Rolling resistance of wide tires

This is basically the math behind what I've been trying to say. The width and the length of the contact patch do not increase equally under load. The width increases much faster:

"Contact patch width is a polynomial (order 2) function of wheel load while there is a linear relationship between tire contact length and wheel load as well as between tire deflection and wheel load."

They have some nice graphs as well in the paper that show this relationship. Basically the slope of the graph of contact patch width vs load is much steeper initially, and then levels off whereas length increases slower and more linearly. This makes intuitive sense, the width of the tire is much flatter so it will make contact with the ground much quicker under increasing load than the curved length.

https://www.researchgate.net/publication/277978476_Relationships_among_the_contact_patch_length_and_width_the_tire_deflection_and_the_rolling_resistance_of_a_free-running_wheel_in_a_soil_bin_facility/link/5812e70a08ae29942f3e9669/download

psi wrote:This is basically the math behind what I've been trying to say. The width and the length of the contact patch do not increase equally under load. The width increases much faster:

"Contact patch width is a polynomial (order 2) function of wheel load while there is a linear relationship between tire contact length and wheel load as well as between tire deflection and wheel load."

They have some nice graphs as well in the paper that show this relationship. Basically the slope of the graph of contact patch width vs load is much steeper initially, and then levels off whereas length increases slower and more linearly. This makes intuitive sense, the width of the tire is much flatter so it will make contact with the ground much quicker under increasing load than the curved length.

<a href="https://www.researchgate.net/publication/277978476_Relationships_among_the_contact_patch_length_and_width_the_tire_deflection_and_the_rolling_resistance_of_a_free-running_wheel_in_a_soil_bin_facility/link/5812e70a08ae29942f3e9669/download" target="_blank"> https://www.researchgate.net/publication/277978476_Relationships_among_the_contact_patch_length_and_width_the_tire_deflection_and_the_rolling_resistance_of_a_free-running_wheel_in_a_soil_bin_facility/link/5812e70a08ae29942f3e9669/download</a>

An Iranian test. In a Spanish agricultural journal? that’s the best you can do?
But OK, now the real disqualifiers.
Of a single size of a low-profile, car tire in soft soil? This shows what?
They varied the size of the contact patch by changing the load and the inflation pressure. OF COURSE the rolling resistance and the contact patch changed when they put more weight on it, or added or released air.
This has literally no pertinent information in it for us. And it’s not even good science for what THEY were trying to do, as the shape of the tire changed with those adjustments and the different shape, not just the contact patch, would have changed rolling resistance. And an improperly inflated tire is always going to have higher rolling resistance.

Our subject is different tires, under the same load, properly inflated. Not the same tire, under different load and pressures ranging from too low to proper to too high.
And there is no explaining away the tests showing the wider contis roll faster, even on BTRR’s smooth drum.

And BTW, yeah, at some level of softness in the surface... def in soft sand.... contact patch is going to go up for a wider tire, but rolling resistance will be lower for a wider tire. There is a reason the g-one speed liteskins beach racing tires I have are huge.

thumper88 wrote:
An Iranian test. In a Spanish agricultural journal? that’s the best you can do?

I'm sorry but if this is what you're resorting to, you're basically admitting you've lost the argument. This paper clearly shows exactly how the contact patch changed under load. It grew quicker widthwise than lengthwise.

This basically means that in most scenarios with varying loads and inflation pressures, a wider tire is going to have a larger contact patch. This is what I was trying to show by using different examples of wider car tires. The polynomial relationship between load and width means that you basically get the width of the tire as a contact patch almost immediately with any load. Then the patch grows linearly from there. So by using a narrow tire you are limiting the contact patch as it doesn't have much free width to use.

My main point is that we should really look at what people are doing in practice. Racers use narrow tires. Ebike riders use massive tires. It seems like as soon as you give people an extra 300 W motor they will ride a fat bike tire all day long. If they have to roll it under their own power they quickly go back to relatively narrow rubber. Upon some digging the math confirms that this is rational behaviour.

I’ve been made to read an Iranian farm study... so I think that’s about as far as I want to go with the discussion.
In the real world, there is still that Conti test.
And there is also Dylan Johnson, the most data-driven gravel racer  alive, and he stays at the widest end of the range that makes sense for any given course. Mostly he’s on 42s that measure out to just under 45mm true, but occasionally pushes out to 45’s that are closer to 47... And for his size and endlessly-measured output, he gets the results.
Cycling equipment choice is of course notoriously driven by peer pressure, tradition and gut...and again and again gut is proven wrong.

Ok this University of Wisconsin study is definitely going to put this discussion to bed. It actually measured the size of the contact patch of various width mountain bike tires at different pressures. You can clearly see in the graph that the widest tires (orange is a 4.0" and yellow is a 3.0") are basically at the top of the graph. If you take the size at equivalent pressure for the tires, maybe something like 12 psi for 4.0, 15 for 3.0, 20 for 2.5, the differences are actually massive. Maybe close to 2x bigger for the 4.0 tire.



https://www.researchgate.net/publication/341129756_Characterization_and_Modelling_of_Various_Sized_Mountain_Bike_Tires_and_the_Effects_of_Tire_Tread_Knobs_and_Inflation_Pressure

Most of the 4 inch “contact patch” in the study isn’t actually touching the ground. They define “contact patch” differently.
This study doesn’t prove, or disprove, anything we’re talking about. It just confuses.

But first, please address your initial claim that a larger contact patch makes gravel tires slower. Can you do that? 
Contact patch per se has little to do with rolling resistance. Rolling resistance comes from the energy required to bend the carcass. It’s lower, for the same carcass layup schedule, the wider a tire gets because you are bending less of the tire.

As to mocking me for noting the previous study was Iranian and in a Spanish farming journal, no, raising that doesn’t mean I’ve “already lost the argument.”
When evaluating how seriously to take the results of any study, you must always consider the source of the study, and the type of scientific journal in which it was published. In this case, even THIS journal apparently took two years before it decided to publish the study.
You also consider the methodology. Mouse studies for example, don’t carry a lot of weight. Mouse studies in minor journals essentially NO weight. A car tire used to, apparently, simulate tractor tire behavior....I’m shaking my head. It’s about like a mouse study. Maybe something useful to point another researcher at a weightier topic or more rigorous approach. For tractors. It’s about 5 layers away from anything useful for bike tires.

Now, back to contact patch. I used to read scientific studies for a living. You need to find one that is on point. Look at how this study defines “contact patch.” 
They’re looking at knobby MTB tires and mainly are not measuring the part of the tire that is in contact with the ground for what they are calling a “contact patch.”
They’re looking at esoteric things like “void ratios” between the knobs and counting void as what we would call “contact." And these contact patches are mostly void.
Throw the variations in pressure into the mix and you still don’t have something directly pertinent.
Look at the amount of rubber actually contacting the ground in the diagram. It’s not significantly different and just is no way to measure contact patch the way we are thinking about it.
The widely-spaced nobs of course deform the tires more until enough knobs are touching to bear the weight.
Read the patter around this image. For our purposes this way of looking at tires opens up all kinds of potential error.

IMO,  many riders are confused about this topic, and many, even pro, riders do not actually choose the right tire as their opinions are colored by the past and by non-scientific "feelings".  Very narrow tires at very high pressures feel "fast" because they bounce all over the place and are sketchy: without actually timing things one gets the impression they are fast.  This is the same as the feeling one gets with a hardtail MTB vs. full suspension on a moderately rough course, the full suspension is usually the fastest choice, but because the rider works more to control the HT, it often "feels" faster.
In road racing, aero properties are much more of an issue than in gravel or MTB, so the width of the tire and how it interacts with the outer width of the rim really matters.  Still in road racing, 28s are becoming pretty standard, when 21s pumped up to insane pressures used to be standard-riders are certainly faster now... on tubeless 28s at much lower pressures.  In MTB, at the World Cup level, 2.35s and 2.4s are common place, at very low pressures-take Nino Schurter for example, he is running the Aspen 2.4 almost all the time.
In gravel, tire choice is very course dependent.  The BWR races are outliers, because they have a greater percentage of road sections, so most competitive folks run a narrower tire than would be ideal for the dirt sections otherwise.
When one considers tires for gravel racing, one has to consider comfort as a factor: if one is going to be racing for 6 hours or more, overall fatigue is an issue, and a more comfortable set up (slightly wider tire at lower pressure) may give advantages later in the race in terms of fatigue: no one is fast if their lower back is screaming in pain from 6 hours of pounding!  I was surprised at Keegan's choice, and pressures, for SBT GRVL, clearly it worked for him, but that set up would never work for me-at 60 years old, I was very happy to be on 40 mm G-one RS tires, at 35/34 PSI, and I was not losing any ground during the road sections, and I passed quite a few folks on the Cow Creek descent, and other descents, who appeared to be really struggling with controlling their bikes (in fact it was a bit scary, as some folks looked like they were about to go down in front of me).
If one is riding longer than a couple of hours, it is important to consider comfort and control as part of the performance equation, and not just RR-and as soon as one gets on dirt, wider and lower pressure is faster to a point-in gravel racing one has to balance that with the percentage of road the course may have, and whether the tactics might dictate being faster on the road sections or the gravel sections.  It is a complex choice, for me, I will err on the side of comfort and control, a crash is also slow, as is a flat!  In gravel racing, tire weight is going to be a total non-issue, the differences are too small to matter, and acceleration is not a factor in a long race like this.  50 grams is nothing.  Plus being able to control your bike and not struggle with it is a lot more fun than being pounded and bouncing all over the place!

I suspect, going forward, we will see more and more wider tires at lower pressures in all types of racing.  Paris Roubaix has already gone this way, so has World Cup MTB racing.  Back in the 90s I raced MTB, and everyone was on 1.95s with narrow rims, at 35-40 PSI, you never see that today.  A wider tire, with as supple a casing as possible (balanced against flat risk) at lower pressure, tubeless, is the best choice for performance.  The longer the race the more comfort/control vs fatigue becomes a real performance issue.  And the larger percentage of dirt sections, the more advantage a wider, lower pressure set up will have.

I prefer real world results. In my world I have never ridden on small diameter textured metal drums with my wheels held in a fixed position.

thumper88 wrote:

But first, please address your initial claim that a larger contact patch makes gravel tires slower. Can you do that? 
Contact patch per se has little to do with rolling resistance. Rolling resistance comes from the energy required to bend the carcass. It’s lower, for the same carcass layup schedule, the wider a tire gets because you are bending less of the tire.

This is basically the essence of my post. I am hypothesizing that wider tires have larger contact patches, and that the larger interface between the tire and the road is part of what makes them have higher rolling resistance. This effect would be something that would be missed when testing the RR of the tire on a metal drum, as the tire would not adhere to the drum nearly as much as pavement. So the BRR tests would capture only part of a tire's rolling resistance, and skew the results to favour wider tires as I would suggest occurred in the Continental GP5k study of different widths.

Despite your attempts to muddy the waters, I believe I've demonstrated that wider tires do indeed have larger contact patches. Yes the tires in this latest study were mainly knobbies, however the contact patches grow exactly the same way as those in the Iranian farm study. In this case they are holding the load constant and varying the inflation pressure. You can see the graph of the length and width do not change at the same rate. The length and width changes exactly as predicted by the other study. In the first study the width grows much more rapidly under load. In this one it shinks much less under increasing inflation pressure. These results point to the exact same phenomenon occuring, namely: width of the tire is much flatter than the length, and so the contact patch width is much more sensitive to load than the length and less sensistive to increased inflation pressure.



So your abstract model of "contact patch always remains the same size under load and only changes shape" is shown to be incorrect by both these studies. You claim to have read scientific studies for a living. Well it's time to read the writing on the wall and consign your pet theory to the dustbin of history.

Me in initial post:
"I think the problem with wide tires that no one mentions is the fact that the contact patch is bigger. You can corner on rails with 40mm tires, but I think that's part of why they roll slower. You have more sticky rubber interfacing with the pavement..."

So now that we are confident that contact patches on wider tires are in fact larger, it remains to be proven that the larger patch of sticky rubber indeed does contribute to higher RR. I would suspect that it does as examples like metal train wheels rolling on metal tracks have extremely low rolling resistance.

Also the effect I mentioned in the first post would point to rubber / pavement adhesion being a part of the RR:

"Leonard Zinn talks about how TT courses are really fast when a bit wet, and I think this is again the same phenomenon. The fact that the tire is not sticking to the road makes cornering quite sketchy but also allows much faster speed in the straights:"

psi wrote:

This is basically the essence of my post. I am hypothesizing that wider tires have larger contact patches, and that the larger interface between the tire and the road is part of what makes them have higher rolling resistance. This effect would be something that would be missed when testing the RR of the tire on a metal drum, as the tire would not adhere to the drum nearly as much as pavement. So the BRR tests would capture only part of a tire's rolling resistance, and skew the results to favour wider tires as I would suggest occurred in the Continental GP5k study of different widths.

Despite your attempts to muddy the waters, I believe I've demonstrated that wider tires do indeed have larger contact patches. Yes the tires in this latest study were mainly knobbies, however the contact patches grow exactly the same way as those in the Iranian farm study. In this case they are holding the load constant and varying the inflation pressure. You can see the graph of the length and width do not change at the same rate. The length and width changes exactly as predicted by the other study. In the first study the width grows much more rapidly under load. In this one it shinks much less under increasing inflation pressure. These results point to the exact same phenomenon occuring, namely: width of the tire is much flatter than the length, and so the contact patch width is much more sensitive to load than the length and less sensistive to increased inflation pressure.



So your abstract model of "contact patch always remains the same size under load and only changes shape" is shown to be incorrect by both these studies. You claim to have read scientific studies for a living. Well it's time to read the writing on the wall and consign your pet theory to the dustbin of history.

Me in initial post:
"I think the problem with wide tires that no one mentions is the fact that the contact patch is bigger. You can corner on rails with 40mm tires, but I think that's part of why they roll slower. You have more sticky rubber interfacing with the pavement..."

So now that we are confident that contact patches on wider tires are in fact larger, it remains to be proven that the larger patch of sticky rubber indeed does contribute to higher RR. I would suspect that it does as examples like metal train wheels rolling on metal tracks have extremely low rolling resistance.

Also the effect I mentioned in the first post would point to rubber / pavement adhesion being a part of the RR:

"Leonard Zinn talks about how TT courses are really fast when a bit wet, and I think this is again the same phenomenon. The fact that the tire is not sticking to the road makes cornering quite sketchy but also allows much faster speed in the straights:"

The instinct to immediately suspect anything that comes from Heine as being questionable is a good one. The instinct to question BTRR and his (large)  rollers... and the use of rollers generally and in rolling resistance tests is also a good one. It’s not great science. But its perhaps the only way to take an affordable stab at it. Results from well-thought-out roller tests -- like blather about bikes, or BTRR -- generally gybe with IRL tests. So they give useful information but not precise.
BTRR also has other issues, mixing in regular tubes for most tests, and using wheels that are way narrow by modern standards. But he’s kind of locked into it, having built a large database using the same methods. It’s better than nothing, and I’m glad he is doing it.
But rollers...Witness the just nutty results the GCN people claimed from their “gravel tire test”... just insane. Anyone who has ever ridden Conti Terra Speeds can tell you they dont cost an extra 75w or whatever over road tires. Their issue seemed to be mainly from using rollers.

But look...the contact patch just isn’t a serious part of rolling resistance for bikes.
Again, the rolling resistance comes from the energy required to deform the tire. THAT is why the railroad car wheels you cite have almost zero rolling resistance. It’s because they dont deform as the roll. Not because of their contact patch.
This is the same reason track tires, which are run on the smoothest of surfaces and don’t need the kind of give gravel tires need,  are pumped up to insane levels that essentially make them solid. They behave exactly like railroad wheels and have negligible rolling resisntance.
Zinn’s comments about wet pavement on TTs aren’t based on any data. The effect is almost certainly minor if real. And if real, it almost certainly has more to do with the water in tiny indentations refusing to compress -- water won’t -- and essentially making the pavement smoother, not less sticky.
Its obviously not true for standing water, prob just for a degree of dampness that is flush with the pavement surface or just lower.
Remember, Zinn is the same expert on physics who said a bike throw couldn’t work.

Generally, a wider gravel tire on rough surfaces will be faster. How rough, how wide? Those are the right questions. 
On smooth hard surfaces, not so much because wider heads in direction of less aerodynamic, unless it’s, say, a smooth file tread type compared with a bit smaller tire with nobs on the edges, which prevent any semblance of laminar flow.

"Again, the rolling resistance comes from the energy required to deform the tire. THAT is why the railroad car wheels you cite have almost zero rolling resistance. It’s because they dont deform as the roll. Not because of their contact patch"

Just to be clear for all readers here, the above has nothing to do with what happens on a bicycle.  Trains only work with low rolling resistance because the surface the wheels roll on is a polished, perfectly smooth, rail.  If we had a bicycle tire with that small a contact patch, and which did not deform, riding in any kind of dirt, rolling resistance would be very high, as the tire would sink into the dirt, and then continuously have to try and climb the "hill" caused by sinking in.  Some degree of "floatation" is required for low rolling resistance as soon as the surface is less than perfectly hard, this requires a tire width scaled to the surface conditions-hence very narrow tires at very high pressures on the velodrome, where the surface is very very smooth and hard, and wider tires at lower pressures on dirt surfaces.  I was thinking about this on my gravel ride today, even on relatively smooth gravel roads, I was still often bounced around a bit with 40 mm tires at 35/34 PSI, note that anytime one feels an impact up through the saddle or bars, one is losing energy (momentum) due to the vertical vector force, something to consider...  Of course this is not the only factor to consider for maximum speed vs watts, but it is a factor to consider.  How often are you feeling bumps through the saddle, every time you do, there is a loss of energy.

thumper88 wrote:



 The instinct to question BTRR and his rollers... and the use of rollers generally and in rolling resistance tests is also a good one. It’s not great science. But its perhaps the only way to take an affordable stab at it. Results from well-thought-out roller tests -- like blather about bikes, or BTRR -- generally gybe with IRL tests. So they give useful information but not precise.
BTRR also has other issues, mixing in regular tubes for most tests, and using wheels that are way narrow by modern standards. But he’s kind of locked into it, having built a large database using the same methods. It’s better than nothing, and I’m glad he is doing it.
But rollers...Witness the just nutty results the GCN people claimed from their “gravel tire test”... just insane. Their issue seemed to be mainly from using rollers.

Just thought I would point out, for our less educated readers, that there are plenty of very smart cycling people who find diamond plate testing to be very reliable and relatable to real world riding and the people complaining about it the most are often the ones whose tires don't measure up.  The below quote is from Tom A. or the "Blather about bikes" guy in reference to Enve trying to say that diamond plate testing is not accurate.

Tom A's reply:
Everything quoted above is demonstrably false in regards to tires ridden on pavement. When tested at pressures one would reasonably run on pavement (i.e. below the pressure "breakpoint") the roller testing correctly ranks the tire performance. Take it from someone who has actually done the testing and the correlations ;-)

They are repeating supposition and conjecture not based on any actual data.

I've been following Tom A's work for years as well as spend entirely too much on Jarno's Brr site.  If you know where to look, Brr addresses all the issues that you describe. Jarno has run tests with butyl, latex and tubeless tires with precise amounts of sealant.  He has also run tests with multiple tires across many tire pressures with rim of different widths.  If others have not read those tests, here's my trite summary for the gravel crowd:
1.  his standard tube vs. tubeless with sealant is worth between 2-5 watts per tire.  Why such a large variation?  Supple tires suffer more than stiffer tires when tested with a tube.  A tire like the Challenge Strada Bianca Tubeless (which he tested at 19ish watts is probably going to be closer to 14 or 15 watts run tubeless.  Heavier duty tires won't be affected as much by testing with a butyl tube (i.e. they won't be at all/or considerably faster when run tubeless).
2.  Jarno found that there was almost no discernible difference between running an 18m a 22 or a 26mm internal rim width with 35 and 40mm tires.  (For road it was also small, but on much wider mtb tires it started to make a difference, but only 1 watt between an 18 and 26mm rim difference.)  So running an 18mm rim with a 35 to 40mm gravel tire is not a problem (from a rr standpoint).

And please, let's not ever compare GCN to either Tom A. or Jarno.  GCN makes up some ridiculous sh--tufff.  

To contribute my own opinion and experience to the discussion, I don't think width or contact patch in and of themselves matter one bit.  The flexibility of the tire and the rubber compound are what makes a tire faster or slower.  I certainly don't lose any speed running a Conti Speedking Racesport as a rear road tire at 25 psi.  I suspect it is at worst even compared to a GP 5000 on typical pavement and faster on the rough pavement I see most often.  

ccand51997 wrote:

To contribute my own opinion and experience to the discussion, I don't think width or contact patch in and of themselves matter one bit.  The flexibility of the tire and the rubber compound are what makes a tire faster or slower.  I certainly don't lose any speed running a Conti Speedking Racesport as a rear road tire at 25 psi.  I suspect it is at worst even compared to a GP 5000 on typical pavement and faster on the rough pavement I see most often.  

Why is it that the universal consensus in the car tire industry is that narrower tires have less rolling resistance?

"Smaller tire footprint / contact patch

No matter how efficient the tread design and tire construction, rolling resistance is unavoidable. Therefore, decreasing the tire footprint -- the overall surface area where the tire contacts the road -- is a straightforward method to reduce rolling resistance. For example, a 6" wide (155 mm) tire will produce 40% less rolling resistance than a 10" wide (255 mm) tire. This is why you’ll typically see narrower tires fitted to hybrid and electric vehicles as original equipment.
"
https://www.tirebuyer.com/education/rolling-resistance-and-fuel-economy

The whole idea that wide tires have less rolling resistance in the bicycle world seems to almost have arisen as a cargo cult type of mentality. You have a couple contrarians such as Jan Heine who declare that "wide tires are just as fast" based on a few roll down tests he's conducted. Oh and coincidentally he's got a whole wall of massively wide extremely expensive tires to sell you. Also you have Schwalbe who say 'wide tires roll faster' (at the same pressure), which is basically a completely junk test. And they of course have a lot of wide tires to sell you.

My personal opinion is that bicycle technology shifted with disc brakes and all of a sudden we have bikes that can fit really wide tires. A few enterprising salespeople decided to convince cyclists that wide tires roll just as well in order to sell them comfort while appeasing their ego, and here we are. But any professional who is actually paid to ride a bike fast seems to use the narrowest possible tire they can, whether that's WT pros on the road or XC pros on mtb.

This is what Jarno himself says after years spent testing tires: "If you want to go as fast as possible, choose your required comfort level and pick the smallest tire that can provide that comfort for you."

"But any professional who is actually paid to ride a bike fast seems to use the narrowest possible tire they can, whether that's WT pros on the road or XC pros on mtb."

The above is just plain factually incorrect.  In the 90s MTB pros routinely rode 1.95 tires at 35-40 PSI, now WC Pros run 2.35s or 2.4s at pressures often at, or below 20 PSI.  The same has happened on the road, where 23 width tires were the standard at very high pressures, now on the road one routinely sees 25-28 mm tires, and even wider is normal for Paris Roubaix, tubeless, at lower pressures, and speeds are higher than ever.

barrows wrote:"

The above is just plain factually incorrect.  In the 90s MTB pros routinely rode 1.95 tires at 35-40 PSI, now WC Pros run 2.35s or 2.4s at pressures often at, or below 20 PSI.  The same has happened on the road, where 23 width tires were the standard at very high pressures, now on the road one routinely sees 25-28 mm tires, and even wider is normal for Paris Roubaix, tubeless, at lower pressures, and speeds are higher than ever.

I was obviously making a generalization. There are compromises to be made around grip, weight and shock absorption. The XC pros are basically running modern slightly wider tires tubeless, but no one is running plus size mtb tires and expecting to go fast. Likewise at Paris Roubaix the pros increase their tire size by a couple mm, but no one is running 44mm Rene Herse tires despite all the claims that they are just as fast. On normal roads 25s or possibly 28s might be optimal with the typical surface roughness. But I think your pavement woud have to be ridiculously bad to actually have 40s be an optimal choice.

I think the whole wider is just as fast thing is basically marketing BS sold to people who want a softer ride but cannot bear the thought that they are actually losing watts. The same people are buying the ceramic speed pulleys and other snake oil that is routinely peddled to us.

So I've finally found the answer I believe, a scientific explanation of why the interaction between a narrow tire contact patch and the road surface can result in a reduction in rolling resistance. Note that this exactly the opposite of what the bicycle marketing departments have been trying to sell us regarding the shape of the tire contact patch.

"The third way is to change the size and shape of the tyre structure, with a larger diameter and a narrower profile. The principle underlying the narrow ‘eco-tyre’ has been known for quite a long time: energy is lost every time a road surface asperity makes a depression in the tread surface. A long, narrow contact patch will encounter fewer asperities per unit distance travelled than a short, wide one of the same total area, and hence will generate proportionately less rolling resistance.

https://the-contact-patch.com/book/general/g1619-hysteresis-losses-in-rolling-wheels

It really behooves anyone interested in this topic to read the short section. It is at the bottom of the page under the heading Car Tyres.

psi wrote:
I think the whole wider is just as fast thing is basically marketing BS sold to people who want a softer ride but cannot bear the thought that they are actually losing watts. The same people are buying the ceramic speed pulleys and other snake oil that is routinely peddled to us.

I've never bought anything ceramic in my life, other than a pot.  I spend a lot of time studying tires, trying to find what works best for me and I take it out on the road and test it with a power meter.  I don't buy snake oil and I know what certain tires are capable of based off tedious power meter testing.  So far, my own testing coincides quite well with the brr results, so I tend to trust the body of his work.  No one has to take this stuff on the word of others.

ccand51997 wrote:I've never bought anything ceramic in my life, other than a pot.

I have a few drawers full of ceramic capacitors, insulators, and other electronic parts.  But I agree with your point, I have no use for ceramic pulleys or bearings.  I'm too slow and too near the end of life to worry about friction improvements of a few parts in a thousand.

[QUOTE username=psi userid=5752889 postid=1333679947]

Why is it that the universal consensus in the car tire industry is that narrower tires have less rolling resistance? ......This is what Jarno himself says after years spent testing tires: "If you want to go as fast as possible, choose your required comfort level and pick the smallest tire that can provide that comfort for you."

 

[/QUOTE    ]

Because car tires are run almost entirely on hard, smooth pavement. It’s as simple as that.
This goes back to the railroad car wheel or track bike performance. If the surface doesn’t REQUIRE a forgiving, soft tire, the best results will come from running the lightest, hardest most aero tire that makes sense.
An ultra low RR car tire, like they fit to say, a Prius, doesn’t deal with rough surfaces, which want a tire that gives to conform to the lumps and bumps. Try a 23mm gatorskin on real gravel. 
I think Jarno there is trying to offer some yardstick as to what works. The goal isn’t comfort per se but trying to figure out what is right to get the right speed. And if YOU are getting jounced around, your tires aren’t working as well as they could with the power you’re laying down. Your watts are being used to jounce you upward.

At some point, wider doesn’t make sense, but at say Unbound, 45mm or even 50mm works great. 35mm, not so much. The fat tire category goes pretty quickly there for example. But likely is slower than the very fastest setups because of the weight and aero penalties more than being too wide.

In the top gravel pro ranks its not unusual to see a range of as much as 8-12mm in chosen tires, and people being effective at both ends of the scale. Though they may be relying on luck with the narrow choices on some courses. At unbound, you see more conformity, with most at 40-42 and some maybe up to a measured 45mm on tires designated 42.

Yes, there is a cult aspect to going wider and wider. On pavement it gets pretty silly. I ride with people now on 32mm road tires and frankly the pavement just isn’t rough enough to have that make sense. I’m not intererested in the cultspeak that’s out there about suppleness, I’m interested in hard data about what works.

ccand51997 wrote:
Just thought I would point out, for our less educated readers, that there are plenty of very smart cycling people who find diamond plate testing to be very reliable and relatable to real world riding and the people complaining about it the most are often the ones whose tires don't measure up.  The below quote is from Tom A. or the "Blather about bikes" guy in reference to Enve trying to say that diamond plate testing is not accurate.

 

Quote:  *From Enve marketing"

Enve was also at pains to point out that rolling resistance tests performed on steel drums are not always indicative of real world performance.In fairness to Enve, it’s not completely wrong in this regard. As steel drums (even ones with diamond plating as used by BicycleRollingResistance.com) tend to be much smoother and more uniform than an actual road, such tests tend to favour higher pressures and stiffer, less supple tyre casings.

Tom A's reply:
Everything quoted above is demonstrably false in regards to tires ridden on pavement. When tested at pressures one would reasonably run on pavement (i.e. below the pressure "breakpoint") the roller testing correctly ranks the tire performance. Take it from someone who has actually done the testing and the correlations ;-)

They are repeating supposition and conjecture not based on any actual data.

I've been following Tom A's work for years as well as spend entirely too much on Jarno's Brr site.  If you know where to look, Brr addresses all the issues that you describe. Jarno has run tests with butyl, latex and tubeless tires with precise amounts of sealant.  He has also run tests with multiple tires across many tire pressures with rim of different widths.  If others have not read those tests, here's my trite summary for the gravel crowd:
1.  his standard tube vs. tubeless with sealant is worth between 2-5 watts per tire.  Why such a large variation?  Supple tires suffer more than stiffer tires when tested with a tube.  A tire like the Challenge Strada Bianca Tubeless (which he tested at 19ish watts is probably going to be closer to 14 or 15 watts run tubeless.  Heavier duty tires won't be affected as much by testing with a butyl tube (i.e. they won't be at all/or considerably faster when run tubeless).
2.  Jarno found that there was almost no discernible difference between running an 18m a 22 or a 26mm internal rim width with 35 and 40mm tires.  (For road it was also small, but on much wider mtb tires it started to make a difference, but only 1 watt between an 18 and 26mm rim difference.)  So running an 18mm rim with a 35 to 40mm gravel tire is not a problem (from a rr standpoint).

And please, let's not ever compare GCN to either Tom A. or Jarno.  GCN makes up some ridiculous sh--tufff.  

To contribute my own opinion and experience to the discussion, I don't think width or contact patch in and of themselves matter one bit.  The flexibility of the tire and the rubber compound are what makes a tire faster or slower.  I certainly don't lose any speed running a Conti Speedking Racesport as a rear road tire at 25 psi.  I suspect it is at worst even compared to a GP 5000 on typical pavement and faster on the rough pavement I see most often.  

Yes, to all this, with noting that I would not faintly compare GCN  “testing” to BTRR or Blather. I’m not going to go back and look for the exact number but they were saying in their lab the Terra Speeds -- which are some of the fastest-rolling gravel tires available -- were costing something like 75 watts. It is just insane, as anyone who has ever ridden them can tell you...and prob because they were bending them over a tiny roller as opposed to a drum.
Tom is right that a big enough drum will rank them correctly for the most part. I’ve done a few roll-downs myself just out of curiosity and BTRR does get results that correlate with IRL. I just wish for more precision, that’s all. And the drum isn’t perfect, its just the best viable tool.
I do get that Jarno does try to fill in the blanks by extra “special” testing wheel width, latex etc. And much of what you need to know about the differences you can figure out by poking around those additional tests and thinking it through. But ... for example many people just go to his chart and pick the listing with the lowest watt number and say, "oh, tubeless is the way to go. Its always faster, so I guess I need to suck it up on using sealant, mounting issues etc."
When in truth, if you use a latex tube in the fastest tubed versions they’re on par.
And no, I dont think contact patch size is particular relevant to gravel tire speed. The tire engineers will flatly tell you - -and they DO test this -- that the compounds make a startling difference in rolling resistance. And of course the casing as well. But for purposes of talking here about the physics of tire size, we are of course trying to think about tires with identical construction but at different widths.