been reading and talking with a lot of people about this topic.

Good question, one I'm sure that will yield a variety of opinions. Here's mine:

If you know what you're doing, and the data isn't junk, tire data can be useful for many things. "Tire data" is a pretty broad term though. Loaded radius and spring rate data can be very important on downforce cars to accurately predict ride heights. Force and moment data can be useful in figuring out what the balance and handling of the car will be. As an easy example you can calculate your understeer gradient and limit under- or over-steer.

With that said, I pretty much wouldn't trust any tire data I wasn't directly involved in testing.. or at least very familiar with the rig, procedures, tires, etc. It's really easy to generate junk tire data. Or as a corollary it's really hard to generate consistently good data. You can have two engineers test the same tire with slightly different procedures and come up with wildly different data sets. You can then have two other engineers take the same set of data and come up with wildly different models. Even then, you can have two analysts with the same model that come up with vastly different performance recommendations.

So it's far from clear cut. The field of vehicle dynamics is small enough as it is. A small subset of that have involvement at some level of the tire data game. An even smaller subset of that do it with some real level of competency. I would say there are only a handful of gurus who truly "get it" from the standpoint of tire physics, to testing, modeling, and analyst work.

Ultimately I would say it's a double edged sword, like many things in motorsports. To give an analog that maybe more people are familiar with... it's like an ADAMS model of a vehicle. Full multi-body representation of every control arm and piece part with geometry and masses and inertias and compliances and all these things. To an expert user you can get a lot of insight out of such a tool and answer some very specific questions. To a non-expert it's probably a distraction and waste of time - or at least you can say there are better uses of your time.

Likewise I would say with tire data, if you've really got some expertise and experience it can be a very powerful tool for handling analysis and car setup. If you're new to everything it can be a distraction, inefficient use of time, and lead you down the road to the "wrong" answer.

Real opinion which I'm sure some would (and should) debate: Do you need tire data to do good suspension design? No..

Really interesting Jersey Tom, thank you very much! Now I see your point about suspension design and tire data, but leads me to another question though. What's a good suspension design!? I've been always told that a good suspension design is the one that makes the tire "happy". I assume you can make a tire work through a good setup, so what are the things I must look at when designing a new suspension?
P.S: from a driver stand point, does having a linear motion ratio (motion ratio change through travel) makes the car more precdictable and easy to drive!? I know there are other factors to consider, but let's get rid of them for now and focus just in the motion ratio.
Why would you want to run digressive or progressive rates!? If you do so...your roll stiffness distribution is going to be a function of chassis roll, isn't it!?
The more I read, the more questions I have...

[MarcelaFus]

Just to add something to this thread; did anyone read an article where it said in a small part of it bigger wheels with the same width tires made the handling worse? I definitely remember reading something like that in Autocars tire test last year maybe? The test where Goodyear Eagle F1 had came up first. This particular info was for 225/40 R18 and 225/45 R17. The unfortunate thing is I cant remember if bigger made it better or worse. Ill try to find my copy.