Sunday, January 22, 2017

Getting Caught Up II's been awhile. Lots of things happening in the last year.

Anyway, It's time to get caught up. Although this tire rolling resistance data was published last spring after I did a collaboration with Jon and Chris of Flo Cycling (see report here), I have been remiss in adding it to the spreadsheet linked to on the right side of this blog. These are tires I roller tested for the Flo tire aero study, models of which I had not already tested. Here are the additions from that testing:

Continental GP4000SII   25C = .0031, 28W for pair @ 40kph
Continental GP Attack   22C  = .0033, 31W for pair @ 40kph
Continental GP4000SII   23C = .0034, 31W for pair @ 40kph
Schwalbe One Tubeless 25C = .0037, 34W for pair @ 40kph
Schwalbe One Tubeless 23C = .0041, 38W for pair @ 40kph
Felt TTR1                      23C = .0048, 45W for pair @ 40kph
Continental Gatorskin     25C = .0048, 45W for pair @ 40kph
Continental Gatorskin     23C = .0052, 48W for pair @ 40kph

The interesting points in there for me are the confirmation that the GP4000SII rolls the same as the previous GP4000S, plus how poorly the Gatorskin models roll at 17-20W worse than the GP4000SII for a pair at 40kph. Wow.

Additionally, in August of 2016, I finally got around to testing a pair of tires that Eric Reid had sent me. One was a brand new model of the Continental Force tire (I had only previously tested a lightly used one) and also a Continental GP TT tire. The latter is a tire that hasn't had much test data on it, so it's something I really wanted to see. Here are those results:

Continental GP Force   24C = .0030, 27W for pair @ 40kph
Continental GP TT       23C = .0028, 26W for pair @ 40kph

That result for the GP TT makes it the new "top dog" for brand new tires I've roller tested. One caveat on that tire measures much larger (24.6mm) than it's rated 23C on my narrow Mavic Open Pro rim, or nearly 2mm wider than a Continental SuperSonic 23C (22.8mm) on the same rim, and is only .0001 lower Crr (~1W difference at 40kph for a pair, or what I consider "tied").

Finally, in October of last year, I roller tested a couple of other tires. One was a newer version of the Continental 20C SuperSonic. I was interested to see if, like some of Conti's other tires, it had gotten any faster since I had last tested that model in 2012.  It did. Here are the results:

Continental SuperSonic 20C = .0030, 28W for pair @ 40kph

That's a fairly significant change from the previous measurement of .0034 for that tire, and corresponds to an improvement of ~3W @ 40kph for a pair, and is just as fast as many tires of MUCH greater width.

The other tire I tested in October was the Specialized S-Works Turbo Tubeless 26C model. This tire was interesting to me because it had been getting some "buzz" about how it was a super-fast tubeless tire (most aren't up to this point).

In this case, I tested it both with a latex tube inside, and then also set up tubeless, with Orange Seal sealant inside (~20ml). Both tests measured nearly exactly the same (within less than .0001 Crr) with the result being:

Specialized S-Works Turbo Tubeless 26C = .0032, 30W for pair @ 40kph

Although that's relatively fast for a tubeless tire, it's not the world-beater it had been hyped to be...especially considering that it's mounted width on my Hed Jet+ wheel for that testing was nearly 30mm!

Anyway...good to be back at it, and I've got some other fun stuff (not so much tire related, hopefully) to be sharing with all of you shortly. Again, all of these updated entries are in the spreadsheet link in the upper right of this page.

edit 23Jan2017: After roller testing a newer version of the Continental SuperSonic in 2016 as described above, I decided to use the newer value in calculating the total power for the H3/Conti 20C SS combination in the chart shown in my last "Win Tunnel Playtime" post. With those changes, the chart looks as follows, and it appears the old H3 has some pretty good speed in it still with that tire:


  1. Hi Tom, thanks for the update! I'd love to see testing for the GP4000S II 28s to see if the "wider is better" trend continues, also the Compass brand tires which are known for being supple and fast (

    1. From what I've seen the 28C should be "better" in regards to Crr, but is going to start causing more aero drag, especially on aero wheels.

      I'm running some Compass tires myself (26"X1.8") on my "all-road" bike. I haven't roller tested them, but they don't feel "slow"...

    2. If you ever end up testing those compass tires for rolling resistance I'd love to see the results. They're not tires for "racer types" but they sure feel supple.

  2. As always, thanks for your testing!!! Would love to get one of the new SS into your hands and see how it rolls.

  3. When we figure in aero dynamics of extra 2mm, or 3mm in some cases, what does that mean. Is the 25mm GP4000SII faster than the 23mm when looking at aero, as well? Looks like the answer would be yes, but just checking.

    1. The answer is "it depends"...on the particular wheel. One example is the Flo study I linked to in the post. On THEIR wheel, the 25C GP4K2 is ~1W faster total (Crr+Aero) than the 23C out to ~10 degrees

  4. Tom - Send me your email and i'll send you the data set tomorrow

  5. Hi Tom, I'm curious how to adjust these numbers/graphs (and the flo one at when thinking of a lighter rider. If a rider was 70kg vs 90kg, how big a difference (if any) would you expect that to make when comparing a 23mm 4000 to a 25?

    Additionally, when testing, how do you choose what PSI to use or is it always constant? If someone was running a 23 they'd generally run a higher pressure then a 25, so I'm curious if you do that as well.

    Thanks for all your work!

    1. Nevermind on the PSI question, I see on the Flo page's footnotes it says all of those were ran at 120 PSI. I assume that's your standard procedure.

    2. Yes, that's my standard for the testing. However, the values reported for the expected on-road Crr and power requirements are estimates for what one would expect with typical pressures on typical other words, a fairly rough estimate. I wouldn't try to put a very fine detail on those...and is the reason I only report those estimates in whole watt values :-)

  6. Hi Richie, since the definition of Crr is literally the ratio of retarding force over load, one can simply reduce the power requirements proportionally for a lighter rider. Just make sure you use the TOTAL weight of the rider plus bike when you do that. For example, the estimates I give in the chart are based on a total mass of 85kg, which is typical for myself, but MY mass is usually ~74kg. So, if I wanted to estimate the power requirements for if I weighed 65kg, then I'd need to add on the additional 11kg of bike and kit before taking the ratio. 76kg/85kg = 0.89, so the expected power would be estimated to be ~90% of the values in the chart. Make sense?

  7. hello, do you test Veloflex tyres. Lots of TT riders in the UK use them as they are considered a fast tyre. How do they compare to newer fast tyres?

    1. Hi Jason,
      One of the first tires I tested was a 20C Veloflex Record (it's on the spreadsheet). It didn't fare as well as the best...

      That said, that test was ~6 years ago. I don't know if there have been any changes to that tire in the mean time.

  8. Hi Tom,
    The width has gone up to 22mm, however not sure about construction and materials used.
    I'll have a look at the spreadsheet. Thanks

  9. Hi Tom, how complex would it be to create an interactive model for people to select wheel and tyre combos? There would need to be a lot of assumptions of course, although as above you could possibly add parameter such as rider and bike weight?

  10. An unscientific, but potentially useful data point corroborating your experience with the GP TT - I swapped some GP 4000 S2's for the TT's (both 25 mm) and did a 9.6 mi time trial 45 seconds faster than my first two runs...and ran 10 W lower average power (290 vs 300)! I'm a heavy (210 lb) rider and run at 120 psi

  11. Hello there Tom,
    My name is Chad or cdw from slowtwitch. I've followed your testing for a while but never actually seen your blog. Seeing the results from the new tires, it got me to wondering if my tire choice is a little dated. I've been out of multisport/cycling for four years while I coached my children's high school running teams but now I am starting back in. I've been using a Bontrager Aerowing tire for a while now because it mates so well to my Flashpoint 60 rim. Do you think the aerobenefits of this tire are losing out overall to the modern tires with reduced rolling resistance? What kind of tire would you put on a Flashpoint 60 if you could pick any tire?