25mm vs 32mm tires, unscientific testing
#76
Senior Member
I was reading comment after comment where you were pretending the data that are there are enough and people pointing out a powermeter was missing were wrong. That literally implies you think a powermeter is not needed and would not have provided significantly more insight. It is not disingenuous at all. You are just digging in. Good luck with that, I am not getting involved any further.
Likes For ZHVelo:
#77
Occam's Rotor
#78
Senior Member
I was reading comment after comment where you were pretending the data that are there are enough and people pointing out a powermeter was missing were wrong. That literally implies you think a powermeter is not needed and would not have provided significantly more insight. It is not disingenuous at all. You are just digging in. Good luck with that, I am not getting involved any further.
#79
Senior Member
#80
Senior Member
Join Date: Jul 2009
Posts: 6,502
Bikes: Colnago, Van Dessel, Factor, Cervelo, Ritchey
Liked 8,313 Times
in
3,301 Posts
Bad science is when you lend credence to the results of a bad experiment because they agree with another experiment. The results of a bad experiment should be ignored, regardless of whether they agree or disagree with other experiments.
Likes For tomato coupe:
#81
Senior Member
Join Date: Nov 2013
Location: Richfield, WI
Posts: 766
Bikes: Trek Domane SL7 Disc, Cannondale F29
Liked 350 Times
in
203 Posts
Recalling my years in engineering school a million years ago, even with "accurate" equipment and testing, test results were typically nothing more than an average of values. There wasn't any "one true answer" to anything. The more data I had, the more sure I was of my results (i.e., smaller variances) and lots of data helped to spot trends and discard bad data. If the OP cares to repeat this experiment ~3 billion times, I think we'll be able to come to a reasonable conclusion w/o the power meter.
Likes For PoorInRichfield:
#82
Perceptual Dullard
2. There are ways to measure rolling resistance without a power meter but they all require, at a minimum, known force.
3. I've presented data and a method here to estimate Crr (and CdA) without a power meter but with a known force.
4. There are ways to estimate Crr (but not CdA) without a power meter or even a speed sensor but with a known force.
5. Rides4Beer's ride was nice but he doesn't measure force.
6. I'm not sure what wgscott's null hypothesis is but it appears to be related to rolling resistance. He's claiming that Ride4Beer's data is "pretty close to ideal" for testing whatever his null is, although it lacks any measurement of force. He also claims that replicating the rides 10 times without measurement of force will make the experiment "more robust." I think it will make Rides4Beer more robust.
[Edited to add] Just as not every ride has to be a race, not every ride has to be an experiment. Rides4Beer had a nice ride.
Last edited by RChung; 06-04-20 at 11:35 AM.
Likes For RChung:
#83
Pointy Helmet Tribe
Join Date: Feb 2009
Location: Offthebackistan
Posts: 4,338
Bikes: R5, Allez Sprint, Shiv
Liked 627 Times
in
295 Posts
If you were to do this in an era without a power meter and have results which would be a little more valid and wouldnt make you the butt of jokes on peer review, you would at the very least:
- repeat the experiment multiple times to average out random error (eg, one run being slightly faster than the other)
- you would use a larger sample size
- you would perhaps control for wind by picking days with consistent weather/temperature
- you would do some runs with thicker wheels first and others with thinner wheels first
etc etc
And even then, you would lack the precision that you can get with a power meter nowadays. Going for a less precise method when a better method exists isnt really scientific either.
If you want to argue whether this level of precision is needed - that's a different story. For the purposes of getting a ballpark estimate of how different the two wheels roll, field tests can often be good enough - and this was a pretty good field test. But a near-ideal scientific test it wasnt, as some people are arguing.
Last edited by guadzilla; 06-04-20 at 12:13 PM.
#84
Senior Member
If it's assumed to be suspected that small differences in tire width produce differences in paved performance so large that they're casually-obvious, then what rides4beer is doing may be an adequate demonstration otherwise.
Insofar as the experiment hopes to determine the actual magnitude (or even polarity) of the difference in performance between the two tire setups, it's woefully inadequate.
Likes For HTupolev:
#85
Advocatus Diaboli
Join Date: Feb 2015
Location: Wherever I am
Posts: 8,855
Bikes: Merlin Cyrene, Nashbar steel CX
Liked 1,613 Times
in
1,061 Posts
1. You can't measure rolling resistance in a wind tunnel.
2. There are ways to measure rolling resistance without a power meter but they all require, at a minimum, known force.
3. I've presented data and a method here to estimate Crr (and CdA) without a power meter but with a known force.
4. There are ways to estimate Crr (but not CdA) without a power meter or even a speed sensor but with a known force.
5. Rides4Beer's ride was nice but he doesn't measure force.
6. I'm not sure what wgscott's null hypothesis is but it appears to be related to rolling resistance. He's claiming that Ride4Beer's data is "pretty close to ideal" for testing whatever his null is, although it lacks any measurement of force. He also claims that replicating the rides 10 times without measurement of force will make the experiment "more robust." I think it will make Rides4Beer more robust.
[Edited to add] Just as not every ride has to be a race, not every ride has to be an experiment. Rides4Beer had a nice ride.
2. There are ways to measure rolling resistance without a power meter but they all require, at a minimum, known force.
3. I've presented data and a method here to estimate Crr (and CdA) without a power meter but with a known force.
4. There are ways to estimate Crr (but not CdA) without a power meter or even a speed sensor but with a known force.
5. Rides4Beer's ride was nice but he doesn't measure force.
6. I'm not sure what wgscott's null hypothesis is but it appears to be related to rolling resistance. He's claiming that Ride4Beer's data is "pretty close to ideal" for testing whatever his null is, although it lacks any measurement of force. He also claims that replicating the rides 10 times without measurement of force will make the experiment "more robust." I think it will make Rides4Beer more robust.
[Edited to add] Just as not every ride has to be a race, not every ride has to be an experiment. Rides4Beer had a nice ride.
#86
Senior Member
I think people are talking around each other. If it's assumed to be suspected that small differences in tire width produce differences in paved performance so large that they're casually-obvious, then what rides4beer is doing may be an adequate demonstration otherwise. Insofar as the experiment hopes to determine the actual magnitude (or even polarity) of the difference in performance between the two tire setups, it's woefully inadequate.
#87
Perceptual Dullard
I just saw that, and sent a note to Josh about it. He should have an altitude field in the air density section.
Calculators like this are handy for steady-state approximation if given CdA, Crr, rho, mu, slope, etc. I think I did a calculator like that back for my old bike club back in the last millenium, based on the equations in Whitt&Wilson.
The issue here is the inverse problem: given average speed (but not power, wind, air density, slope, etc.) can you derive what CdA and Crr were? Or, even more dauntingly, can you estimate *the difference in Crr between two tires* given average speed but not power, wind, air density, etc. That's a lot harder. That's sort of like asking, can you estimate the difference in black and white mortality from COVID-19 from two different measurements of population average death rates? (This turns out to be related to a real problem I'm working on right now, which is why it's on my mind).
Calculators like this are handy for steady-state approximation if given CdA, Crr, rho, mu, slope, etc. I think I did a calculator like that back for my old bike club back in the last millenium, based on the equations in Whitt&Wilson.
The issue here is the inverse problem: given average speed (but not power, wind, air density, slope, etc.) can you derive what CdA and Crr were? Or, even more dauntingly, can you estimate *the difference in Crr between two tires* given average speed but not power, wind, air density, etc. That's a lot harder. That's sort of like asking, can you estimate the difference in black and white mortality from COVID-19 from two different measurements of population average death rates? (This turns out to be related to a real problem I'm working on right now, which is why it's on my mind).
#88
Advocatus Diaboli
Join Date: Feb 2015
Location: Wherever I am
Posts: 8,855
Bikes: Merlin Cyrene, Nashbar steel CX
Liked 1,613 Times
in
1,061 Posts
I just saw that, and sent a note to Josh about it. He should have an altitude field in the air density section.
Calculators like this are handy for steady-state approximation if given CdA, Crr, rho, mu, slope, etc. I think I did a calculator like that back for my old bike club back in the last millenium, based on the equations in Whitt&Wilson.
The issue here is the inverse problem: given average speed (but not power, wind, air density, slope, etc.) can you derive what CdA and Crr were? Or, even more dauntingly, can you estimate *the difference in Crr between two tires* given average speed but not power, wind, air density, etc. That's a lot harder. That's sort of like asking, can you estimate the difference in black and white mortality from COVID-19 from two different measurements of population average death rates? (This turns out to be related to a real problem I'm working on right now, which is why it's on my mind).
Calculators like this are handy for steady-state approximation if given CdA, Crr, rho, mu, slope, etc. I think I did a calculator like that back for my old bike club back in the last millenium, based on the equations in Whitt&Wilson.
The issue here is the inverse problem: given average speed (but not power, wind, air density, slope, etc.) can you derive what CdA and Crr were? Or, even more dauntingly, can you estimate *the difference in Crr between two tires* given average speed but not power, wind, air density, etc. That's a lot harder. That's sort of like asking, can you estimate the difference in black and white mortality from COVID-19 from two different measurements of population average death rates? (This turns out to be related to a real problem I'm working on right now, which is why it's on my mind).
Yeah, that would be perhaps more interesting.. have all of the variables, including power be able to be entered in, except for just one data field (user choice), and the calculator could figure out the missing value. Eg. you could tell the temperature outside by filling everything else in.
#89
Senior Member
Join Date: Sep 2014
Location: Siberia West .. aka Central Wisconsin... USA
Posts: 308
Bikes: 2000 Litespeed Appalachian, 1998 Litespeed BlueRidge.. 1977? Schwinn LeTour 12.2 'Rain Daze'
Likes: 0
Liked 17 Times
in
14 Posts
https://www.bicyclerollingresistance...000-comparison
This guy does some good work.
And.. the 32's are fastest.
This guy does some good work.
And.. the 32's are fastest.
#90
I just saw that, and sent a note to Josh about it. He should have an altitude field in the air density section.
Calculators like this are handy for steady-state approximation if given CdA, Crr, rho, mu, slope, etc. I think I did a calculator like that back for my old bike club back in the last millenium, based on the equations in Whitt&Wilson.
The issue here is the inverse problem: given average speed (but not power, wind, air density, slope, etc.) can you derive what CdA and Crr were? Or, even more dauntingly, can you estimate *the difference in Crr between two tires* given average speed but not power, wind, air density, etc. That's a lot harder. That's sort of like asking, can you estimate the difference in black and white mortality from COVID-19 from two different measurements of population average death rates? (This turns out to be related to a real problem I'm working on right now, which is why it's on my mind).
Calculators like this are handy for steady-state approximation if given CdA, Crr, rho, mu, slope, etc. I think I did a calculator like that back for my old bike club back in the last millenium, based on the equations in Whitt&Wilson.
The issue here is the inverse problem: given average speed (but not power, wind, air density, slope, etc.) can you derive what CdA and Crr were? Or, even more dauntingly, can you estimate *the difference in Crr between two tires* given average speed but not power, wind, air density, etc. That's a lot harder. That's sort of like asking, can you estimate the difference in black and white mortality from COVID-19 from two different measurements of population average death rates? (This turns out to be related to a real problem I'm working on right now, which is why it's on my mind).
seems like you're the most knowledgable guy in this thread. So let me just ask you this.
I think all these "tire rolling comparisons" are not a bit obsessive by the layman. Unless one is doing a flat and steady state TT, tire rolling resistance means VERY little in real world races. Why? And here's my question to you. Have you or your collegues tested ACCELERATION of a tire? and come away with some quantifcation of the cost of acceleration of using different tire sizes? I ask because, in a race, or even in a spirited club ride, NOBODY is cruising stead-state, people are attacking out of the saddle... and if the tires are squishy.. then the rider is losing major wattage... and losing ground... and we all know that to make up even a few lost positions will cost the rider dearly over the course of a race or ride. I, personally, have found that the larger the volume of the tire, the harder it is to accerlerate the bike to speed,.. it does me NO GOOD if I can be "efficient and comfortable" at cruising speed but lost everything when the attacks come because I cannot keep up... and once you've lost the draft.. it's game over,.. you may resume rolling steady state by all by yourself.
So, I ask, what is your take on this matter of prioritizing acceleration over steady-state rolling efficiency.
#91
Senior Member
Join Date: Jul 2009
Posts: 6,502
Bikes: Colnago, Van Dessel, Factor, Cervelo, Ritchey
Liked 8,313 Times
in
3,301 Posts
https://www.bicyclerollingresistance...000-comparison
This guy does some good work.
And.. the 32's are fastest.
This guy does some good work.
And.. the 32's are fastest.
#93
Senior Member
Join Date: Jul 2009
Posts: 6,502
Bikes: Colnago, Van Dessel, Factor, Cervelo, Ritchey
Liked 8,313 Times
in
3,301 Posts
#94
Senior Member
https://www.bicyclerollingresistance...000-comparison
This guy does some good work.
And.. the 32's are fastest.
This guy does some good work.
And.. the 32's are fastest.
#95
Perceptual Dullard
And here's my question to you. Have you or your collegues tested ACCELERATION of a tire? and come away with some quantifcation of the cost of acceleration of using different tire sizes? I ask because, in a race, or even in a spirited club ride, NOBODY is cruising stead-state, people are attacking out of the saddle... and if the tires are squishy.. then the rider is losing major wattage... and losing ground... and we all know that to make up even a few lost positions will cost the rider dearly over the course of a race or ride. I, personally, have found that the larger the volume of the tire, the harder it is to accerlerate the bike to speed,.. it does me NO GOOD if I can be "efficient and comfortable" at cruising speed but lost everything when the attacks come because I cannot keep up... and once you've lost the draft.. it's game over,.. you may resume rolling steady state by all by yourself.
So, I ask, what is your take on this matter of prioritizing acceleration over steady-state rolling efficiency.
So, I ask, what is your take on this matter of prioritizing acceleration over steady-state rolling efficiency.
#96
Batüwü Creakcreak
I just saw that, and sent a note to Josh about it. He should have an altitude field in the air density section.
Calculators like this are handy for steady-state approximation if given CdA, Crr, rho, mu, slope, etc. I think I did a calculator like that back for my old bike club back in the last millenium, based on the equations in Whitt&Wilson.
The issue here is the inverse problem: given average speed (but not power, wind, air density, slope, etc.) can you derive what CdA and Crr were? Or, even more dauntingly, can you estimate *the difference in Crr between two tires* given average speed but not power, wind, air density, etc. That's a lot harder. That's sort of like asking, can you estimate the difference in black and white mortality from COVID-19 from two different measurements of population average death rates? (This turns out to be related to a real problem I'm working on right now, which is why it's on my mind).
Calculators like this are handy for steady-state approximation if given CdA, Crr, rho, mu, slope, etc. I think I did a calculator like that back for my old bike club back in the last millenium, based on the equations in Whitt&Wilson.
The issue here is the inverse problem: given average speed (but not power, wind, air density, slope, etc.) can you derive what CdA and Crr were? Or, even more dauntingly, can you estimate *the difference in Crr between two tires* given average speed but not power, wind, air density, etc. That's a lot harder. That's sort of like asking, can you estimate the difference in black and white mortality from COVID-19 from two different measurements of population average death rates? (This turns out to be related to a real problem I'm working on right now, which is why it's on my mind).
Kudos.
#97
Senior Member
https://www.bicyclerollingresistance...000-comparison
This guy does some good work.
And.. the 32's are fastest.
This guy does some good work.
And.. the 32's are fastest.
A larger tire at the same pressure is a harder spring. Just try riding a 40mm tire at 80 psi; it's going to be a total boneshaker. A 23mm tire at 80 psi is going to offer a hell of a lot of cushioning (unless you are fat, in which case it will pinch flat and you should ride fatter tires).
We know overinflated tires have the least rolling resistance on a smooth drum test, but in reality on the road it doesn't quite pan out. When the tires are inflated so that they offer a similar amount of suspension, the rolling resistance evens out. What's left is tire aerodynamics, which doesn't matter that much in the pack but you're not going to see triathletes or timetrialists on 28 or 32mm rubber; 23/25 F/R mix is the norm.
If tire width was such a rolling resistance win, you wouldn't expect Continental to make their GP5000 more narrow than GP4000s at the same nominal width
#98
Senior Member
This is true, but with respect to surface irregularities, I wonder if it doesn't get overstated. At a given pressure, the vertical spring rate of a tire appears to change much less with respect to tire width for small deflectors than for pressing a wheel against a flat surface, at least in the static case. So the method that BRR uses to equalize tire pressure for comfort - seeing how far a tire drops against a flat surface - potentially overestimates how much lower the tire pressure needs to be in the wider tire to satisfy the suspension needs of a given use case.
It would be interesting to see real-world measurements characterizing how the suspension breakpoint on various surfaces changes with respect to tire width.
It would be interesting to see real-world measurements characterizing how the suspension breakpoint on various surfaces changes with respect to tire width.
#99
Perceptual Dullard
The connection to cycling is that we have large aggregative measures of speed and power and we're trying to indirectly back-out plausible values for rolling and aerodynamic resistance. The speed-power model is pretty well understood, so there are parts of that model that constrain what those plausible values are. Cycling data, no matter how lousy, are cleaner than the data I normally work with, so applying analytic methods developed for crappy data to relatively clean data turns out to work pretty well.
Likes For RChung:
#100
Senior Member
Join Date: Sep 2014
Location: Siberia West .. aka Central Wisconsin... USA
Posts: 308
Bikes: 2000 Litespeed Appalachian, 1998 Litespeed BlueRidge.. 1977? Schwinn LeTour 12.2 'Rain Daze'
Likes: 0
Liked 17 Times
in
14 Posts
Yes.. true. Yet suggest that specific test w same flavor rubber significant. I can tend to agree drum data not real world in many respects.
Any way rolling data is sliced.. from whose perspective.. the needle is going to tilt. I test rolling off a minor hill... was very surprised when the 32mm
Ultra ll's 48/60psi set up tubeless where near the S Pro Ones 25mm. Yet my test is very limited view... S Ones on long rides hot days give much higher average speeds. But IMO that is just a 'weather report'.