Tire width/traction based upon vehicle weight - Exocar build question

[hipowernut]

thread68-102250

I read through the "Wider is Better? Or Not?" thread and certainly gained some insight to my dilemma. However, the question that remains for me is how do I select the best width/size of tire for a custom tube frame exocar that I'm building?

The car will weigh about 1800lbs with a mid-engine setup (40/60) and about 400hp at the rear wheels. Usage will be on the street and some autocross events.

You can see my build thread here -

http://forums.nasioc.com/forums/showthread.php?t=2093090

I have searched and searched all over the internet for a formula that could at least lead me to a size selection based upon my parameters noted above, but I've had no such luck.

Then I stumbled upon the thread noted above and more specifically the member tireman9, with hope of enlightenment to my situation.

So, can anyone help me determine the best size tire for my car?

I already have a set of street wheels and Nitto 555 tires (Front = 17x8; 235/45-17 and Rear = 17x9; 255/40-17).

I am considering an alternate autocross set that could be used for travel to and from the events. Hoosier A6 tires >> (Front = 18x10; 275/35-18 and Rear = 18x12; 315/30-18). Many are commenting that this larger set will not provide adequate traction (heat up properly) due to the light weight of my car.

I am wondering if that will be true considering the use of the lower temp Hooser A6 combined with my low CG (turbo Subaru flat 4), low ride height and wider track with the larger wheels.

Professional insight is greatly needed and appreciated.

Hipowernut

 

[CapriRacer]

A couple of thoughts:

First is that within a given tire size there is a wide range of values for grip. This is due to differences in tread coumpound. The tread compound also affects rolling resistance and treadwear (It's a 3 way compromise) That means it is very difficult for the average guy to sort out the "Wider = better grip" based on personal experience.

Second is that the size of the footprint has a pretty minor affect on ultimate grip compared to the tread compound characteristics. Again, it would be difficult to sort this out for the average guy.

Third, the best way to get you within a good range of values is to determine the load on an individual tire, then look at the load vs inflation pressure curve (also known as the load table). You're looking for something where the load on the tire is on the low end of the curve. Some of the tire sizes seem too large for your vehicle. Plus, in autocrossing, you don't get enough time to truly bring the tire temps up. So you're kind of stuck with the grip at lower temps.

Again, in autocrossing, I think it is more important to get a quick response out of a tire - and sacrifice some ultimate grip to get it. Low aspect ratios is one way to do that. High inflation pressures is the other.

I'm afraid there isn't a definitive answer to your question. as there are so many variables.

 

[CapriRacer]

Addendum.

I ran across an article on the Ariel Atom which is very similar to what you are doing.

I would look at what they did.

 

[BrianGar]

Is it not unusual that you built a car incl suspension, without considering the size of the tyre contact patch?

Brian,

 

[coyoteboyuk]

I'm quite interested in this topic myself and I'm in the middle of reading a lot of surrounding literature on the matter (scratch building a lightweight mid-engined spaceframe car with up to 400hp). While looking at the calculations I'm using it is obvious that I need to choose a wheel/tyre combination almost as one of the first tasks, so that I can identify required camber control, identify engine/transmission mounting heights to maintain ground clearance etc etc. What the OP has done, I think, is buy a kit and be attempting to adapt the tyre/wheel combo to suit it best. For this I suspect we would need to know the original spec of wheel/tyre combination and see how the suspension would cope with a change?

From reading here and elsewhere, and a post in the recent history, it would seem mid engined cars may require stiffer sidewalls than mid-front engined RWD cars. The locost 7-esque community rarely venture from 13" wheels, sometimes to 15's, and would suggest larger rims just reduce the drivability of the car and are not needed. Even some of the big V8 and V12 engined cars are running relatively small rims, high sidewalls and what I would say, intuitively, looked like tyres that are too narrow. I've not yet been able to answer the question for myself, I'm still hunting and I guess it's always a compromise. For my own car I could run anything from a 13" rim to a 19" rim with ease, anything from 185 section width to 300 or 330. Even the 300hp Atom only runs 225 16s (all round) IIRC and seems fairly well renowned for handling, but something like the caparo T1 (admittedly twice the power and about the same weight) runs ~300mm 19" super-low sidewall tyres. It's a minefield for someone heading out down this route.

 

[Tmoose]

3rd hand info - A recent (Car and Driver, Motor Trend ?) magazine article about driving (and spinning) some Porsche (?) on a race track mentioned Milliken advocated tire widths proportional to F/R weight distribution for balanced handling

 

[GregLocock]

I'd have thought the biggest problem you'd have is the absolute dearth of objective performance data for a tire available to the retail customer (don't worry, it's not just you, OEMs pay to have tires tested).

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[coyoteboyuk]

Indeed, there's pretty much nothing available for non race tyres so it's bordering on pointless, barring a general design strategy. The variation of road tyres from one make to the next in exactly the same sizes is vast and undocumented (I find it somewhat odd that you can't see the precise performance data of the product you're buying as a consumer - the tyre manufacturers must have this data). Presumably this is why lots of the car designers simply pick a tyre from a similar known-good tyre combo. No help to the OP or me!

 

[CapriRacer]

Coyote,

I feel compelled to correct some misconceptions you have. First let's talk about what info IS available on tires:

Every tire has to have a UTQG rating. It rates treadwear wear, traction, and "temperature".

The treadwear rating has some quirtks in it, but at least it tells you something about treadwear - especially considering how much variation can be experienced just by differences in driving style.

The traction rating is also quirky because it is so coarsely granulated. On the other hand, tire traction testing has it's own problems and in some respects, the coarseness of the grading is a reflection of the state of testing at the time the regulation was formulated.

The temperature rating is really speed rating. I think this rating is an American "Not Invented Here" response to the European Speed rating system.

So what about the high range of values between tires of the same size? Each consumer has his own ideas of what tires are supposed to do, and since there are compromises involved, a single make and model can't do everything. The UTQG rating helps clear this up a bit. But that's why reading reviews is important. In much the same way car magazines review cars, there are reviews and rating of tires of various sorts - if you look for them.

So what do YOU want in a tire? Great treadwear? Great traction? Sorry, can't have both! Those are tradeoffs! Somewhere in between? What about fuel economy? That's the 3rd leg of the compromise.

If you are posting in this forum, you're an engineer - and you should be aware that testing brings with it a whole series of issues. Let's take Rolling Resistnce as an example of the problem.

Obviously there are different Rolling Resistance tests. The numbers recieved from each test is different. So which one should be published?

Then there's the problem of tire size - rolling resistance varies by tire size for otherwise identical tires. Test only a representative sample? Test all sizes? (Hint: with tire lines having 60 different sizes and a couple dozen different lines of tires, this number becomes huge in a hurry!)

And lastly - Greg, back me up here! - The car manufacturers don't go shopping for tires the way you and I do. They specify the properties they want and the tire manufacturers design to those specs. They NEVER select a currently available product. It is ALWAYS custom designed - and customer designed.

Personally, I have a problem with the specs the car manufacturers use. They greatly reflect the needs of the car designer and much less so the needs to the consumer. The result is generally tires that have treadwear and traction issues. But it is the way these things are done - and has been for decades and isn't about to change any time soon.

So while it would be nice to have all this data availble, the sheer volume of testing needed is just beyond rediculous. Precision? Unfortunately, tire testing has some inherent variablility that is always going to impact tghe varaiblility. Precision is just not going to happen - and that is just the way it is.

 

[NormPeterson]

Why isn't more tire information available?

The average tire-buying consumer simply wouldn't understand it or be realistically able to relate the tire performance envelope to his own mix of requirements - if he even has more line items than round/mostly black/holds air and is at least somewhere near the right size on that list. We're talking about people who either don't notice turn-in response softening as tires gradually lose air in the first place, or don't relate the effect back to the cause.

I'm not even convinced that the average non-motorsports-enthusiast engineer would have a whole lot of use for more data than UTQG. Most people don't ever intentionally drive hard enough for it to matter, and if they occasionally find themselves up there they're a bit too busy to notice.


Norm

 

[NormPeterson]

coyote - the best resource for comparing some tires against each other seems to be Tire Rack's own comparative testing (which limits the selection to only the products that they sell). Even that has quite a bit of variation if you compare the results for the same tire over two or more separate test sessions.

Consumer ratings and reviews are generally less useful. The ones acknowledging competition or other hard use and go into some detailed descriptions are the only ones worth reading.


Norm

 

[coyoteboyuk]

CapriRacer - cheers for your input, always happy to have misconceptions corrected by those in the know. Clearly all the factors are a tradeoff but if you have raw numbers from a standardised test structure you at least have something to go by. However I still don't think there were any misconceptions - the info that is available is, as you say, fairly coarsely assessed. I've two different makes of tyres on my normal road car with identical ratings of wear, traction and temp on the sidewall. Both are "budget performance" if there ever were such at thing! Their behaviour is light years different - one set of tyres (Toyos) roll their sidewalls far easier at the same pressure than the other set (Federals) and while their wet traction is somewhat better than the federals, the toyos flex too much and create massive turn-in understeer when up front. Switch ends and the handling is MUCH improved. Clearly this is partly down to the suspension setup etc but the like-for-like swap and change of effect is not reflected by the rating system at all.

I find it rather shocking that there can be such a staggering variation in performance with no hint of it in any figures available to the consumer. The argument that "there's virtually no one who'd know what to do with it", as NormPeterson says, is not really valid, there's millions of people out there driving cars with no idea how the power and torque figures affect their vehicle performance, but the manufacturers will provide that info so you can use it to select between models and makes. Considering the key properties of the tyre are fundamental to the reason for buying it, how can we not have realistic info about their performance at the point of sale?

But that's a side issue really, the point is that most tyres do not have more than the basic info available, making it impossible to design a car around. You say OEM's order tyres to their spec - that's fair enough and as I would expect of course, but most people don't stick to the tyre designed for their car - how do they know how their car will handle with a different manufacturer fitted, who created a same-size tyre with different characteristics? End users can be educated, they shouldn't have to choose blindly. Yes testing causes problems of "but your test doesn't show this, or accentuates this" - that's not really the end user's problem. The test should be selected by a combination of tyre and vehicle designers to ensure it picks out the 5-10 key features that reflect handling and performance of the vehicle and ensure that the same test is carried out on all tyres, all sizes of each model.

 

[GregLocock]

"They specify the properties they want and the tire manufacturers design to those specs." Deepnds a bit on the manufacturer, and the sales volume. I know at Lotus we'd often work with a given existing tire for the entire program, and tune the car to the tire. Given our sales volume that made more sense.

But for most mass volume cars yes, we specify what we want, and then the tire company gives us what we are prepared to pay for, and so it goes on, for anything between 6 and 40 iterations. The problem with developing a custom tire is that the PAS calibration is optimised for a given cornering stiffness (in particular) so as each new submission comes in we have to revisit the PAS cal. For sports SVPs again we often pick an existing tire, and develop the tune around that (springs, shocks, sta bar).








Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[CapriRacer]

Coyote,

Thanks for responding - and a special thanks for responding in such a positive manner. Discussions of this type can quickly evolve into pissing contests - especially when one (or both) sides want to prove they are right - at the expense of illuminating the truth.

I'm also glad you have delineated the problem you have. This is an area that has particularly vexing issies. But first let's talk about testing and how it should be published.

As you suggested, there has to be an agreed upon standardized test - and to illustrate this, allow me to discuss the UTQG treadwear test.

UTQG testing is a government regulation. It has to be! Voluntary cooperation just doesn't work - just too prone to cheating!

When this regulation was in the discussion phase, there were lots of points raised about what test to perform. This discussion went on for a long time - a decade comes to mind - and finally, the feds got fed up and decided on the test we know and love (?) today. It's a 10K test with a comparison to a standard tire, known as the SRTT (Standard Reference Test Tire).

At the time the test was intiated, the SRTT was a bias tire and it was assigned the value of 100. Every tire was to be compared to that tire. A tire that got half the wear would get a value of 200.

Today we have a different SRTT - a Uniroyal radial tire and it has a different (higher) value based on a whole set of correlations. - But the principle remains - a comparison.

But what about tests at different facilities? Shouldn't there be a continuing correlation between facilities to make sure a numbered result is the same regardless of where the test is performed? In the case of the UTQG treadwear test (traction, too!), there is only ONE (!!) approved test factility. Problem solved? Sort of!

So how has this panned out? Well, some tire manufacturers are conservative when it comes to these ratings, so they tend to understate. Others tend to take the max approach. The only governmental stipulation is that the tire manufacturers can't OVERSTATE the result. No penalty for understating.

Then there is the marketing guys. They want to differentiate between lines of tires - and even though there may only be a 100 point difference between lines of tires, they may decide to understate one of them in order to have a selling point for the one they are charging a preminum price for.

Have you ever heard folks say you can't compare UTQG ratings between brands? Not so, as I hope you can see above.

And let's talk cost for a moment. The last time I checked - and I'll bet that was 10 years ago! - doing this kind of test cost about a $1.00 per mile PER TIRE. That means the 10K UTQG test with a control costs $80K to perform. I'l bet it is over $100K now.

So let's get back to the situation you presented. One of the other compromises in tires is the tradeoff between - and I'll use engineering terms - ride harshness and steering crispness. (BTW, I'll bet I get feedback from others who visit this website who say they use different terms. No surprise there.) This is the difference between the ride quality and how responsive the steering is.

First, these parameters have always been evaluated by the seat of the pants, by ride engineers. In other words - it's subjective. That creates all sorts of interesting problems - the least of which is differences between evaulators. There are also differences between how tires react on different vehicles. This makes the test very imprecise.

But allow me to draw a conmparison to cars. Do car manufacturers have a test they publish that evalutes the ride qualities of their vehicle? What about how a test on how the steering feels? Nope, that is what car magazines do. Should this be different for tires? If the car manufacturers can't seem to find a way to communicate these kinds of differences, how can tire manufacturers?

That's where a group such as Tire Rack becomes so valuable. They hope that if they do these types of evaluations, you'll buy from them - and that seems to be working.

Consumer Reports also does thes types of evaluations - in the hope you'll buy their magazine.

Does the ordinary guy take advantage of these resources? Well, I had a guy who bought an Audi with a sport package (40 series tires) complain about the ride. He thought that since it was expensive, it ought to ride as good as his Buick!

- and you touched on a sore point with me about vehicle manufacturers when they go way out on a limb when selecting the characteristics they specify.

But allow me to point out that "5 to 10 key features" for each tire size is an unmanageable amount of testing.

When NHTSA proposed its rolling resistance test (and this is a lab test that doesn't take too long to perform), it was pointed out that if EVERY test facility in the world were utilized, it would take 3½ years to complete - and that during that time NOTHING ELSE could take place. No research, no development, no breakthroughs! What you have proposed would easily take 3 times as long.

Sorry, but this just not as easy as it sounds.

But I am glad we could have this discussion.

 

[coyoteboyuk]

@GregLocock - Did you, at an OEM, have to pay to get access to the tyre data or did you simply test and re-define changes in those iterations? At the end of the day I, and the OP to some extent I guess, are trying to get the best data we can to base a design on. What seems to be openly available is "suck it and see", but I suspect OEM's get far more detailed info to start with.

@CapriRacer - I'm not here to argue, I'm here to learn and maybe contribute at some point, and there's plenty of people on here with a lot more experience than I have. Part of my learning is to create the best vehicle I can for my own satisfaction - no point taking part in self-defeating arguments. That's some interesting info you put across there, it's somewhat mindblowing that the testing costs could be that high, I'd have expected tyre manufacturer to have in-house testing facilities that would be checked by a national overseer - much the same way that petrol station pumps are maintained and run by retailers but legally get checked for accuracy by a central government body. I can't understand the pricing of that, but that's not fully appreciating what goes into the tests I guess. Thanks for showing me some of the less visible sides of the industry.

Well (and still trying to keep the thread within the realms of the OP's post, though I have hijecked this somewhat) - how can Joe Public Eng. take the best approach to designing a vehicle suspension system and accurately size the tyres required based on torque and vehicle weight etc? It seems that without the detailed info a race tyre manufacturer my provide, or an OEM may have access to, the best that can be achieved is a rough guess (makes me shudder!) based on the review agencies that you specify? I recognise that a narrower tyre may provide more progressive slip at the limit, a wider may provide ultimately higher lateral grip, but without some figures how would someone such as yourself with plenty of knowledge and experience of the field attack the design problem? I am personally almost tempted to simply pick a tyre/rim combo based on cost and availablility in sizes similar to similar weight vehicles (such as the Atom) and suck it and see, but that could become a very expensive, not very engineering-esque, method of getting to the ideal!

 

[patprimmer]

cyote

I am sure if you can order 100,000 tyres to be delivered over one year, or if your prepared to pay say $10,000 per tyre for 4 tyres you might get better service.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[coyoteboyuk]

Naturally!

 

[GregLocock]

We pay someone like Smithers to test each tire. Typically that costs around 20k, and 4 tires.

Sometimes the tire manufacturer supplies Pacejka coefficients for a particular submission, but they tend not to be a full characterisation, ie including parking and limit. Sometimes they just give us a cornering stiffness.

We've found that it is possible to model a lot of the changes that we ask for, for instance it is relatively easy to predict what will happen if pressures are varied by a few psi, or a different rim width is used, or a 17" version of a 16" tire is made, or small changes in width, etc.





Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Tmoose]

Hi Greg,

You said "it is relatively easy to predict what will happen if ..... a different rim width is used..., "

In one of those handling articles by Jim Hall in Car and Driver back in 1965 I think he said something fairly categorical, like, wider rims reduce slip angle. Does that sound right, yesterday and today?

thanks,

Dan T

 

[GregLocock]

Yes, increasing rim width increases the cornering stiffness (up to a point).

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?