FLO Cycling – Wheel Design Series Step 5 – Wind Tunnel Results


In 2014 we sat down to redesign our FLO Cycling wheel line. This five-step design process took 15 months to complete and this blog series covers the design process in detail. This is Step 5 of our five-step design process. To learn more about Steps 1-4, please check out the links below.


Step 5 – Wind Tunnel Results

Looking at the New FLO 90 Carbon Clincher at the A2 Wind Tunnel

Proving Our CFD Results in Real Life

In Step 4 of this series we discussed how we drastically improved the aerodynamics of our rims using computational fluid dynamics (CFD) software. Getting results in CFD is one thing, but to get proof that your designs provide a real-world improvement, you need to test them in a wind tunnel. We visited the A2 Wind Tunnel to put our wheels to the ultimate test.

Looking at the New FLO 60 Carbon Clincher in the A2 Wind Tunnel

Visiting the A2 Wind Tunnel in North Carolina

We took our 2012 FLO 60, FLO 90, and FLO DISC models to the wind tunnel and did a direct comparison to our new 2016 model wheels. The results that you get from a wind tunnel don’t mean much to the reader unless they understand the full testing protocol.  We’ve always believed in fully disclosing our testing protocol, so we’ve listed it below.

FLO Cycling Wind Tunnel Protocol
  • Tare was calculated and removed from all tests.
  • A Mavic Open Pro with 32 round spokes was used as our baseline wheel.
  • Each wheel was swept from 0-20 degrees of yaw, in 2.5 degree increments.
  • The same tire was used for each test.
  • The FLO DISC had the valve cover taped shut.
  • Each measurement was taken twice and averaged. 

  • All tires were inflated to 95 psi and calibrated with a digital gauge.
Setting Tire Pressure to 95 psi with Silca’s “Truth”
Setting Up the FLO 60 Carbon Clincher at the A2 Wind Tunnel

Tire Selection
We used the new FLO 60 Carbon as a baseline to test 20 different tires. After finding the best two tires, we tested all of the new wheel shapes with these tires. The two tires used for testing were the following. 
  • Continental GP 4000 S II in a 23mm size
  • Schwable Ultremo ZX in a 23mm size 

Continental GP 4000 S II on the new FLO 60 Carbon Clincher at the A2 Wind Tunnel

The Results

On top of showing an aero graph of grams of drag versus yaw angle, we have always believed in showing our results in what we like to call a Net Drag Reduction Value (NDRV).  NDRV uses a weighted averaging function that calculates drag saving by considering the percentage of time spent at each yaw angle. From our data collected in Step 1 and our analysis in Step 2 of this series, we have a very good understanding of what type of yaw angles a rider experiences while on the road. Combining this data with our NDRV formula gives a rider a realistic estimate of how much time they will actually save using FLO wheels.  We do this instead of assuming a rider spends 100 percent of their time at a wheel’s fastest yaw angle, and then calculating the times savings.  For a full understanding of NDRV, check out our NDRV Article.  

Looking at the Results at the A2 Wind Tunnel

We have updated our 2012 NDRV formula. In 2012 we thought a rider spent 80 percent of their time between 10 and 20 degrees of yaw. Now we know that a rider actually spends 80 percent of their time between zero and 10 degrees of yaw. This discovery forced us to focus on designing wheels that were more aerodynamic at shallower yaw angles instead of focusing on designing wheels that were fast between 10 and 20 degrees of yaw. Our old estimates for how much time a 2012 FLO wheel would save you aren’t as accurate as they could have been, and we have recalculated them below. We have compared those to the new 2016 models.

Time Saving Estimates and Percent Improvement for Continental GP 4000 S II in 23mm Size
Time Saving Estimates and Percent Improvement for Schwalbe Ultremo ZX in 23mm Size


    Discussing Testing Results and Protocol at the A2 Wind Tunnel

    Final Thoughts on the Wheel Design Series

    When you start any new engineering project, there are always a list of unknowns.  Designing our 2016 wheel line was no different. We knew we wanted to create the fastest wheels we had ever designed, and in order to do so, we knew we had to take an approach that had never before been taken. Over a year ago when we started this process we had an idea, and a lot of hope that the entire plan would come together when we stepped into the wind tunnel in November of 2015. We were confident that our approach could work, but until you get the seal of approval from a wind tunnel, your design process is still just a idea. After seeing the results, we were beyond happy with our accomplishments. We were able to improve the aerodynamics of our best-selling front wheel, the front FLO 60, by over 28 percent, we were able to develop a 45mm wheel that was as faster than our old 90mm wheel, and we were even able to improve the aerodynamics of our DISC wheel, which is not an easy task to accomplish.

    On top of redesigning our current Aluminum + Carbon FLO 60, FLO 90, and FLO DISC, we were able to produce four brand new Carbon Clincher wheels to add to our product line.  The new Carbon Clincher FLO 45, FLO 60, FLO 90, and FLO DISC wheels are not only more aerodynamic than our old wheels, they are nearly a pound lighter per set, and more compliant on the road.  

    Everyone who has ridden the new product line has been very excited about its performance and we are confident that you will be too. Here’s to the 2016 season, and to our new product line.

    Having Fun Trying to Stay on Two Feet with 80-Mile-per-Hour Wind at the A2 Wind Tunnel
    Please let us know if you have any questions about the article. We’d be happy to answer them for you.


    Take care,

    Jon and Chris

    Join the Conversation

    24 Comments

    1. Thanks for this. I really appreciate you giving the end user insight into what is customarily a black box process.
      One question. Based on the tables in this post, it would appear that the time savings favor the use of the Schwalbe Ultremo ZX in 23mm instead of the Conti4000s 700×23, but elsewhere you recommend the Conti4000s. Is your tire recommendation based on Crr instead of drag effect?
      Thanks,
      Hy

    2. It is not by accident that you've developed a superior wheel line. You push your own limits! You've invested a lot of time and sweat equity, and proven it all through extreme and grueling testing. Congrats!

    3. Hy,

      Great question. We looked a number of things. The Schwalbe, while fast does not have the best rolling resistance. We take multiple things into consideration. At this time, the Continental is a tire that has it all.

      Take care,

      Jon

    4. Echoing the thanks for providing this insight. Very interesting to see the process, and I think it gives the potential buyer even more confidence that you guys are producing wheels that rival the biggest names in the industry.

      Also makes me wonder how often those bigger names put in this much work when producing new lines. Seems like one or two companies made somewhat radical changes 5-6 years ago, and then have made just subtle changes since. Possibly just doing some wind tunnel tests but nothing huge. And then others seem to just copy those designs and then rely on their reputations to sell at inflated prices.

      These blogs really show that any similarities your wheels have with other brands is due to your research coming to similar conclusions, and not just seeing a popular wheel and trying to copy it. And even after putting in so much work just a few years ago on your first lineup, you asked more questions, did even more research, and built up a whole new line almost from scratch! This is why you guys will continue to succeed and grow. You guys rock, and I really hope I can get my hands on some Flo wheels soon 🙂

    5. Sebastian Phong,

      Thanks for writing. All of our new wheels are tubeless compatible. That means the rim bed is designed to work with a tubeless set up, if the user chooses to do so. All you will need is a conversion kit similar to a Stan's Kit.

      I hope that helps,

      Chris

    6. Ryan W,

      I really appreciate the kind words. We always loved our first model of FLO wheels but we knew there was a lot more we could do from a design stand point. With more time and resources available, we were really able to do what we wanted this time around!

      Take care,

      Chris

    7. Hey guys, I wonder if you could comment on the NDRV differences between the Aluminum + Carbon and the pure Carbon. Specifically I was wondering why the Carbon Clinchers are faster across the board with the exception of the disc. Thanks!

      P.S. As an engineer I really appreciate the data and more importantly the methodology behind the data! Keep up the good work!

    8. Lucenzo,

      Thanks! Let me try and answer your question. The biggest reason the Carbon Clinchers are faster is because you have pretty much full control of the entire rim shape including the brake track. The Aluminum + Carbon wheels include the carbon and aluminum "seam" which limits your ability to shape in this area. Our Carbon Clincher FLO DISC is still a spoked wheel like our Aluminum + Carbon DISC, the only exception being the rim is a full carbon rim. With the Carbon Clincher DISC the carbon fairing is still bonded to the carbon rim, and the "seam" is still there. We expected the Caron Clincher DISC and the Aluminum + Carbon DISC to have nearly identical drag numbers, and they do.

      Take care,

      Chris

    9. This is awesome and terrible news for my wallet after having bought the old models a few years ago 😉

      Keeping in mind I'm not a mechanical engineer, some of the questions that I have are (sorry for length):
      1. Would it have been better to mount the data collection rig lower, in front of the wheel, to get more appropriate wind data for the area you are trying to optimize? Is that just not practical?
      2. Would having the wheels mounted to a bike/rider in the wind tunnel give you the same results in terms of NDRV? I just wonder how much we can trust these numbers with just an isolated wheel.
      3. Just want to confirm that what you are saying is that drafting did not significantly affect measurement values vs. non-drafting scenarios. That seems very counter-intuitive.
      4. I'm not clear how relative velocity factors into the final results. Was it only used to discard outliers?
      5. How is your yaw angle chart in step 2 affected by relative velocity? Do the percentages in each yaw zone change much at a low relative velocity vs. a large relative velocity?

    10. alex_ygd,

      Thanks for writing. Let me address your questions.

      1. No, I don’t believe so. Mounting low in front of the wheel could interfere with the readings and does not allow you to get close to a rider in a draft.  We are not just optimizing for the front of a wheel, we are optimizing for the whole wheel.  Mounting on top clears up more issues than it creates.  

      2. The results will vary from bike to bike and rider to rider. Testing our wheels on a P2 with a certain rider only gives relevant data to that bike and that rider. Which means our data is useless to a different rider or the same rider using a different frame. You can certainly trust the numbers. We aren't hiding anything in the numbers. We are currently working with Ryan Cooper from Best Bike Split to try and get some good time saving estimates of a full bike and wheel system over several popular Ironman courses. We'll have that data up as soon as we get it.

      3. That is true.

      4. Relative velocity was used to identify outliers. If your relative velocity is 0mph you have no drag so there is no point in considering those yaw measurements.

      5. Relative velocity does effect your yaw angle but when things average out over an entire ride, you still get the same or very similar result.

      I hope that helps,

      Chris

    11. Good morning,

      I find your research very interesting and I am curious about some of the factors/constraints you used that may have potentially limited the benefit realized by the redesign. Specifically the factors such as the lacing pattern, the shape of the spokes, nipples, the size of the hubs, etc.

      Did you consider sourcing different components or for that matter designing different components to help contribute to achieve the performance benefit.

      I am looking forward to hear more about how the wheels perform in the races.

      Regards,

      David

    12. David Prairie,

      Thanks for writing. Let me address your questions below.

      We studied lacing patterns, spoke count, and spoke shape while in the tunnel this year. We'll be publishing that data soon. We use Sapim CX-Ray spokes because they give an aero advantage over round spokes… something we were able to confirm in the tunnel in November. We did not study hubs.

      That said, remember that the wind tunnel tests were done with the full wheels, tire, rim, spokes and hubs. Any advantage over the Open Pro was recorded. This means all of the benefits realized by the redesign are what you can expect… meaning you won't have to wonder how much drag a hub will add because it was already in the wheel when studied in the tunnel.

      I hope that helps,

      Chris

    13. Hi Chris,

      I was checking the new carbon wheels (you did a great job!), namely comparing the 60 to the 90, and it came to my attention that up to 5 degrees of yaw the 60 has lower drag numbers than the 90.

      60C VS 90C
      107,9 VS 127,6 @0
      112,6 VS 122,3 @2,5
      93,6 VS 101,9 @5

      Taking into consideration your new found yaw distribution, this means that for a fast rider in a calm wind day the 60 should be faster, right?

      Greetings,
      Raul

    14. Raul,

      Thanks for the kind words. Our studies showed that overall the 90 is still faster. That said, with the right conditions (like you mentioned), the FLO 60 could be faster.

      Take care,

      Chris

    15. I've been using 60/DISC on my Specialized SHIV 2x week for last 2 years with zero issues. Best set of wheels PERIOD.
      Better than others twice the cost and it got better? outstanding job guys. Chapeau !

    16. Loved the details! (Warning, I am a phycisist). It looks like you are NOT rotating the wheel consistent with the airspeed. You will find significantly more drag, and much more sensitivity to spoke details, if you rotate wheel in wind tunnel. This is because the top of the top spokes will be traveling twice as fast though the air, meaning about 4X the drag, which will outweigh the 1X reduction in drag of spokes when near the bottom of rotation. Maybe next time you will have even more fun spinning your cool wheels!

    17. Phillip in Katonah,

      Thanks for writing. We DID, in fact, rotate the wheel consistent with the airspeed. We understand the principles you are talking about and have taken those into consideration.

      Take care,

      Chris

    18. Is it not misleading to use a 32 spoke wheel as the baseline? Especially on the chart on your aero page. I understand you are selling full wheels and not rims, but it seems more truthful to use a shallow/narrow 20/24h wheelset like the Shimano RS010 which matches your spoke counts at 20/24 and is a true budget/training wheelset.

      I did a quick cda calculation based on some numbers I found online on spoke count, and it seems the baseline would be much closer to your other wheels at zero yaw if it had fewer spokes.

    19. Andrew,

      I think it would be misleading if we didn't call out exactly what we tested, but we did call out exactly what we tested. The industry has used that wheel as a baseline many times, which is why we chose it. Yes, we could have chosen another wheel as a baseline, but we simply did not. I'm not sure what numbers you found online or what calculations you did, but I feel the best way to get a good answer is to test everything in the tunnel. Perhaps the next time we are in the tunnel we can include a lower spoke count training wheel.

      Take care,

      Chris

    Leave a comment

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