FLO Cycling – A2 Wind Tunnel Tire Study Part 2

In Part 1 of this series, we looked at the aerodynamic drag created by different tires on a front FLO 60 Carbon Clincher wheel.  As many of you know, when searching for the fastest tire, aerodynamic drag is only half of the equation.  In order to know how much power is required to rotate your tire,  you must also consider it’s rolling resistance.  Only when you combine the wattage required to overcome both the aerodynamic drag and the rolling resistance, can you tell which tire is the fastest.  

To calculate the rolling resistance generated by each tire, we sent our box of tires to Tom Anhalt from Blather ’bout Bikes.  Tom is a well respected engineer, and a wealth of knowledge on all things cycling.  Tom has his own rig for testing rolling resistance, and it was perfect for our application.  After Tom was finished calculating the rolling resistance of each tire, he calculated the total wattage each tire consumed.  

Let’s get to the numbers!

Step 1: Calculating Rolling Resistance

To calculate the rolling resistance values below, we used the following test procedure.

  • – Bike Speed 35km/h or 21.75mph
  • – Wheel load 45kg or 99.21lbs
  • – Rolling resistance numbers were measured on a Mavic Open Pro wheel at 120 psi. *

* Tom noted the following: “I’ve found that the Crr on the (smooth) rollers of a tire the Mavic Open Pro wheel at 120psi, is approximately equivalent to the Crr on a 21mm internal width wheel at 100psi.”

Below are the rolling resistance values and the power to overcome that rolling resistance for each tire.  They are listed in order from fastest to slowest.

Step 2: Combining Aerodynamic Drag, and Rolling Resistance

For each value plotted on the graph, a tire’s aero drag and rolling resistance values were converted to Watts and added together.  This combined value represents the total wattage required to power the tire.  

Note: When combining the both the aerodynamic drag values from Part 1 of this series and the rolling resistance values, it is important to note that beta correction of the data was completed by Tom.  Tom’s Beta correction calculation assumes that the ambient wind is at 90deg to the direction of travel (pure crosswind) for the resultant yaw angle for a given bike speed.

Because there are so many tires on one chart, we felt is was best to display this data on an Interactive chart.  

Step 3: Applying our Net Drag Reduction Value Formula

So what does all of this data mean?  Which tire is truly the fastest?  At FLO we have developed a Net Drag Reduction Value (NDRV) formula that gives a cyclist a realistic estimate of how much time a wheel will save them in the real world.  This formula was developed using the 110,000 real world data points we collected in our 5 step design process.  Since we know how much time a cyclist spends at each yaw angle on average, our NDRV formula can calculate a weighted average power value across all yaw angles.  In the table below, we have listed the tires from fastest to slowest based on the numbers generated by our NDRV formula.  Keep in mind that the lower the power number, the faster the tire.  Wattage numbers are for one individual tire.

I hope you have enjoyed this article.  Please leave your comments and questions below.  I’d like to say a special thank you to Tom for all of his help with the Crr data.

Take care,


Join the Conversation


  1. Hi Chris, what type of surface were the tires tested on? Average asphalt, smooth, other?
    Like tire pressure it makes a difference, right? Tks., Raul

  2. Great data. Onset of stall is clearly visible as increasing drag when yaw angle increases past 12-15 degrees. Maybe scope for Iimprovement by tricky casing texture to energise the boundary layer and delay lateral stall.

  3. Although the tires were tested on smooth rollers, the Crr values are adjusted to represent what would be expected on an "average" asphalt surface by using a 1.5X multiplier.

  4. Thanks for doing this.

    One of the advantages of running a 25mm tire vs a 23mm is the ability to drop the pressure and increase comfort. It looks like this isn't taken into account in this particular data set?

    How might the aero and rolling resistance change for the 25 change if one were to drop the pressure by 10 psi?

    Thanks again!


  5. Damen,

    While we didn't specifically test how the aerodynamics of a 25mm tire change with changes in PSI, we did test that effect on a 23mm tire. You can see that study here: http://flocycling.blogspot.com/2016/05/flo-cycling-does-tire-pressure-change.html

    With respect to how changes in pressure effect your rolling resistance, I'd read this article with Josh Poertner from Silca. It's a great piece: http://www.trstriathlon.com/talking-tires-with-joshua-poertner/

    Please let me know if you have any additional questions.

    Take care,


  6. For what it's worth, if you look at TomA's blog he's done work in the past correlating his smooth roller data to data he collected outside. There's a very strong correlation. Basically the rollers just provide a controlled environment to tease out differences in the crr from different tread compounds and casing construction.

  7. Hi, great data and article. If the speed was closer to 30mph would the results change? I ride TT with typical speeds of 28MPH plus. Mark.

  8. Mark Denney,

    Things do change as the rider's speed changes and as the rider's weight changes. We may be be able to build something interactive that allows people to adjust those variables. I'll see if it's possible. That said, as speed changes the same Continental tires are still at the front with respect to Crr, and aerodynamics. I think the end results will be quite similar.

    Take care,


  9. I know you can't include every tire, but why not include the Schwalbe Pro One here? It's widely available and a completely different construction to the previous version. No one is buying the original One, and the Pro One is a direct competitor to the Turbo and GP4000II. Same problem happened on the VN test.

  10. tony,

    I'm sorry you are disappointed with our selection of tires. At the time of our testing that tire was just announced and not readily available. After polling a list of our users, that tire did not come up, but the Schwalbe One did many times. Many of our customer were using the Schwalbe One. We tested what we had available to us at the time.


  11. Unknown,

    Can I ask you to clarify what you mean by "standard clinchers"? Are you asking what tube was used for all of the Crr testing done by Tom?



  12. Understand the crr for the force is 0.29. But some testing shows 0.34. However the force is more aero than the gp4000s and having the same crr 0.34. So can I say that the force is still faster than the gp4000s even if it has the same crr as the gp4000s?

  13. It's interesting how the 4000s IIs continue to do so well in your tests. I know it's standard to do these comparisons based on the manufacturers' reported sizing, but it seems like it'd be valuable to include actual sizes as well. For instance, the 4000s II 23mms are probably about the same size as S-Work Turbo 24mms, and the 4000 25mms are likely around the same size as S-Work 26mm.

    Given both variants of the 4000s beat the S-Work, I wouldn't expect any of the standings to change, but given how large those tires actually are it seems like a better comparison to do larger versions of other tires as well.

  14. > Can I ask you to clarify what you mean by "standard clinchers"? Are you asking what tube was used for all of the Crr testing done by Tom?

    I'm not sure what Unknown was asking, but yes, knowing if these were ran with butyl or latex tubes would be informative. In theory if these were done with butyl, doing them again with latex should drop all the crrs. That should not affect the relative rankings though, so not sure how useful that information is.

  15. Stan lee,

    At the current time the latest rolling data we have for the Force is the 0.0029 so we are going to leave it displayed as that. If anything changes, we will update this article.

    Take care,


  16. Another interesting question is how rapidly would the tire CRR increase if the wheel size is decreased?

    The most aero tire combo with the flo wheel was around 14.4 Watts of power loss through aero drag, and 18 Watts of power loss through rolling resistance.

    If one went to 17 inch wheels as per the orgininal moultons bicycles, the aero cross section area would decrease, and assuming CdA goes proportion to area, would save 5.6 watts of aero drag.
    Would be really interesting to see (if different sizes were available) the dependence of rolling resistance on wheel size – maybe smaller wheels could end up giving less drag. (plus smaller wheels mean a tighter pace line spacing and better draft.

  17. Hamish,

    Those are definitely some interesting things to consider. At some point you are limited to how much you can change on the bike. For example, developing a custom wheels size would mean redesigning frames to accept that wheel size.


  18. Is the tested Continental Force "the" Force from the Attack & Force Set?
    So this Tire is intended to be on the backwheel, not on the front? Does this matter?
    Cheers Joe

  19. Jack,

    Thanks for writing. This is the Force tire from the "Attack & Force" set. There shouldn't be any issues with running this on a front wheel.

    Take care,


  20. Maelstorm,

    This started as an aero study. We have studied Vittoria tires in the past and they performed poorly aerodynamically. That reason, combined with the fact that we had few if any requests to run Vittoria tires when we polled our readers, resulted in not testing them.

    Take care,


  21. Many of the time-loss columns in the last table looks wrong to me when comparing to the differences between tires in the watts column. One example: tire no. 4 in the list is 0.39W worse than tire no. 3 on the list and I loose 1 second over an Ironman by choosing tire 4 instead of tire no. 3. By picking tire no. 5 instead of no. 4 I give up another 0.14W, but suddenly I loose 37 seconds more over an Ironman. I know the watt/time-saved relationship is not 1:1 linear, but That just seems wrong. I believe there are more such errors further down the list also, just from eyeballing it. But great to have the watt data! Thanks!

  22. Seems like Conti has the "formula" down for a fast tire, looking at their website they have a TT tire that they indicate as an even faster tire than the Force, any reason you didn't test that or is that the Supersonic with a new name?


  23. DrJ,

    At the time of this test (nearly a year ago now), that list of tires is what our followers asked us to test. From what we could see, the tires we picked were the best/most interesting choices. The TT could definitely be a good tire to test, but we had to draw the line somewhere. Perhaps we will test the tire next time.

    I hope that makes sense,


  24. So obviously these tests were in regard to the front tire. What are the thoughts on the rear tire choice? Are aerodynamics as important/less important? Or is rolling resistance more important?

  25. DrJ,

    I'd still use the tire with the lowest total watts possible for the rear wheel. Different bikes, riders, set ups etc. will change the rear wheel aerodynamics for almost every rider, so it's hard to say exactly what is more important, unless you study a specific rider individually.

    I hope that makes sense.


  26. One issue I have with the combined data(Drag + resistance) the drag was recorded @95psi while the resistance was @ 120psi, so the combined result of you number could never be ridden in reality.

    You should redo resistance test @ 95psi to be inline with the drag test.


  27. Matt,

    The pressure was set to 120psi, because a thinner rim was used. When compared to our wider rims, the 120psi on a thinner rim produces pretty much the exact same results as 95psi on a wider rim.

    I hope the makes sense,


  28. agreif,

    Yes, this is for all new tires. The hardest part about testing tire wear, is you have to test the exact same tire. To do this we would have had to have made multiple trips to the wind tunnel after wearing out a list of 20 tires.

    I hope that makes sense.


  29. Hello,

    Thank you. I found this information very useful.

    I noticed that your aero/speed data/testing uses tires with tubes. Any testing with running these wheels tubeless (combo based or otherwise), and/or any recommended tire options for running tubeless (i.e. FLO recommends the Continental Force as the fastest tire for your wheels, anything for if you run tubeless)??


  30. Andrea,

    Thanks for writing. The Hutchison Atom and the Schwalbe One ran in this test were both tubeless tires. They are both very close to each other in the list so both tires would yield a similar result.

    I hope that helps,


  31. Hello,

    I am just wondering if you guys have ever tested any Michelin tires. I have used the Continental 4000's before but I feel the Michelins cut up less and wear very well.

    Thank you,


  32. Delighted to see this data, because I was already a 25mm GP4Ks fan, and thought I might be sacrificing pure speed performance in favor of ride quality and feeling secure to stay in my aero bars when the ride gets technical.
    Pleased to learn I can have the best of both worlds!

  33. Thomas Hands,

    Thanks for writing. Unfortunately, without specifically testing the FLO 45 with the two tires in a wind tunnel, it's tough to say exactly which tire would be faster on the FLO 45.

    I'm sorry that I can't give you more specific information.

    Take care,


  34. It's known that 25mm tires usually have lower rolling resistance than 23mm tires, and 28mm are even lower.

    The 25mm GP 4000S II was probably already wider than the brake track on the wheels, I'm guessing probably 26-27mm, so 1-2mm wider than brake track.

    28mm will probably balloon out to ~31mm = 7mm wider than brake track.

    So, the big question. Are 28mm tires [especially the GP 4000S II] faster than 25mm tires, even with the ballooning effect of a wider tire on a narrower rim?

  35. Andrew,

    Unfortunately, we did not test the 28mm tire in the wind tunnel or on the rollers, so I can't answer if they are faster. The reason we did not test the 28mm tires is because even the 25mm tires have an issue fitting on a list of road and triathlon frames. The extra wide 28mm tires would have a very hard time fitting on a lot of common bike frames so we chose to emit the tire from our testing.

    Have a great day,


  36. Chris – to follow up – if the frameset clearance is not an issue, is there any reason you would NOT recommend 28mm tires on a set of FLO CC's? I have a 45/60 combo running 23mm but am looking to buy a new frame and keep my wheels, but would love to ride 28's on them. I'm not as concerned about aero/rolling resistance as much as reliability and comfort.

  37. David P,

    You can certainly use the 28mm tires on our CC wheels if you have no frame interference. They will definitely help with comfort.

    Have a great day,


  38. Mark Scanio,

    Thanks for writing. This entire wind tunnel study was performed on a front FLO 60. We rarely use rear wheels when doing wind tunnel tests. The Conti force can be used as both a front and rear tire.

    I hope that helps,


  39. Was the Specialized Turbo 24mm tested the older version with the all red graphics or the current version with the world champion stripes graphics? I was told that the older Turbo version uses the same compound as the current Turbo Cotton!?

  40. Simon Oxenham,

    Thanks for writing. We used the Specialized Turbo 24mm with the "World Champion Stripes" graphics. The sidewall says "Gripton Compound" and the tread pattern looks essentially identical to the Turbo Cotton.

    Take care,


  41. Hi, a great study. Something I have picked up though is that in part 1 of your test the results graph for aero performance shows the 23mm GP4000s to be a faster tyre than the Force. In part 2 the interactive graph show this result as the other way around ? Which is correct and has the correct data been used when combined with rolling resistance to calculate which is the fastest overall tyre ? Thanks

  42. Graham King,

    Thanks for writing. The chart in Part 1 only shows the aerodynamic performance. When you only consider the aerodynamic performance the 23mm GP 4000s II tire is faster than the Force tire.

    Part 2 of this series shows the aero performance combined with the rolling resistance value. Because the Force tire has a lower rolling resistance, it ends up being faster than the 23mm GP4000s II tire when you consider both aerodynamics and rolling resistance.

    I hope that helps.

    Take care,


Leave a comment

Your email address will not be published. Required fields are marked *