Tire Rolling Resistance

[doctorSpoc]

The problem with a test that is designed to reflect "real-world" surfaces is that there isn't any such thing. Road surfaces are different. Again, as Scott said, so long as the rankings don't change and the magnitude of the differences is proportional, it's not necessary to test under real-world circumstances for purposes of choosing a tire. For other purposes (e.g., predicting TT times), it would be useful to test on real-world surfaces, but which one?

for sure... what is real world?

but this doesn't address the question I'm raising... DO bumps, texture (which are really just small bumps) cracks etc. affect the relative rating of Crr of tires in a way not reflected in these test? we don't know this because we haven't seen any data that demonstrates that the ranking of tires would or would not change when tires have to go over surface textures that even approach what tires would be subjected to on the road. it's not enough (for me at least) for you to just state that it doesn't and that's the end of it... it may or it may not.. there are a lot of variables and as i said before why make all these assumptions why not just test a few surface textures with cracks, bumps etc. and see what the deal is?

and if we do know one thing for sure.. "real world" is not a perfectly flat shiny drum...

[Edit]come to think of it doesn't even using a drum change the shape of the contact patch and possibly the skew numbers as well?[Edit]

 

[RapDaddyo]

for sure... what is real world?

but this doesn't address the question I'm raising... DO bumps, texture (which are really just small bumps) cracks etc. affect the relative rating of Crr of tires in a way not reflected in these test? we don't know this because we haven't seen any data that demonstrates that the ranking of tires would or would not change when tires have to go over surface textures that even approach what tires would be subjected to on the road. it's not enough (for me at least) for you to just state that it doesn't and that's the end of it... it may or it may not.. there are a lot of variables and as i said before why make all these assumptions why not just test a few surface textures with cracks, bumps etc. and see what the deal is?

and if we do know one thing for sure.. "real world" is not a perfectly flat shiny drum...

Well, I guess we have two choices. One, we can find reasons that the available test data are not representative of real world and reject the data and use anecdotes such as the LBS Cat1's perception above. Two, we can base decisions on the available data and assume that the relative rankings and proportional differences between specific tires will not change significantly on real-world surfaces. I'll take the latter approach and await your real-world test results. When do you plan to publish your results?

 

[acoggan]

DO bumps, texture (which are really just small bumps) cracks etc. affect the relative rating of Crr of tires in a way not reflected in these test? we don't know this

Actually, I believe that we do know this, i.e., to my recollection people have shown that the surface texture of the drum (or even whether it is a drum at all) doesn't change the rank order of various tires. You'd have to dig back through old discussions on r.b.tech to find a reference, though.

[Edit]come to think of it doesn't even using a drum change the shape of the contact patch and possibly the skew numbers as well?[Edit]

Now here you are on to something...but again, while the use of an excessively curved surface can inflate the CRR values, they are still valuable for relative comparisons.

Of course, if you want absolutely "real world" data, you can always go to an indoor velodrome and rides lots of laps at different speeds using a powermeter...but then some might reject that data as not being applicable to a less smooth surface!!!

 

[doctorSpoc]

Well, I guess we have two choices. One, we can find reasons that the available test data are not representative of real world and reject the data and use anecdotes such as the LBS Cat1's perception above. Two, we can base decisions on the available data and assume that the relative rankings and proportional differences between specific tires will not change significantly on real-world surfaces. I'll take the latter approach and await your real-world test results. When do you plan to publish your results?

i have a third... armed with the knowledge that:

- tires with high thread count, fairly thin, supple casings all have relatively similar Crr (i.e. not a crappy tire to the best tire)
- that Crr compared to things like drag of rider and bike at race speeds is pretty much insignificant.. and that a small differential in a numbers that is fairly insignificant is itself insignificant and even more so....
- and that the traction--> handling, comfort --> less fatigue are significant attributes of a tire etc. etc
- using tests that don't even approach real world condition and analysis that has all kinds of assumptions and pit falls built into it is not really much better than using anedotal evidence is not that useful

i just ignore the charts read the thread coun't numbers, make sure it doesn't include doubled over casing.. feel the tire to see how thick the sidewalls are and how supple it is and finally try the tire out and see how well it handles under race conditions and how nice they roll (comfort) and be comfortable in the knowledge that maybe some others out there may have tires with maybe some fraction of a percent difference in Crr because at the end of the day it's really not here or there... 'cause maybe they need to slow down quite a bit to get though a corner and lose position every time and spend lots of time and energy with losing a making up placings or their muscle are more fatigued at the end of a race because their tires are at 140psi and they have been riding a jackhammer for the past 4-5hrs etc, etc, etc.

 

[RapDaddyo]

Crr compared to things like drag of rider and bike at race speeds is pretty much insignificant

Not sure I buy that one.

 

[scotmart]

- tires with high thread count, fairly thin, supple casings all have relatively similar Crr (i.e. not a crappy tire to the best tire)
- that Crr compared to things like drag of rider and bike at race speeds is pretty much insignificant.. and that a small differential in a numbers that is fairly insignificant is itself insignificant and even more so....
- and that the traction--> handling, comfort --> less fatigue are significant attributes of a tire etc. etc

I disagree with all 3 of these...

#1 because #2 is patently not true (such that even small differences are significant).

#3 because traction and handling are similar, and comfort is primarily a function of inflation pressure.

Scott

Edit: To qualify the small difference part... Even small differences in Crr can equate to 10+watts to the road. Between now and nationals I am hoping to eek another 10 watts out of my FTP. The amount of effort and pain I will endure in the next 3 months to try to accomplish that is very significant in my eyes.

 

[doctorSpoc]

I disagree with all 3 of these...

#1 because #2 is patently not true (such that even small differences are significant).

#3 because traction and handling are similar, and comfort is primarily a function of inflation pressure.

Scott

Edit: To qualify the small difference part... Even small differences in Crr can equate to 10+watts to the road. Between now and nationals I am hoping to eek another 10 watts out of my FTP. The amount of effort and pain I will endure in the next 3 months to try to accomplish that is very significant in my eyes.

the point of #3 was to show that things like rubber compound and inflation pressure... things that effect handling and fatigue ... things unrelated to Crr, but related to tires can have very significant effects on performance.

 

[mikesbytes]

Being a nob on tyre rolling resistance, I am interested to see the compartitive figures between tyres. What your posts imply is that the figures do not accurately reflect real-world situations, however they do give a general perspective and this could be used to compare tyres of different brands to your current ones and therefor increase your choices.

What I see here is;
1. There is not a lot of difference between the performance of Tublars vs Clinchers. Each has their pros and cons.
2. There is a opitum pressure to run tyres and when over that the bumps in the road push you back, negating the benfits of lower rolling resistance and decresing the confort levels.
3. Some parties say "ignore the data and count threads".
4. One person says that tyre x is **** while another says its wonderful. Personal choices are abundant.

So does anyone have a link to some data, even if they disagree with it.

 

[earthman309]

So does anyone have a link to some data, even if they disagree with it.

http://www.rouesartisanales.com/article-1503651.html

SciGuy

 

[MichaelB]

20W??? 20W at what speed? I cant see where it says the speed it was tested at.

Its like saying your car has 100bhp when actually it produces 100bhp at one single point on the rev-band.

Did they quote a speed in that data?

 

[MichaelB]

Alright forget that. It does say on the main page. It was tested at a measly 30km/h ~18.6mph.

 

[mikesbytes]

While the data does leave a lot of questions to be answered, it do show 20w difference between the two ends of the results. So is that 20w per wheel (40w total). And what performance difference does 20/40w make?

BTW here's an example which shows the impact of tyre pressure. http://bike.terrymorse.com/rolres.html

 

[doctorSpoc]

Michelin seems to understand that on real road there is an optimum pressure for low rolling resistance and that it is a myth that higher pressure equals lower rolling resistance on real honest to goodness pavement...

"...In theory, higher tire pressure equates to less rolling resistance. But theory doesn't count for much. The cycling world exists in reality. For this reason, Michelin's tires have a recommended pressure dependent on a rider's weight. Cyclists who over-inflate pay a price on the road when energy meant for horizontal acceleration is funneled into vertical fluctuation. Even tiny imperfections in the pavement can cause this fluctuation. This loss in efficiency is often compounded by an over-emphasis on the down-stroke while pedaling..."

http://www.bicycling.com/article/0,3253,s1-15146-P,00.html?category_id=361

 

[frenchyge]

Michelin seems to understand that on real road there is an optimum pressure for low rolling resistance and that it is a myth that higher pressure equals lower rolling resistance on real honest to goodness pavement...

I suppose, but look how smooth their test drum is. Those dummies will never learn....

 

[doctorSpoc]

Haven't read the whole thing but here is the comparison i was looking for, for bicyle tires done for car tires.. not likely completely applicable since cars have a suspension system and on bicycles the tire constitues the entire suspension system...

anyway, here are the conclusions though...

seems they say that yes the relative ranking of tires is the same on rough and smooth surfaces is the same but the differential between the tires on rough surfaces is not as great as they are on smooth surfaces. this is the type of non-obvious things that just fall out of real world testing but impossible to intuit when doing unrealistic testing and making tonnes of assumtions and approximations...

http://www.iea.org/Textbase/work/2005/EnerEffTyre/sandberg.pdf

Conclusions (tentative, based on pilot study)
•Trailer measurements of RR are possible
•…but critical regarding characteristics of test track or road
•Ranking of tyres (on drum) basically the same on rough as on smooth surface texture, but RR much higher on rough surf
•Very clear surface influence on RR
•50 % diff due to surface for large surface range
•25 % diff between surfaces in common use ( ⇒energy saving potential !)
•60 % diff between tyres on smooth + even surface (100 % diff forwider selection)
•Air drag on tyre and hub gives speed-dependent RR contribution
•When not taking air drag into account one gives wide tyres a favour
•Much smaller difference between tyres on uneven, realistic surface than on surface with unrealistically low unevenness (such as drum)
•The smaller difference probably caused by unevenness contribution to RR
•The lack of unevenness on drum may mean that tyres are suboptimized
•The lack of air flow on drum also may mean that tyres are suboptimized
•Solution: Introduce more realistic surface on drum & calculate the air drag contribution and add it

 

[doctorSpoc]

have taken some time to read the presentation and they would seem to indicate as i suspected that potentially their are some real problems with the testing of tires on smooth drums.

- even though we see that in general tires have the same ranking we can see that some tire have good ranking on smooth an very low ranking on rough... with one tire going from best of 12 on smooth to 8th of 12 on rough and the rougher the surface the less the differential between tires.

- interestingly they suggest folding in the wind drag into the measure which seem to make all kinds of sense since there is a trade off between tire size (wider roll better and thinner have less drag) and having a rating that had these two factor rolled into each other would make, making sense of the number much easier and obvious.

- some of the assumptions and conclusions extrapolated from the smooth drum test are suspect... i.e. the difference in RR between differrent tires on smooth and rough road will be constant, that a tire's RR on rough surfaces will go up by some constant factor on rough surfaces when this test suggest that some tires can perform much better relatively on smooth surfaces than on rough ones and vise-versa etc. etc.

Question - why don't TTrs use big honkin' tires for their back wheels.. it would seem from all test that this would reduce RR and the drag is not as much of an issue since the rear tire hides behind the seat tube? come to think of it why don't all cyclist do that?

 

[doctorSpoc]

scottsmart,

taken from the Michelin Pro2 Race article above.. here is exactly the point I was making with #3...

i'd be willing to take a higher RR to get tires that grip in corners and in the rain... lower RR doesn't look that good if to acheive it you are losing 3, 4, 5 places into each corner of a crit 'cause you don't trust your tires or sliding out and crashing in the rain because maybe you trusted them more than you ought to.

"...The original Pro2 Race had a reputation as one of the most efficient, free-rolling tires on the market. The problem was that the tire was notorious for shady handling in slick conditions..."

 

[WarrenG]

scottsmart,

taken from the Michelin Pro2 Race article above.. here is exactly the point I was making with #3...

i'd be willing to take a higher RR to get tires that grip in corners and in the rain... lower RR doesn't look that good if to acheive it you are losing 3, 4, 5 places into each corner of a crit 'cause you don't trust your tires or sliding out and crashing in the rain because maybe you trusted them more than you ought to.

"...The original Pro2 Race had a reputation as one of the most efficient, free-rolling tires on the market. The problem was that the tire was notorious for shady handling in slick conditions..."

The grip has been improved, and there is a "grip" version of the Pro2 Race. Choose your tire on the day?

 

[scotmart]

scottsmart,

taken from the Michelin Pro2 Race article above.. here is exactly the point I was making with #3...

i'd be willing to take a higher RR to get tires that grip in corners and in the rain... lower RR doesn't look that good if to acheive it you are losing 3, 4, 5 places into each corner of a crit 'cause you don't trust your tires or sliding out and crashing in the rain because maybe you trusted them more than you ought to.

"...The original Pro2 Race had a reputation as one of the most efficient, free-rolling tires on the market. The problem was that the tire was notorious for shady handling in slick conditions..."

I have never had any problems using PR2's, nor has anyone I've known. My guess is that that statement came from one of the many cyclists who mistakenly believe that slicks have less traction than tires with some tread (the PR2s are slicks), especially in the rain.

The fact that tread pattern has nothing to do with bike tire traction, and very little to do with Crr, is also why using car tire tests is a poor comparison. The link you gave, and several others, have been brought up before. The problem is car tires have a much larger contact patch relative to the are of sidewall that is subjected to deformation. Tire compound and especially tread pattern play a significant role in car tire Crr, and very little in bike tire Crr. Also, the particular pattern of the tread can have a significant effect. A tire with a rough tread pattern will pay a relatively larger penalty on smooth roads compared to rough roads (where less tread deformation occurs due to irregularities). So this is one way in which a tire that performs poorly on smooth roads can get better (relatively) on rough roads. It's not that the tire is better, it is just paying less penalty for its tread.

Bike tire tread patterns are too small in relation to surface irregularities, and most importantly, are neither deep enough or robust enough to resist deformation on any road surface. That is why they make no difference in traction, have minimal effect on Crr, and why comparing car tires tests to bike tires is innappropriate.

Scott

 

[doctorSpoc]

I have never had any problems using PR2's, nor has anyone I've known. My guess is that that statement came from one of the many cyclists who mistakenly believe that slicks have less traction than tires with some tread (the PR2s are slicks), especially in the rain.

The fact that tread pattern has nothing to do with bike tire traction, and very little to do with Crr, is also why using car tire tests is a poor comparison. The link you gave, and several others, have been brought up before. The problem is car tires have a much larger contact patch relative to the are of sidewall that is subjected to deformation. Tire compound and especially tread pattern play a significant role in car tire Crr, and very little in bike tire Crr. Also, the particular pattern of the tread can have a significant effect. A tire with a rough tread pattern will pay a relatively larger penalty on smooth roads compared to rough roads (where less tread deformation occurs due to irregularities). So this is one way in which a tire that performs poorly on smooth roads can get better (relatively) on rough roads. It's not that the tire is better, it is just paying less penalty for its tread.

Bike tire tread patterns are too small in relation to surface irregularities, and most importantly, are neither deep enough or robust enough to resist deformation on any road surface. That is why they make no difference in traction, have minimal effect on Crr, and why comparing car tires tests to bike tires is innappropriate.

Scott

the statement seems to have been the belief of michelin themselves... the article is about changes to the original Pro Race 2 tire to fix it's lack of grip.. but the point wansn't even meant to be a slight of the Pro Race 2.. it would seem it's grip problems have been fixed, it was just to point out that other factors related to tire are as or more important, depending on how you look at it, and as WarrenG stated race day dependant... low RR isn't going to do you much good when you are sliding along the pavement... or having to put out 50-75 more watts than other around you out of every corner because your tire's grip sucks.

edit - and thetread pattern explaintion seems to make sense.. but in the end we still don't really know because the real world tests haven't been done.. for instance maybe a lot bicyle tires with puncher restistance will do relatively even worse on rough roads.. will some rubber compounds fair worse on rought roads, forgetting about the casing for a minute.. will the lack of realistic cooling due to wind in these test skew results with some rubber compound able to deal with heat better... etc, etc.. who knows?