Tire stiffness PER SIDE of tire 1

[KevinK2]

Tire stiffnes is a well described subject. For zero camber, a simple 2D model of the tire could be represented by 2 side-springs, one at each side of the wheel. However, I recently went from a 17" set to a 16" set, and found that, when subjected to a non-symetric load condition, ther was a surprising large effect on the "side-spring" rate .... much softer.

The test was my driveway's slightly uphill entry. There is a concave drain gutter along the road I have to cross to enter the driveway at about a 45 deg turn-in angle to the road. The tires cross this effective bump one tire side at a time. I think the big improvement in compliance is due to each sidewall distorting in a non traditional way, acting as a much "softer" side spring for the 16" set, vs the 17" set (same OD).

Any data or comments on this effect?

Kevin

 

[NormPeterson]

Even if both sets of tires are the same tire mfr and model (implying that similar internal construction to be more likely), lateral tire stiffness is influenced by things like sidewall height and wheel width relative to the tire tread and maximum section widths. Chances are that the 16's are taller and mounted on relatively narrower wheels, both effects trending toward softer lateral stiffnesses.

There's more that just lateral effects here. That you're crossing the drain at an angle brings the the total roll stiffness effect into play, where the vertical tire spring rate at each corner is a spring in series with the suspension (and ultimately the chassis structure itself). Changes affecting this effect alone can cause a noticeable change in the amount of lateral head toss that passengers experience.

Were the 17's H-rated or higher and the 16's T-, S, or some other lower than H speed rating? If so, it is possible that there's a carcass construction detail "missing" on the 16's that you had on the 17's.


Norm

 

[KevinK2]

Sorry I was not clear Norm.

The "side-springs" I spoke of would be a vertical spring pair, for each tire. The load from the tire is only transfered to the wheel at the tire bead, so I was suggesting using 2 side-springs (side by side) for a tire, vs the normal single spring. For example, a "wheel rate" based on metal suspension parts may be 120 lb/inch, and a typical tire rate may be 900 lb/in, so my side-spring pair approach would be two 450 lb springs at the wheel bead seats.

"..the vertical tire spring rate at each corner is a spring in series with the suspension (and ultimately the chassis structure itself). Changes affecting this effect alone can cause a noticeable change in the amount of lateral head toss that passengers experience"

This head toss effect is exactly where I was experiencing a big difference with the 16". The 17's were 93V's and the 16's are 93Y's, so you would expect a slightly stiffer/thicker sidewall, contrary to the improvement.

My first thought was that when crossing the drain with an angled approach, I could explain the 16" improvement by considering half springs sequentially crossing the drain lip, for each tire. The rate for the 16's is apparently more controlled by the sidewall height, vs it's thickness parameters. Upon reflection, it could be even a softer rate as this is a "punch" load, vs the basic flat pavement stiffness.

 

[cibachrome]

I doubt whether the constructions are anywhere near similar. Just look at the sidewall recipe. A 186+ mph tire probably doesn't have any steel in it, Bet its Kevlar?. It's probably a 2 or less tread ply tire, also. 2 or 4 plys are probably not gonna reach that high speed for very long.

Also, you've obviously changed wheels. Every tire (size, pressure, rim diameter, rim width) is designed optimally for a specific wheel constraint set. If you didn't use the recommended wheel and pressure (a tire dealer probably won't know that) your results are just annecdotal, as we say.

The high speed tires have to have a very light carcass and tread to keep them intact at speed and under abuse conditions. They also have less tread rubber (which you would probably also have misjudged if you compared your worn 17's to your new 16's.

Fill us in on the specific brand, size, pressure,rim width and construction number if possible. (You might have to call on the tire manufacturer to get that. I don't mean the Model designation). These are all O.E.M. tires or are they aftermarket jobs? (If so, this whole discussion would amount to pure speculation).

Given the construction number code, all the phenoms you describe can be explained. A F&M test result will surely show this up, too. (Which is why you run tests. It saves us all the trouble traveling to Wonderland).

 

[KevinK2]

Thanks for the info Cibachrome.

Time for me to post basic data:

Old P215-50 Tire on 17x7" wheel, 8 years old, ~.18" deep tread.

Michelin HX MXM4 93V XL ~32 psi
Tread = 2 steel 1 polyamid (kevlar)
Sidewall = "polyester body cord"

Note I googled this data, as old tires are gone, and this model apparently discontinued. The tread on this tire hardened quickly, vs same tire model that was 3.5 years old on another car where tread remains soft ... appeared to be a tread compound change as it remains soft.

New P215-55 Tire on 16x6.5" wheel, New, full tread
Continental DWS 93W SL ~32 psi
Tread: 1 polyetser, 2 steel, 2 kevlar

The 16" tire and wheel combination is the same sizing used on the Saab 9-3. Where the 17" tire was barely wider than the rim (lots of curb rash), the 16" bulges 1/2" beyond the rim outer edge. The 55 series drops the contact width to 6.5", same as rim bead seat width.

Yesterday I stopped the car as a rear tire had started to angle over the gutter's high edge. The tire had a punch load on the outboard side with high deflection there, but when I looked under the car at the inboard side of the tire, that had not yet fully reached the gutter peak, it showed little if any deflection.

Without doing a 3D FEA model (as I did when I designed Lance Armstrong's TDF TT 3-spoke wheel) of the tire loading, I think I
can support the following conclusions, for similar, standard tire groups (like all season group, and excluding DOT race tire group)

1)_ For non-conventional static loading, the tire can be considered to be 2 springs with 1/2 the typical single spring rate "K"(non-linear).

2)_ For punch loading one side of the tire (not a flat surface) the sping rate on that side is likely softer than the K/2 described above.

3)_ The "softer than K/2" rate described above, is mostly based on the sidewall height (to some power), although material of constuction parameters also influence this non-linear spring rate. The same statement also applies to the basic K/2 rate.

 

[NormPeterson]

Maybe you need to think in terms of having an applied displacement case superimposed on the static (deadweight) force case. One of your "side springs" is being compressed further going over the peak while the other is (relatively) in extension because it hasn't gotten up to the peak yet and the load has shifted. I suspect that some sort of warping stiffness in the tread gets involved as well.


Norm

 

[KevinK2]

Norm, Superposition would be one possible path, but I think the basic K/2 relationship for force vs displacement on a flat surface is lost when a punch load is introduced. Thats why I'm biased toward a new stiffness factor for punch loading, which is softer than just K/2. Tread and sidewall warping is very likely, agreed. Thank's for you interest.

Kevin

BTW, I guess I can't edit my last post, as I meant conventional, and not non-conventional, for item 1)

Also attached a tire load/defl curve, and for that small size, the spring rate is more linear than I had thought.

http://i1010.photobucket.com/albums/af226/duckants_bucket/tyreload.png

.

 

[NormPeterson]

Look for anything written by Hans Pacejka on the topic of tire modeling.


Norm

 

[GregLocock]

I agree the sidewall has some stiffness, but it ignores the balloon mode of the tire.

Cheers

Greg Locock


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

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

 

[cibachrome]

Congratulations on a fine choice of replacement tires, I believe all can be explained by the brand/model description you provided. You did make a major screwup, though, so it's still a SAAB. [You know what that stands for, eh? (Still Ain't A BMW)].

The construction recipe elements for the Conti's usually includes stiffer chafer and breaker 'parts' in the sidewall as it leaves the bead area. This is part of an improved 'road feel' tire construction and is part of the BMW road feel experience when you get Conti's OEM BMW tires. (You know how to tell a BMW factory tire, right?) Michelins, less so. Goodyear Beemer tires would probably also do pretty well in your experiment.

But, you screwed up by getting 6.5" wheels. What were you thinking? If you still have some winning lotto money, go out and get some 7" or 7.5" wheels (steel if you aren't stylish, do just the fronts if you are still paying child support). Then your right brained driving experience will be better that sex with a 21 y.o. blond Hooters babe (well at least as good, lets say).

Meanwhile find a road with some re-rod sticking up out of the concrete and run over it with one tire. The damage will demand replacement (get your insurance company or the local road commission to pay for the new meat. Make sure you get back the blown donut and cut a 1 inch cross section out of it with a band saw (bead wire is REALLY tuff stuff). Then get one of the old Mics out (you saved them for autocross, right?) and also cut out a 1" cross section. Staring you in the face will be the amazing inner details of the sidewall spring model you are contemplating. In fact you can push it down with you fingers or hands in order to run the same test analogue that your poor car is being put through.

Then go out and get the 7.5" wheels and put the 16s on Craigslist. Get some young Tuner type to suck it up and put them on his Hyundai....

 

[KevinK2]

I went with the 6.5" width to avoid curb rash. I had the optimum fitting with the 17's, enjoyed it, and decided the thrill was not worth curb-trashing the rims on this "utility" car. I always have the Rx7 FD that I set up for HPDE track events, when I need to put my nose in shear on a corner.

" .. Staring you in the face will be the amazing inner details of the sidewall spring model you are contemplating .."

Not me, and a bit condescending. The spring model in my mind would be consistent with an Instron test with the inflated tire on a flat boundary, and perhaps a 1.5" square bar placed on that flat surface, perpindicular to the plane of the tire, and about 1/3 of the way across the nominal contact width of the tire.

Hope this clarify's stuff for you.

 

[NormPeterson]

Suppose that instead of comparing your above observed sidewall deflections against each other that you compared the heavily-deflected one against the static sidewall deflections with the car at rest. How close to double the at-rest static deflection is the curb-deflected sidewall, given that the other sidewall is displaced approximately zero? What would that say about the stiffness of a single sidewall (assuming that it's OK to lump all of the various tire stiffness contributions together and divide the total by two)?


Norm

 

[KevinK2]

Suppose that instead of comparing your above observed sidewall deflections against each other that you compared the heavily-deflected one against the static sidewall deflections with the car at rest.

That's what I meant by one case "flat boundary" and another case with a 1.5" sq bar on "that flat surface". It was confusing.

How close to double the at-rest static deflection is the curb-deflected sidewall, given that the other sidewall is displaced approximately zero? What would that say about the stiffness of a single sidewall (assuming that it's OK to lump all of the various tire stiffness contributions together and divide the total by two)?

Good questions. In the absences of an Instron now, I could resort to my garage testing again. Last test: I drove up on flat 2"x6" boards under diagonal corners, and then measured the slope of the bumbers (transverse to forward direction) to see if the car roll stiffness was forward or rear biased.

 

[KevinK2]

from cibachrome:

... But, you screwed up by getting 6.5" wheels. What were you thinking? If you still have some winning lotto money, go out and get some 7" or 7.5" wheels ...

I get your point, and worried about the narrow rim choice a bit. Do you have any data or links to testing the effect of using the various rim widths "allowed" for a given tire?

Kevin

 

[KevinK2]

Norm, proposed garage "Instron" test.

Goal: test various support conditions of a rear wheel (FWD), using a constant load on a rear tire and measuring tire deflection.

Method: Jack-up RF wheel, and add blocks under tire until the RR tire remains off the ground. At that point, the LR tire will see a constant load = 2X the normal static rear wheel force.

Logical?

.

 

[KevinK2]

Sounds like what you are describing is dealing with all 4 load points active. By putting one rear in the air, unless there is a longitudinal bending moment created through the chassis structure, starting at 2 pinned boundary constraints at the 2 front tire contact points, the load at the grounded rear tire is constant ... isn't it?

To take it a step further, I could leave it in neutral, and provide front tire stops, so the chassis rotates off the front wheel bearings.

 

[cibachrome]

" Do you have any data or links to testing the effect of using the various rim widths "allowed" for a given tire?" Yes I do, that's why I made the comment. While usually known as a means to increase cornering stiffness (handling/maneuvering) coefficient of a tire, a far bigger effect is on the tire's transient response properties (rollout); again, a handling 'feel' ingredient. And, this is a result of beefing up the sidewall and fore/aft stiffnesses of a tire, in fact the very parts of a tire and the effect you are focused on. In spite of your decrease in rim diameter (for the same O.D.) a smaller rimmed tire doesn't necessarily have a softer sidewall because of "more" cross section. It depends on what parts are in there. (which is what the 1" cut sections would show you.

BTW, the really good tires already have a rim protector band added to the tire above the bead, sort of like a thin whitewall section, except that it's black. Its for the purpose of preserving the rim seat. And why are you hitting curbs, anyways? (Or do you have kerbs instead?) Get some metal curb feelers. They were really cool in the 60's

 

[patprimmer]

If in doubt I just tilt the kerb side mirror down so I can see the kerb and the rear rim. and I steer the front so the corner of the tread touches first.

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

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

for site rules

 

[KevinK2]

" ... and I steer the front so the corner of the tread touches first."

That was impossible with the oem 17's, as the rim touched first. This car (Mazda6 5 door) obstructs more front ground vision than most other cars I've driven.

" ... In spite of your decrease in rim diameter (for the same O.D.) a smaller rimmed tire doesn't necessarily have a softer sidewall ... "

I know this is true, since the 16" Kuhmo 700 Victoracer DOT race tires on the 7 have very stiff sidewalls, such that they can be at 10-15 psi and still look fine. For similar tire unofficial grouping (ie high performance summer, vs grand touring all season) I think the taller sidewall vastly improves ride comfort, as I experienced.

 

[cibachrome]

If a stiff sidewall makes your day, had you ever considered a runflat? Then you can eliminate the spare, the jack, and the tools, too...