When the new Felt AR was announced, serious questions about the performance of the bike relative to the increasingly crowded aero road bike category were raised. Felt promised to respond to these questions with a technical whitepaper. Well, Interbike is over and we have the data. We’d recommend you sit down and get comfortable for this one. Let’s get started.
Test Methodology
Felt really did their due diligence in the tunnel for the new AR. They went to the San Diego Low-Speed Wind Tunnel – widely regarded as the gold standard for bike testing. The yaw sweep is not just the non-drive side of the bike, either. The AR and the six other bikes (technically seven, but we have to throw one out for reasons we’ll get into later) were run from -20 to +20 degrees yaw at 2.5 degree increments. To put that into perspective, each bike was measured seventeen times during their sweep for a total of 119 datapoints from Felt (136 if we include the second run of the AR with Felt’s test brake). Each bike’s components were identical and transferred between them for each bike’s run, as well as being setup for the same rider (i.e. stack, reach, distance to pedal, etc.). Also, the frames were all 56cm “or equivalent,” according to Felt. This is as close to an apples-to-apples test that we are likely to ever get.
Component list according to white paper:
| Wheelset | Zipp 404 Firecrest Clincher |
| Tires | Continental GP4000 (width not listed, photos show 23mm) |
| Component Group | Shimano Dura-Ace Di2 9000 |
| Brakes | Shimano Dura-Ace 9000 F+R (Propel tested using proprietary, integrated brakes. AR tested using direct mount rear brake. New AR w/ Felt Brake using proprietary design) |
| Saddle | Selle Italia SLR |
| Stem | 3T Integra (120mm, -10 deg rise shown in photos, Propel tested with standard stem due to 1 ¼ steer tube) |
| Handlebars | Zipp Vuka Sprint |
It is worth noting that Felt makes the following claim: “The choice of components was based on a typical, high-end aftermarket configuration and was entirely unbiased and not meant to influence the performance of any one bicycle frame. The specifics of the setups, such as the cable routing were optimized for improved aerodynamics and actual function.” The included photos from the whitepaper (gallery below) appear to bear out their claim of optimization and adherence to consistency of setup, which is admirable in this day and age of “tunnel data marketing.”
Test Data
As noted above, in the interest of fairness of analysis, we will not be discussing the entry for “New AR w/ Felt F Brake” as it presents some fairly astounding drag numbers for an addition of equipment that we have no information on. Leaving that aside, let’s take a serious look at the drag graph above as it relates to the competition in general.
For those doubting Cervélo’s claim about the S5’s dominance of the aero road bike market, you have been both vindicated and disproved in the same tunnel test. Prior to the AR’s debut a month ago, the S5 was the clear frontrunner in the aero road bike category by a comfortable margin and is still arguably the best choice from -5 to 5 deg yaw. Impressive for a three-model-year-old bike, but the full sweep tells a different story, one that begins and ends with Felt’s new AR.
Out past +/- 5 degrees, Felt’s new wonder bike is literally unrivaled in the category, posting a drag curve that plunges to a 90g drag delta on the non-drive side—compared with the S5 at 12.5 deg—and maintains that lead through the rest of the sweep. On the drive side, the test AR maintains a lead of around 75g from -10 to -15 against the Cervélo, though the gap closes a touch out past -15 yaw. Against the S5, the newest bike from Felt Racing Division handily takes the win in the war on drag. Against the rest of the competition in the category, the data speaks for itself—the war was over before it even began—2014 sees a new king of the aero road bike in the Felt AR.
Time Savings
Felt’s new whitepaper also asserts claims as to “how much faster is bike X versus a Felt F1” and extrapolates them out to time saved versus the same bike, though we are somewhat short on details other than “…we used a weighted average with an emphasis placed on more common angles of attack such as 10, 12.5 and 15 degrees. The weights were derived from our experience of aerodynamic testing in the wind tunnel and real world.” We will leave them here, without comment, for the reader to make of what they will. For our part, we believe that the drag data speaks for itself: Felt has come out on top and been as transparent about it as is possible, which makes their victory all the more impressive.
Final Thoughts
We think the data is clear regarding the 2014 Felt AR and its trip to the tunnel with all of its friends—there’s a new sheriff in town and his name is Jim Felt. The competition has quite a bit of catching up to do. And according to this test, that is going to be no small task. Felt did everything right in testing their bike against the competition and came out ahead. From every angle, this is a test that shows what everyone has long suspected, and what the next generation has to compete with. Long live the Felt AR.
[Update: Felt Test Procedure – Felt 2014 AR v Comp Aero Test procedure Rev B-1 – AG]
[Update 2: The complete Felt AR 2014 White Paper – 2014_AR_White_Papers_8_26_13]
Can you post a link to the actual whitepaper?
Sorry it took so long, but we attached the complete test procedure document from Felt today.
Thanks for posting this, I appreciate it. Do you guys have access to the entire whitepaper or just this appendix?
Now you are just getting greedy. We just added it to the article as well. Enjoy the reading!
As is often the case the appendix is more interesting than the actual whitepaper. Kudos to felt for testing with a consistent handlebar height despite the fact that it appears that they had to use more spacers than their closest competition to get the same height, which you would think would be a detriment.
No waterbottles, might make a big difference with some frames.
Pingback: 11-17-2013 WiR | AeroGeeks·
They always leave out the water bottles. I’ve never ridden my road bike without them. I know the propel was designed around use of water bottles.
This whole test was useless. No watter bottles, no dummy. Felt themselves asked Cervelo to be able to use their Foam Dave. I think that says a lot
Felt asked Cervelo to use their Foam Dave? Where did you hear/read that?!
Time save (minutes) on witch basis? 1 hour ride at 300W?
Watts save (watts) at witch speed?
Honnestly you can make numbers speak as you wish
Why is there never any discussion about mechanical efficiency and fit in these tests? You can have the most aero bike in the world and if it doesn’t fit you and it’s a complete noodle, all that aero advantage adds up to nothing.
Hi Devo,
Regarding mechanical efficiency take a look at http://www.friction-facts.com/. They do very specific testing to find which components truly help you bring the most power to the road.
As to fit – one of the things we emphasis here at AeroGeeks is that before you buy any component – fit has to be one of your prime considerations. For example knowing that the S5 and AR are the two most aero bikes in this shootout – finding that the S5 fits you best should be the bigger consideration over the enhanced aerodynamics of the AR. But all that being said – this was meant to be an aero only shootout that provides you to make a more informed decision once fit factors have already been weighed.
Pingback: Cervelo S5 – First Ride | AeroGeeks·
I like the transparency with Felt’s testing–very refreshing. Perhaps the only slightly misleading element is the 30mph test speed. Few people who can ride 30mph in the wind for long periods get to choose their frame.
The relative differences in drag are much smaller at 25mph or 20mph, where mortals ride. On this page – http://www.novemberbicycles.com/blog/2013/1/21/the-rail-prototype-wind-tunnel-data-and-calculations.html – November has a comparison at the 3 different speeds for their wheels. First time I had seen that (and they are another good example of transparency–their wheel didn’t even win their test).
The only issue I have with the testing is it appears they used a Giant Propel, not the Propel Advanced, which has markedly lower drag, so it’s unclear from these results whether the new Felt is actually the KOM when it comes to aero…
I believe the only difference between the Propel and the Propel Advanced, aerodynamically wise, is the aero handlebar. I also believe that Felt held this aspect constant in theirs tests by using the Zipp handlebar across all the frames. So the Advanced, in this case, would not have made a difference.
I’m trying to find the chart, but I believe I recently saw a drag chart that had both the standard and the Advanced, and the Advanced was significantly lower. I could be wrong, but that’s my recollection.
You are probably correct….but the reason for the Advanced being better than the normal Propel is that Giant’s testing for the Advanced places the Aero handlebar on the Advanced and not on the regular Propel. The aero handle bars make a difference in the overall numbers. Giant’s testing compares the packages you can pick up at your dealer, where as Felt’s testing tries to compare just the frames. Neither way is wrong, just two different end goals for the testing methodology.
Fascinating, thanks!
Hi David, sorry for the late reply. Based on the images that that Felt provided of the testing. It did appear to us that they are using a Propel Advanced SL. Notice the seatmast versus a standard seat post (http://aerogeeks.com/2013/09/24/2014-felt-ar-wind-tunnel-data-analysis/propel-2/#main).
There are multiple difference in the propel and the advance sl that would affect Aero. Our team rides a mix of the two. The Advance SL is much more Seamless.
I have been proven wrong…can you state what the differences are so I can know for future conversations?
Other than differences already stated above the advance seat tube and clamping system is significantly chunkier to a point it is almost an eye sore. The seat mast isn’t even the same shape as the SL which is very apparent with the bikes sitting next to each other.
Pingback: Culprit Croz Blade One Bike Arsenal – First Look | AeroGeeks·
Pingback: Felt IA Wind Tunnel – Data Analysis | AeroGeeks·