Help: Go with Cervelo S3 or Felt AR1 frameset for new bike build?

I need some help. I want to get a new frame and want an aero road bike as my racing is starting to branch out. I have narrowed it down to the new '14 Cervelo S3 or the new '14 Felt AR1 framesets. Both of these will fit me perfectly in a 54 as the geometry is similar to my Cannondale CAAD9. The S3 just has a little higher stack which means I can run the stem all the way down vs. the Felt maybe needing to run one spacer until my flexibility gets to the point I can run it without any spacers. Cervelo I could always run with a -17 stem in the future if my flexibility improves like I am hoping.

I will be running either frame with Ultegra Di2. It appears the Felt was designed with Di2 in mind slightly more in terms of the internal battery setup with the mount in the seatpost. Not that Di2 won’t run easily on the Cervelo, just isn’t a perfect design like the Felt. Slight edge to the Felt?

Both frames are priced in the same ballpark, $2500 for the Felt and $2300 for the Cervelo. Edge to the Cervelo.

Aero. This is a tricky one as the Felt looks very fast on the data they provided…but they didn’t test with a mannequin nor did they test with water bottles on the bike. It might be the fastest in the windtunnel, but not sure that correlates into the real world as it doesn’t take into account moving legs or water bottles on the bike. The S3 is 4 watts slower than the S5. My gut feeling is to give the Edge to the Cervelo S3 in this, but the Felt AR does appear to be fast. So, Edge to the Cervelo?

Comfort? I don’t have any experience with the new AR or the new S3. Just the old ones. From what I have been told, the new frames are significantly more comfortable than their old selves. I am waiting to get a test ride in as my local Felt and Cervelo dealers don’t have either frame in my size currently in. S3 gets the RCA seat stays that look quite forgiving as well as a change in other tube shapes and carbon layup. The Felt is also quite comfy from what I have read with their carbon layup and that new seatpost. Might have to call this a draw unless someone has a bit of time on both frames.

I also like that the Cervelo has the traditional rear brake. I plan on running a Tri Rig Omega brakes anyways, front on Felt if I go that route. The felt bb brake doesn’t look to complicated to work on, but it is a bit more than the Cervelo. Edge to Cervelo?

Both frames appear to be very close to each other in terms of being aero. Is the Cervelo a clear cut winner between the two of them?

You can’t go wrong with either frame. Both are great. Get the one you think looks better, or get the frame from the shop you prefer to give the business to/get service from.

I couldn’t tell from the website, but is the Felt fork wide rims ready? That is one of the bigger features of the new S3, if you aren’t going to run wide rims, then it probably doesn’t matter. I like the paint of the Felt better, but I’m thinking I will just have the S3 painted black and call it good.

A fellow competitor who beats me on a regular basis runs Enve Smart 6.7s on an AR1, so I’d say it probably works fine with wide rims. One of my teammates runs Reynolds Aero 46s on his new S3, so seems to work fine on that frame too.

Cervelo. It’s just better. It’s red.

Disclaimer. I have a P3.

Felt! I love my AR. It’s the old model but it’s awesome.

I need some help. I want to get a new frame and want an aero road bike as my racing is starting to branch out. I have narrowed it down to the new '14 Cervelo S3 or the new '14 Felt AR1 framesets. Both of these will fit me perfectly in a 54 as the geometry is similar to my Cannondale CAAD9. The S3 just has a little higher stack which means I can run the stem all the way down vs. the Felt maybe needing to run one spacer until my flexibility gets to the point I can run it without any spacers. Cervelo I could always run with a -17 stem in the future if my flexibility improves like I am hoping.

I will be running either frame with Ultegra Di2. It appears the Felt was designed with Di2 in mind slightly more in terms of the internal battery setup with the mount in the seatpost. Not that Di2 won’t run easily on the Cervelo, just isn’t a perfect design like the Felt. Slight edge to the Felt?

Both frames are priced in the same ballpark, $2500 for the Felt and $2300 for the Cervelo. Edge to the Cervelo.

Aero. This is a tricky one as the Felt looks very fast on the data they provided…but they didn’t test with a mannequin nor did they test with water bottles on the bike. It might be the fastest in the windtunnel, but not sure that correlates into the real world as it doesn’t take into account moving legs or water bottles on the bike. The S3 is 4 watts slower than the S5. My gut feeling is to give the Edge to the Cervelo S3 in this, but the Felt AR does appear to be fast. So, Edge to the Cervelo?

Comfort? I don’t have any experience with the new AR or the new S3. Just the old ones. From what I have been told, the new frames are significantly more comfortable than their old selves. I am waiting to get a test ride in as my local Felt and Cervelo dealers don’t have either frame in my size currently in. S3 gets the RCA seat stays that look quite forgiving as well as a change in other tube shapes and carbon layup. The Felt is also quite comfy from what I have read with their carbon layup and that new seatpost. Might have to call this a draw unless someone has a bit of time on both frames.

I also like that the Cervelo has the traditional rear brake. I plan on running a Tri Rig Omega brakes anyways, front on Felt if I go that route. The felt bb brake doesn’t look to complicated to work on, but it is a bit more than the Cervelo. Edge to Cervelo?

Both frames appear to be very close to each other in terms of being aero. Is the Cervelo a clear cut winner between the two of them?

If you’re looking at using Ultegra Di2, why not choose the AR2 complete i/o the AR1? Already have parts?

I haven’t had the chance to ride or test and S3 but the Garmin pros seemed to embrace it more than the S5. In Felt’s testing the S5 offered better low-yaw drag numbers but much worse at higher yaw. I haven’t studied the drag curve on the S3 vs. S5. In a -5 to 0 to 5 sweep the S5 was ahead of the AR, in part by design as the AR’s tubes and optimal performance envelope was pushed out beyond 15 degrees. in a -10 to 0 to 10 or -15 to 0 to 15 sweep, the AR was well ahead. Stiffness and weight, while less important, were two areas that the AR was well ahead also. I have no idea about the S3 stiffness vs. the S5/S5Team lay up.

Comfort on the AR is incredible. Much smoother than my F-series bike. A different world vs. the S5 in terms of vertical compliance. The AR1 with the VR post is in the Felt Z & Spec Roubaix realm. The graph doesn’t scale to include other aero road bikes we’ve tested. Unfortunately the latest S3 is excluded from our stable of test bikes.

Consider where you’ll be buying the bikes from and your intended use. It may be possible than an S3 with a -17 configuration is indeed faster than an AR1 with a 20mm headset spacer stack and a -6 stem. Sure having 3 seatpost options on the AR is nice for a Felt Dealer but once you’ve determined which one you need, it is unlikely that anyone would own all the different versions.

Fit, finish, and relationship with your retailer will probably have as much to do with your enjoyment and the bike’s overall performance for you as any drag data or frame comfort metrics I could provide.

Cervelo makes great bikes, the S3 seems to come with changes in the design where much of the criticism the S5 received. These are the two bikes ahead of the rest of the class by a generation or two. You’ve narrowed your choices wisely.

If I can offer any additional information to help you make your decision, let me know.

-SD

Pure class as always. Thanks for your contribution to the thread as I am looking at a similar choice and my dealer carries both.

If you’re looking at using Ultegra Di2, why not choose the AR2 complete i/o the AR1? Already have parts?

Thanks for your response. Honestly the answer is yes. I have some of the parts and I have been buying others over time as they go on sale and I find them on discount. Every thing is brand new from an authorized dealer. After doing the math and subtracting the 3T wheels that I don’t need or want, I am still coming in $1200-1300 less ($2300 before selling wheels) than what you guys are selling the AR2 complete for. That is worthwhile for me to spec out the build myself.

I haven’t had the chance to ride or test and S3 but the Garmin pros seemed to embrace it more than the S5. In Felt’s testing the S5 offered better low-yaw drag numbers but much worse at higher yaw. I haven’t studied the drag curve on the S3 vs. S5. In a -5 to 0 to 5 sweep the S5 was ahead of the AR, in part by design as the AR’s tubes and optimal performance envelope was pushed out beyond 15 degrees. in a -10 to 0 to 10 or -15 to 0 to 15 sweep, the AR was well ahead. Stiffness and weight, while less important, were two areas that the AR was well ahead also. I have no idea about the S3 stiffness vs. the S5/S5Team lay up.

I have a question about this. From my understanding, Felt did not design the new AR frame with a mannequin on it nor the use of water bottles on the frame whereas Cervelo did. From my limited understanding of windtunnel data, the Cervelo test seems closer to my use of the frame; I sit on it and pedal and I carry water with me. Whereas the Felt design ommited that, right? I know you guys use CFD, just seems like it would be obvious that the water bottles affect the airflow around them and the tube shapes downwind from them. But I’m no expert.

Comfort on the AR is incredible. Much smoother than my F-series bike. A different world vs. the S5 in terms of vertical compliance. The AR1 with the VR post is in the Felt Z & Spec Roubaix realm.

I have heard the new AR is very comfortable, part of the reason I am considering it. Have also heard the same thing about the new S3 which uses the same shapes of the chainstay as the R5/RCA.

Fit, finish, and relationship with your retailer will probably have as much to do with your enjoyment and the bike’s overall performance for you as any drag data or frame comfort metrics I could provide.

The relationship with my bike shop will be minimal as I will just need them to special order the frameset in my size and I pick it up the same day it arrives. Will do my own wrenching when I get it home. But the finish of the frame is important and I believe both Felt and Cervelo are similar in this regard. I would inspect either frame anyways before taking ownership and paying for them.

Both frames seem great. I just wish the Felt had maybe used a mannequin and bottles on the frame during the design process. Since I am buying an aero frame and both frames will fit me fairly well, I just want to buy the fastest frame I can at this point since I am spending the money on it.

I’ve been training and road racing on the AR2 since late January. It’s a fantastic bike, right up there with comfort of SLR01 and handling of SL4, but of course the benefit of an aero frame. I’d consider those two facets above any marginal difference in one being more aero than other. It’s likely to be so minor that personally, I wouldn’t get too hung up on it.

Someone else asked about wheel/tire width. I race on 303 FCs with 25 Conti 4000s, no problem on AR2. If comfort is a significant consideration then going to 25s will make a noticeable difference.

Dave,

Thanks for posting on here. It’s great insight into the workings of Felt and the bike’s development.

Question: Is there any difference in the frames between all of AR models in the line-up? They seem to have the exact same tech description

Also, any chance of an Aluminum version of an AR hitting the market? Matching that with the carbon seat post and I’d guess that you’d have a comfy crit-slaying affordable machine.

Thanks,

Nick

Dave,

Thanks for posting on here. It’s great insight into the workings of Felt and the bike’s development.

Question: Is there any difference in the frames between all of AR models in the line-up? They seem to have the exact same tech description

Also, any chance of an Aluminum version of an AR hitting the market? Matching that with the carbon seat post and I’d guess that you’d have a comfy crit-slaying affordable machine.

Thanks,

Nick

Great question! Us privateers would love an heir to to S1’s throne of race worthy pop-cans. I love my CAAD10 but it’s an aerodynamic nightmare.

SD~
Does the AR1frame come with “plugs” for lack of a better term to fill the ports on the down tube where cables would go if using a mechanical group? I can’t tell because the of the black on the AR2, but it looks like those ports aren’t there. Is the AR1 available sans those ports if you are planning on using Di2 or does the FRD come like that? This is sort of in line with Cervelo’s future cable proofing the cable ports that they did.

thanks,
Ben

Dave is obviously the expert here but the AR2 is Di only, with the cable entry points on top tube just behind stem (none on down tube). The bike is super clean as a result. The AR1 can go either Di or mechanical.

That’s good to know, thanks!

Felt does have a hydroformed aluminium AR mode but according to the official AR thread here it will not be introduced to the North American market

The AR platform will get an aluminum model in 2014 but it will not be offered in North America although I’d consider it a great crit bike. I hear Americans love their crits so we’ll see how long we can keep it out of this country. The frame has an hourglass head tube ~38mm wide at the narrowest point. Fully hydroformed front triangle and seat stays including the patented gap-sheild tech borrowed from the DA (and also on the AR carbon). The cables are internally routed and the frame uses the identical AR seatpost and clamping system (also patented) to allow for one of the most comfortable riding aluminum frame we’ve ever offered.
It is expensive and the market is currently slanted toward inexpensive carbon, especially our competitors Sora equipped carbon $1500 bikes.

-SD

source

Edit: Also between the different tiers of AR models mainly its the components, and as mentioned the AR2 is Di2 only, and for the FRD model the use of light&stiffer composites.

well that about covers everything haha

I thought I had read somewhere about the alu AR frame but looks like you found it!

Thanks

Sorry the information we’ve presented appears incomplete. Of course the impact of bottles and a rider were studied as they’ve been with all of our tunnel-developed products. If you plan to run one round bottle, stick it on the seat tube. If you are running an aero bottle, stick it on the seat tube. The absolute lowest drag number may be found using aero shaped bottles on the downtube as the delayed separation or less abrupt separation (less lift, drag increase/stall) improves with the right additional shape. I’ve mentioned the effect of “helping the air re-attach”. That’s not exactly accurate but I used it to demonstrate what is occuring. Separation is a process, not an on/off event. Our higher yaw shapes may begin separation at 15 degrees. Adding the aero shape of an Elite bottle for example reconfigures the net shape and helps delay that separation. It is NOT possible however to reverse stalled/highly turbulent air downstream of the separation. The churned airflow will not just stop being turbulent unless it is allowed to stabilize much further downstream or you put energy into the system to do so.

The drag penalty of round bottles could be as high as 55-80g depending on frame size and yaw but what you add that 55-80g to is a frame that already has far lower drag than anything else beyond about 7.5 degrees of yaw. We could have made the frame-alone have higher drag and the frame + bottles a similar NET drag by making an airfoil that required or favored a bottle but that would not have made for lower drag necessarily.

A frame with ~600g of ave -20/0/20 drag that was designed around round bottles and has a net gain of 0g of drag when bottles are installed is not faster than a frame with ~500g of ave -20/0/20 drag that was designed with airfoil shapes which add an additional 80g of drag when a round bottle is behind/ahead of them.

580 < 600.

I’m oversimplifying the approach our engineers took here but the idea of creating the lowest system drag (frame, components, rider, bottles, cables, etc…) was indeed the priority with both the AR and IA projects.

Publishing the comparative data was a big step for us as it required exhaustive legal review and countless verbiage revisions. Prior to 2014 we did not elaborate on who or how our testing against competitors was done nor did we open the door to our development so widely.

-SD

Thanks Dave for explaining that about the bottles.

You just briefly hinted that you did study the impact of a rider on the bike but didn’t go into any detail like you did on the bottles. From what I know, I didn’t think Felt had a DZ like mannequin like Cervelo does. How did you guys use a “rider” during your testing and also when you compared your frames vs. other frames for the white paper?

Also, to add to Dave’s post above, I believe in the AR thread, he noted that two Elite Crono bottles have a 0g impact on the AR’s drag, which is what I have chosen to ride with. There are tradeoffs to using those bottles, but for me it is worth it to not spoil the engineering that went into the bike itself.