I’ve long followed Tom Anhalt’s writings on this topic. He uses rollers for testing, which are vastly smaller than the drum used by BRR and he comes up with similar rankings. I’ve done similar testing myself with rollers and a power meter and my rankings have come up similar to his and BRR, and match reasonably well with my subjective perception of how the tires roll on very smooth asphalt.
It’s an interesting assertion that manufacturers stack the deck with regards to performance on the RR tests. It’s true that a tire that’s wide relative to its labeled width will roll faster at the same pressure on a drum test, but a wider tire is also heavier and I’d say that metric is given at least as much attention, if not more, by the general riding public as rolling resistance tests.
IIRC the GP4K was actually only a moderately fast rolling tire in these tests and significantly slower than the 5K. It wasn’t the most supple tire but it wasn’t some sort of garden hose-stiff tire either. If stiffness were the only factor in determining how a tire fared in drum tests then tires like the Gatorskin would do very well in drum tests and we see that’s not the case. I suspect that part of that much-touted suppleness attribute is not just casing flexibility but also a softer tread compound, which may add grip and a nicer ride at the cost of rolling resistance.
Tom Anhalt has asserted that the Wheel Energy rig - despite adding damping - still only has approximately 10% of the damping of a human body, and this is borne out by their tests of gravel tires where the breakpoint at which rolling resistance increases is much higher than any sensible real-world pressure for those tires. So I’m not sure the damping really adds that much useful information.
I agree that drum tests do provide some useful data as long as you understand the limitations and any quirks of the testing protocol. They are reasonably predictive of hysteresis losses and rolling resistance on smooth pavement, but not predictive of the effects of tire width, pressure, and tread on rolling resistance over varying and rougher surfaces.
This is how you validate the roller testing. Take the same tires and VE field test them and show that the rank order is the same. The percentage differences hold between tires when doing this. This makes sense, because below breakpoint pressures, by definition, the main contributor to rolling resistance is the hysteresis losses in the tire casing and tread, even on rough surfaces. And, as it turns out, roller testing is a very effective method of evaluating those losses.”
Come back and ask again when you’ve managed to reach a level of comprehension that allows you to distinguish between a testing protocol being questionable and a source having bias versus a testing protocol being questionable to suit the sources bias. One is a statement I made and the other is just something you imagined… and until you can comprehend the difference, you’re definitely not gonna comprehend any “informed and reasonable explanation”
Calling out your poor reading comprehension as poor reading comprehension isn’t insulting anymore than you claiming I’ve made statements that I never made was insulting me.
Let’s try this another way… go ahead and quote where I claimed that:
Nonsensical sookery about insults notwithstanding, their protocol does incidentally favor some tires over others but I neither suspect nor assert that to be the intent of their protocol. The smallish 30" drum and poor representation of actual road surface just tends to favor tires with stiffer casings. In the real world though, you don’t really need to look beyond their results to see that there’s a flaw. In the real world, you will not get results that show a consistent and steady reduction in resistance either as pressure increases or decreases. I think it’s Tour Magazine, if I recall correctly, who did some flat surface testing a few years ago that demonstrated this really well. This is also well illustrated in the images I linked regarding breakpoint and was borne out well in on road testing conducted by Poertner, both when he was with Zipp and after he bought Silca.
You spoke of a “deep and endless bias in favor of Conti”
If I reconstruct that sentences’ meaning by translation and re-translation, I get a feeling that both a neutral and a negative meaning is possible: it’s either a tendency due to technical constraints or a willful distortion.
Now that you claim the first meaning, I hope this helps everyone to relax a bit. Unless, of course, we all agree that a shouting match in a forum is just what we needed after too many days with the family (well, I don’t)
You issued a vague and rather inflammatory statement about “deep and endless bias” and despite your protestations you didn’t specify exactly how that manifests itself or provide any proof. Then you expect us to carefully parse your vague statement and it’s on us if we happen to slightly misinterpret what exactly you meant. Now it sure sounds to me like you’re backing off your claim of “bias” towards Conti, and your criticisms of drum testing are just the same general ones others have made.
As I’ve pointed out above, if the tests were biased towards stiffer casings then very stiff, slow rolling tires such as Gatorskins would rank highly, but they don’t. My own theory is that drum tests are actually biased towards hard tread compounds, not stiff casings; but that’s just a theory.
I am truly interested in the topic and did spent few hours reading so still have few hopes to get something out of this but I am not getting your point. @Because_Mechanics mentioned:
a protocol limitation and you seem to agree (on the wheel diameter and the stiff casings at least)
he mentions that the rougher the road, the sooner a pressure increase starts deteriorating overall power consumption and you just pulled a graph backing-up this (inflection @110psi on good road, 100 on coarse and high 50s on what seems almost gravel type).
@tbro21 you mention: "Didn’t recall this bit (“you need to be riding Paris-Roubaix conditions to have those pressures -80psi- starting to cause damping issues (for the best tires)”) so just re-read the test and confirmed it… this is never claimed by either Wheel Energy nor Velonews.
But isn’t it what their paris-roubaix tires test show? a degradation around 65psi for most and 80/90 psi for the bests.
Then what am I missing that makes invalid to get findings comparing one same tire reference at different width, adjusting the pressure for same “comfort” across the 4 sizes (around 80 PSI for the 28mm, so before the inflection point for more classic road conditions than PR) ?
My dude… you’ve already made your poor reading comprehension abundantly clear, there’s no reason to continue working so hard to explain it. Repeating fallacious claims based solely upon your linguistic ignorance and inability to comprehend the written word doesn’t make you incrementally more right, it just makes you repeatedly wrong and calls your integrity into question.
The only thing BRR “illustrates” is a deep and endless bias in favor of Conti and questionable testing protocols.
There are two distinct statements separated by a conjunction, “and,” which exist independently of one another.
Jared, or whatever his name is, demonstrates a consistent bias for Conti… AND
BRR utilizes questionable testing protocols…
See how that works?
And, to be absolutely clear, I’m not in any way backing off either of those statements.
I read what you “pointed out” and, like most of what you’ve typed, I summarily disregarded it as uniformed and generally inaccurate. Favoring stiff casings doesn’t mean every tire with stiff casings will get top marks.
Again, not sure if English is a second language for you and that’s causing your struggles or if you’re a native speaker and just really bad at the nuances of comprehension but favoring one thing doesn’t mean that one thing will always overcome other factors to win out… if that were the case, rather than saying the tests favor stiffer casings, I would have said the test relies only upon casing stiffness in determining the resistance.
No you won’t. The rotational inertia of a bike wheel is a tiny fraction (around 1% IIRC) of the linear inertia. So even if you were just accelerating the bike wheel alone along a piece of road (without the bike and rider) you could hardly tell the difference between mass at the hub and mass at the rim.
The reason is quite simple: bike wheels spin at very low RPMs compared to other rotating system (where rotational inertia can dominate linear inertia)
One way manufacturers could stack the deck is by selecting faster rolling but less grippy compounds. This is often levied at Continental.
However, the deliberately wider tire bit could be done but isn’t - Continental has made each successive generation of it’s tires from the 4000SII to 5000 to new 5000S narrower.
That said, the performance of previous GP5000TL tires on BRR is probably a slight outlier; while aerocoach testing also finds them to be on top of the pile when it comes to rolling resistance of road tires, they differ in how much, and BRR finding them to perform as good as most TT specific tires seems a bit odd.
While it’s possible that variations in protocol and accuracy account for the difference, I don’t think we can assume that all tires labeled the same from a given manufacturer will perform exactly the same. Manufacturing and batch variations are certainly possible, and there have certainly been instances in which manufacturers have made unannounced rolling changes.
I vaguely remember Josh Poertner talking in a podcast about how there were certain manufacturing lines of the GP4000S II that would consistently produce significantly faster tires than others. In the same podcast, he mentioned that Conti confessed to not really having a clue what made the GP4000S II faster than previous models - but I don’t recall if that was about the rolling resistance or the aerodynamics of the tire. If it was the Crr, that wouldn’t sound like Conti would (or even could) bother to stack the deck with regards to this kind of testing.
I’m almost certain that Josh was referring to the aerodynamic properties of the tire in a number of wind tunnel tests. I’m sure Conti does their own rolling resistance testing, and it’s highly improbable that they conducted any wind tunnel testing on their tires, at least at that time. So the aerodynamic performance was likely serendipity.
From what I’ve come to understand is that making tires is a dark art. This at least was the experience from fellows who, under the umbrella of a professional tire manufacturer, made custom tires for a Formula student race car.
There seemed to be a rather big margin between a design and reality.
Yeah that’s one way to think about it. Also, wider (newer) rims are usually more U-shaped vs V-shaped so when the air leaves the wheel it has a gentler profile to reattach to the air on the other side of the wheel.
But when the tire is wider than the rim it creates a ‘shelf’ that creates turbulent air that has a harder time remaining tight on the rim and therefore causes less stability in cross winds.
Im new here but I thought I’d find a little information before I start riding this year on tire widths. I average 20 to 30kph depending on who I’m riding with/solo. My bike came with Vittoria Zaffiro Pros that are 32mm wide on an externally 25mm rim. They feel good on my local treacherous roads which eat tires and axles of cars and bikes. But i do feel slightly slower at high speed vs when i had 23mm tires. I can’t compare speeds because I only ride once a month or so but i was hoping someone would have this info before i buy a new set of tires. Is it seriously detrimental at all speeds or just high speeds? Am I only losing 1 or 2 mph? I wanted 28’s but if they are no better than 30’s I’d rather go with the comfort. I’m kinda shocked no bike manufacturer has run the tests indoor with a controllable AI or robot pedaling at constant speeds to test the aero affects…tire pressure could be tested too. The human element brings in too many variables to be reliable short term sometimes.
At 30kph, you’re not losing much of anything to a 32mm tire. A 28 would be faster strictly in technical terms but in terms of real world, on the road effect, that’s the kind of thing that might be 20 seconds across an hour at those speeds. If you were averaging 30-40kph, the benefits would be substantially more noticeable but under about 32kph average aero gains are really minimal.
I think companies eschew robotic testing specifically for the reason you mention, the human form introduces so many variables that results not using a human form don’t translate well… but that’s purely me guessing as to why they don’t.