Interesting that they have have only the two yaw angles. When I read about the new QR design to divert the air around the drive side to reduce drag, my immediate reaction was that it was an intriguing new approach, and I could see how you could get benefits with relative wind from straight ahead and some yaw angles.

What I cannot see is how you could get equally good, symmetrical results through all yaw angles. Given that many of the benefits are supposed to derive from asymmetrical movement of air, I would expect performance to be not as good from certain yaw angles (I'm not saying the drag numbers would be BAD, just not as good). After all, a design that gets at least some of its benefits from pushing airflow away from the drive side would lose that part of its design advantage when the relative wind it too far from 0 degrees yaw to prevent that.

(Disclaimer - I am not an aeronautical engineer, but I have taken aerodynamics courses. I was a Navy pilot for 10 years, and my life often depended on understanding these kinds of issues)

While I applaud the presentation (much more than typically seen), there are still a few things left out:

• Wheels spinning or still?
• Wheel spec? Is that a Hed lenticular disc?
• Wind speed drag force measurements are relative to?
• What year P3C and P2C? Gerard has mentioned in the past that there are differences in the fork performance across years.
• Tunnel repeatability/uncertainty? It's hard to put the numbers in context without "error bars".
  

Actually...my hunch is that the majority of the differences shown is due to the just the differences in the "heads" of the exposed seatposts on each bike. That's one of the issues with that particular test setup. Heck, with the data shown, I'd have a hard time choosing a Cd0.1 over a Seduza (they're basically tied)...but, my suspicion is that the Seduza does as well as it does just because of it's tiny seatpost head (especially as compared to the other bikes, and the Cervelos in particular).

Not only is the lack of a saddle not a "fair" comparison, but IMO the lack of bars also unduly "hamstrings" the bikes with lower headtube heights. A simple cylinder hanging up there in the breeze is much worse than a cylinder with a stem attached.

---

http://bikeblather.blogspot.com/

I've been intrigued by the off center design and presumed benefits. As much as I love the toys, I'd have a hard time justifying (even to myself, let alone my wife) saving 3 watts with a $3300 frameset vs a $1999 complete bike.

Gerard has been quoted many times saying that he is perfectly happy with always coming up second in everyother windtunnel test preformed by a competitor.

That said, not having aerobars or a saddle on the bike seems wierd. if they are going to do that they might as well take everything else off the bike too.

---

Grant
----------------------------------------------------
Proudly sponsored by Desoto Sports
Please Support CAF every little bit helps http://raceforareason.kintera.org/grantreuter

By the photos shown on QR website, the P3 looks like the 2006 version, most likely with the Wolf TT fork, which Gerard has stated mnay times is slower than the new 3T fork and even the Wolf CL fork.

---

People sleep peaceably in their beds at night only because rough men stand ready to do violence on their behalf. George Orwell

moreover the only people on this forum who TT between 0 and 10 degrees yaw are bjorn, napoleon and saulraisin!

As the proud owner of a QR Seduza, can I just take a second to say how pleased I am at the results of this test? :)

---

Coach Rob Wilby
http://team.oxygenaddict.com/

OxygenAddict Podcast: http://www.oxygenaddict.com/podcast

I wouldn't say the old carbon bikes were poor performers. The Cervelo data from BRAIN bike (take it for whatever bias you perceive there):

P3C = 690g of drag for a size 56 bike
Trek TTX = 690
P2C = 705g
Pinarelo = 705
Felt DA = 725
QR Lucero = 735
Kestrel Airfoil = 760
Kuota Kaliber = 780
Guru Crono = 790
Scott Plasma = 800
Orbea Oro = 810

I rode a Lucero and really love the frame. The geometry was not great with 76.5STA; I could just barely get my saddle where it needed to be. But other than that, I really found it to be quite a fast bike. The shapes of the tubes - and especially the seatstays - had a fair bit of development in the tunnel according to Brad at QR. John Cobb, IIRC, was the man behind the seatstay design and actually contribued quite a bit to the frame overall. Even going by the numbers above, the Lucero/Seduza/Caliente is certainly a fast bike (aerodynamically). The minor tube diameters are very small - 28 or 29mm, though I think I have them written down somewhere from when I mic'ed them - which I suspect is a big part of it.

---

"Non est ad astra mollis e terris via." - Seneca | rappstar.com | FB - Rappstar Racing | IG - @jordanrapp

 
BTW, I think I'm going to need a bit more explanation about how the steerer tubes were "equalized above the headset top cap". When I took the pics in that .pdf and lined up the BB heights of the Cd0.1 and the P3C, there was a fairly large difference in the height of the top of each steerer tube.



Doh!

---

http://bikeblather.blogspot.com/

when I start riding a bike without handlebars, cables and a saddle, the QR goes to the top of the list!

The older QR tested well in Cervelo's testing, so it's no surprise it did well. The new bike certainly looks the part so I'd expect it would do well also. It's worth noting though that the Trek white paper showed a 30 gram difference between 0 and 10 degrees for a P3C--QR shows an 80 gram difference. Who's right? (I suspect both-when you add cables/handlebars, a rider etc).

While I think the 'directing airflow' concept is an interesting idea, one thing I wonder about is 'how does it perform at both directions of yaw'. For that matter, I wonder that about any bike.

I'm pretty much left to conclude that they're all fast bikes...but that's about the only conclusion I can reach out of that data.

difference BB heights? just a quick WAG for what 'equalize' means...

One detail to note is that in my interpretation the wind axis cd is the primary number that is displayed. I could be wrong but I have to push the bike along the bike axis and not the wind axis. The body axis numbers show a much closer race and the P3C comes out on top of the seduza w/ the p2c close behind. All with an old inferior fork. The steerer tube is suspect as well in my opinion but I will take their word that the steerer tube is equal length above the headset cap. Is the headset cap height equal heights? I'm also suspect of a testing protocol that involves leaving a round tube sticking up in the air, but I haven't done comparison testing of these protocols so I all I have is unfounded opinions.

1. What air speed are these numbers for.
2. Body axis drag is our main concern with regards to aerodynamics correct?

I don't know why, but for some reason I feel like I'm being a fed a partial story with regard to the bikes.

Here is my take:
1.Leaving off the handlebars would be fine if it didn't also mean that cables are then excluded. Am I to believe that no matter where cables are routed they have no effect?
2. Were wheels spinning? I believe it has been stated that the rear fork brake placement is a bigger problem on spinning wheels.
3. Why not a sweep from left to right from 20 degrees yaw all the way across?

I guess it's cool that they posted it but it also leads me to believe that Felt is on to something when they say, "our bike is fast, trust us". Cause people(myself included) will rarely ever be happy.

Sine the P3 is obviously absolute crap,

I will volunteer to take it off their hands.

where can I pick it up?

---

---

Kat Hunter reports on the San Dimas Stage Race from inside the GC winning team
Aeroweenie.com -Compendium of Aero Data and Knowledge
Freelance sports & outdoors writer Kathryn Hunter

Tom A,
Good call on the pic error. The pic of the P3 was from a pull where the P2's Wolf TT was sub'd in. I tested many combinations of frames and forks and did attempt to calculate out any subtle variations. The actual data of the P3 is raw (not factored or calculated with change) and using a 3T that was trimmed to spec. Nice call and debate me as busted (laughs and respect).

Jordan,
Great plug in of the BRAIN data which confirms my respect and reasoning for making a very simple comparison with the Cervelos. They are the class of the field and have provided the benchmark to beat with regards to production bikes. Fortunately, I've been able to benchmark against many prototypes and concept bikes that yielded better (and worse) results. Needless to say, the P3 is a great performing aero rig.
As for your success on the 1st GEN Lucero: your input has been invaluable for the changes that have gone into the Seduza, Caliente, Lucero mold. A steeper seat angle and aero tuning are what the 2009 model feature. The 1st gen results (as you rode) are right in line with the BRAIN classification. Now, instead of patting John on the back for his contributions to the old Lucero, how about giving yourself some acknowledgement on what led to the SHIFT (Cd0.1) concept. It was that cold morning in Flagg as we sat at Late for the Train when you posed the question of "why not fair the chainring?". I went home and built a prototype and furthered the design until August of last year. That's when I gave up on getting the performance gain across a 40deg (+-20) yaw sweep. I spent countless hours in the tunnel adding and subtracting material from that proto. I should send it to you for 'wall art' (laughs). Until I removed the beloved fairing and started tuning what was happening behind the front wheel, the idea didn't work well enough in all conditions. The concept was simply too extreme. I want to publicly thank you for your contributions to QR as an athelete, friend, and evil genius.

footwerx,
A2 did a really cool study by installing a wind probe on a PRO and JOE AVERAGE and putting them on the same courses (same days) in a huge variety of conditions. Initially I felt 14-16 degrees was the sweetspot that the bike should be tuned to. Eventhough the bike does work well there, the sub 10 degree category is what the latest research shows as important for all of us. As you will read in the data, I'm very interested in what benefit we can realize from our much lower side loads experienced in QR's latest work. I think this could be more important than drag when considering stability and efficiency of effort. Thoughts anyone? This will be a fun field of study moving forward.

Tom A,
I love your insights. You're bright (respect). Wheels are spinning, Stinger front, Proto super-fat disc that Steve (Hed) gave me for this testing protocol. This thing is wicked fast and I had to develop and evaluate with a standard. This was what he and I felt was best to baseline with as well as use for compatibility. It didn't fit in one of the Highroad prototypes (super narrow with 60mm BB shell) that Lars allowed me to test with. Wheel and wind speeds were varied and the design was tuned accordingly. By all means, the last thing we wanted to create was a "Pro Only" bike. Testing was done with a variety of forks and each bike was reported with the best fork available for/with each. Thanks again for busting me on the bad photo selection ... these were pulled from hundreds. Please trust me that the data is presented in a very fair and virtuous manner. No cheating has been allowed.

heloguy,
You are correct, drag is always higher on the drive side. That was the primary inspiration for this design: reduce air flow across the drive train. Please know that the data presented is an easily digestable 'snapshot' presented for the sake of explaining a benefit of the design. We've purposely shown this in a simple, debatable format if for no other reason than the questions that will arrise. The reasoning of only showing the Cervelos in the comparison is purely respect. They are wonderful bikes and have set a standard. Much like the BRAIN data, the Cervelo result in our production bike data is of the best. The information is not presented as a slam on Cervelo but to respectfully show how QR's efforts and newest bikes compete with what we feel is the best competitor.

To comment on the title of the thread: Until now, the Body Axis drag has been the category of focus. The P3 actually trumps our QR Seduza in this category and shouldn't be considered as 3rd. The other stats are left for all of us to debate.

Thanks all,
Brad DeVaney

My participation in sport started on the football (soccer) pitch in England. Anyone that tried to explain how good they were was met, inevitably, with the following :

"Show us your medals"

Which in THIS context (i.e. evaluating the relative drag of bike frames), the "medals" are the numbers read off of the wind tunnel balance.

Nice try... ;-)

---

http://bikeblather.blogspot.com/

With good reason: it is drag along the body axis that determines how much power must be generated to travel at a particular speed.


Well there you go - good on you for admitting it.

 
OK...can you explain the process of "steerer tube equalization" then?

Also, seeing as how there are WIDE differences between the tops of the seat posts for these bikes, can you explain the rationale for doing the testing without a seat? It wouldn't be hard to pick a single seat and go with that, wouldn't it? In fact, to account for the fact that this portion of the bike is fairly sheltered by the rider in use, it might have made sense to create "standard block" which would be placed on top of each seat post. The same could be done at the front of the bike to account for the bars and riders arms, no? I guess this is why Cervelo made the "Zabriskie dummy", huh? :-)

Lastly, I'll ask again about tunnel uncertainty/repeatability. With numbers this close between setups, that can be a significant detail.

That said, please don't think I'm just randomly throwing darts at your presentation. I think it's great that you did this and actually published the results. As I said, I'm just trying to place the results in the proper context, and hopefully some of the things in this discussion will help to improve the testing protocols.

---

http://bikeblather.blogspot.com/

Manufacturer testing is so cute! It's like going to the hair salon. Since your paying you can design it just like you want it.

---

customerjon @gmail.com is where information happens.

I don't suppose you've had a look at the effect that the legs have on altering the airflow that you're moving to the non-drive side?

I ask because I was quite shocked by how much the air flowed sideways behind a vertical leg (ie. one leg bent at 90deg and the other straight) when I ran this simulation:
http://www.biketechreview.com/aerodynamics/CFD.htm

Jordan....since you have the inside info, how meaningful are any of these tests without a rider.

Add the body and 2 spinning legs on each side and it completely changes how the air moves around the bike. I really cannot see how any tests without a rider are meaningful at all. You may as well not even test as far as I am concerned. There tests are completely bogus without a rider and have no meaning in the real world as far as I can see.

Dev

The publically-available data don't support your conclusion. To wit: the only frame design for which a significant bike-rider interaction effect has been demonstrated was the Trimble. That is not to say that subtle differences might exist between, e.g., an offset vs. a centrally located downtube when tested with and without a rider, but to label tests conducted without a rider "completely bogus" is, well, bogus.

Sorry, but would the airflow around the back end of the bike not completely change with or without rider, thus changing the impact that the back end design might have with and without rider? Sorry man, show me the tests with the riders, I'm not really interested in tests without riders. I guess I also need riders of various sizes off various bikes so I can pick which rider is the closest shaped to me :-).

Tests without rider are great if the bike is going to ride itself around the race course....oh I forgot, just go buy your P4 and you should be able to pull off a 4:47 split and get out of T2 before Belinda Granger. Just go to T1 exist, give the bike a push and it returns to T2 4:47 later :-)

You know, when you have guys splitting 6 hours for an Ironman obsessing about windtunnel data of one bike over an another, the marketing has won over....better to go out and ride some more kids :-)