Gerard,
For the benefit of all parties contained within this BB, please confirm for me that an aerodynamic frame is at its best at 0 yaw. As the yaw angle increases, the drag on the frame increases until it reaches the stall point.
Please also confirm that this doesn’t mean an aero frame is slower. It is faster at low yaw angles before stall and no slower at high yaw angles when the tubes are stalled.
Thank you.
It’s simple aerodynamics…
I know this, but my word and engineering isn’t worth anything to some (read Pubmed) people. I am hoping that if someone with the bicycle credentials as well as the engineering credentials confirms what I know and you know, then we can put this to rest.
He will not. Let’s just wait for him to answer, though, rather than assuming. IMO, you don’t know enough engineering to make that statement, or you think I don’t know enough, so let’s just wait for someone whose authority we can agree on to answer.
Are you saying that a 170 lb rider on a P3, producing, say, 250 watts on a flat course, is going to go faster in a 10 mph cross wind than in a 10 mph tail wind?
An “out and back” course is not a pure headwind/tailwind situation. Can you not have a crosswind on an out-n-back course?
Quit worrying about saving 1-2 seconds with all this “aero this and that”
Get your ass out riding more and “earn” minutes the real way.
Show me them. Show me where it says, exactly, that an aero TUBED (not one of those super aero bikes inside of a cover) is more aerodynamic in a crosswind than in a straight headwind.
Captain PubMed is right here, we don’t need Gerard… give it a rest.
Do you have a link?
Although, not like it matters, I did what I should have done in the first place, and actually plotted some numbers for a couple of symmetric airfoils:
For NACA 0010:
Cd @ 0 - 2.343742E-4
Cd @ 10 - 6.3236733E-4 (269% of Cd0)
Cd @ 20 - 0.0017066579 (269% of Cd10)
Avg: .000858
For NACA 0015:
Cd @ 0 - 4.9189327E-4
Cd @ 10 - 5.996097E-4 (122% of Cd0)
Cd @ 20 - 8.8862097E-4 (148% of Cd10)
Avg: .00066
For NACA 0020:
Cd @ 0 - 8.5916906E-4
Cd @ 10 - 7.674214E-4 (89% of Cd0)
Cd @ 20 - 5.160523E-4 (67% of Cd10)
Avg: 0.000714
For NACA 0025
Cd @ 0 - 0.0013343904
Cd @ 10 - 0.0010642904 (80% of Cd0)
Cd @ 20 - 3.2913827E-4 (31% of Cd10)
Avg: 0.000909
You will notice to be fair, I picked two high aspect ratio and two low aspect ratio wings. The high aspect ratio wing has a lower drag at low angles of attacks and the drag increases as the angle of attack (yaw angle in our case) increases. The low aspect ratio wings have higher drag at low yaw angles, but do better as the yaw angle increases.
(This may be why Zipp put the bulge in on the 404’s, to increase the aspect ratio (just a theory, but it makes sense) for improved crosswind performance.)
But in the end, the best overall drag performance comes from a low AR wing, which does decrease in aerodynamic efficiency as yaw angle increases.
So theoretically, both you and/or I could be right, depending on what airfoils were used.
If you want to play with airfoils yourself, here you go:
http://www.ecs.syr.edu/centers/simfluid/red/af.html
So, again, I will wait for Gerard to answer regarding the bikes. But I hope you will at least admit now that just because it is an airfoil, it doesn’t mean it does better in crosswinds…
I thought you were backing me here. I have given you real numbers. Now stick to your guns…
Especially if you go by Pubmed’s arguments that the high yaw angles are much less since wind speed is lower close to the ground, then it definitely makes sense to pick the airfoil that is better in the low yaw angle situations. That would mean an airfoil like the NACA 0015, which had the lowest average drag and which does in fact go up in drag as the AoA (yaw angle) increases.
Has Pubmed given you any numbers? Nyet.
Can I ask a stupid question for Captain Pubmed and Rappstar? Which direction is the Cd pointing? Is it pointing opposite of forward motion, or is it pointing in the same direction as the flow?
yaw angles? drag coeffiicients? stall? Are we talking bikes still or did we suddenly reach speeds that allowed for take off? So many techno arguements on this forum lately…holy crap…you know it’s the off season when…the pocket protector gang is done racing and goes back to pointless arguing…smoke a fatty and go watch the Simpsons or something will ya…relax. It would interesting to be a fly on the wall the day all the STers gather into one room and chat it up face to face.
Drag is opposing forward motion so it will be in line with the direction of travel, not parrallel to flow. Lift and drag are vector components of the result force from the pressure differential. This is why you have drag but no lift for a symmetric airfoil and lift but no drag with a straight updraft. Any other case and you can’t have one without the other
But is the data for frame only with rider, or does it include aero wheels? If it includes wheels, the superior drag reduction of the wheels in crosswind could mask the decreased performance of the frame. If it is for just frame+rider (no wheels or non-aero wheels), then your data shows the same thing as happened for the NACA 0020 and 0025 airfoils in my little model. I concede that it is possible to build such a bike.
But that doesn’t mean that a bike couldn’t be built with 0010 or 0015 airfoils that would exhibit the opposite stall characteristics. That bike would be superior in a headwind and worse in crosswind, as the data I posted shows.
I concede that a bike can be built that will perform better in crosswinds than in headwinds. But that same bike could be built to be faster in a straight headwind, and would then suffer in crosswinds. It is a function of aspect ratio, and the effect of aspect ratio is not debatable for NACA airfoils. I showed the data. So I am willing to say truce, a bike could be built with either aerodynamic characteristic. Neither of us is wrong (or we both are wrong and right).
I would struggle for weeks to write something as funny as what you just did Kentiger. Thanks for making my day! 
NACA profiles were developed long ago, there are many low-speed sections out there that are more useful for use in bikes.
Actually, a common convention is for drag to be defined “as the component of the aerodynamic force parallel to the relative wind. The lift is defined as the component of the aerodynamic force perpendicular to the relative wind.”
Depending on how you define the drag direction, makes a difference when you and the Capt. start comparing drag coefficients. You both might want to make sure that you’re using the same convention first of all. The above convention is common in aerodynamics, and the calculations you made I would presume are from an aerodynamics book. I wouldn’t think this convention would be that useful in bike aerodynamics, so they might be using a different one, and wouldn’t make you’re comparisons all that useful.
It seems to me that you are contradicting yourself here: how is that aero wheels can show a greater reduction in drag under crosswind situations, but frames cannot?
I am not contradicting myself. I didn’t say that a frame that performed better in a crosswind could not be built. I just said that isn’t necessarily what you are riding. And your data doesn’t prove that it is what you are riding either. Show me some frame only test results for some recent frame designs. You have to sacrifice something in terms of speed in low yaw angles. Is is not a choice. Take track frames for example. Do you really think they want them to better in a crosswind than in a straight headwind? Would that make any sense at all? Nope. But that doesn’t mean a track frame is going to be the best frame to ride outside for a TT (and not just because of surface irregularities and UCI rules).
An aero frame that performs better in a crosswind can be built. But it will not be as aero in a headwind as a frame that was built for low yaw angle optimization. Look at the NACA numbers I posted. Or, if you want cycling proof, look at the H3 vs. Hed Alps. The rim profile is the same on both wheels, but the difference is in the spokes (or blades vs. spokes):
Wind Angle 0° 5° 10° 15° 30° Alps 0.406 0.392 0.304 0.4 0.415
H3 0.409 0.397 0.321 0.313 0.278
The Alps is best in a straight headwind and lower yaw angles, all the way up to 10deg of yaw and then it gets worse from 10 to 20 deg. It is also more aero than the H3 from 0-10deg. The H3 however, is vastly superior from 10-30deg. One wheel is better in a crosswind, but worse in a headwind, and vice versa.
If you want to talk about contradiction, I would like to point out the obvious contradiction in your arguments against what I have said:
So if 5mph is VERY HIGH, then that would be the upper end of the crosswinds we might expect to see on course, according to you. And in order to have the highest possible YAW angle from a 5mph wind, we need to have it coming from 90deg to our path of travel, which is going to be even rarer just based on natural distribution. It is not like wind is predisposed to blow directly across our path. So based on cyclist speeds of 20-30mph (which would seem reasonable for a bike that is meant to be ridden in the aero position), that would mean that 10-15deg of yaw would be the some of the highest we might encounter. (Which does seem a little odd to me since the guys at Zipp, who do this for a living, say that 10-20 deg of yaw is the most common in real world experiences.)
… SAY WHAT? And you accuse me of being contradictory? You spend one thread claiming that the wind speeds at ground level, and therefore the wind angles, I mentioned are unrealistic and another thread claiming that the bike should be built to preform better in the wind angles you just spent bashing. Makes perfect sense to me!
And regarding NACA profiles:
First, I never said anything about NACA profiles. Second, as Gerard mentioned previously Cervelo has gone away from using NACA profiles - and they are not the only ones.
No, you did not mention NACA airfoils. I mentioned them because they are relevant for showing how aspect ratio affects the performance of an airfoil shape at various wind angles. Yes, Cervelo has gone away from using NACA profiles. Gerard’s exact quote was “You are absolutely correct that NACA profiles have its limitations for bicycle designs. Delaying the moment where the airflow stalls is I would say the biggest part of our work in the tunnel. Of course once it stalls all the bikes basically have the same drag, so the aero frame is overall still faster as it combines “non-stalling and stalling course stretches”. How big these respective stretches are depends on the riders speed, the wind angle and the exact airfoil shape. That’s why our new airfoils are no longer NACA profiles, but profiles designed specifically for the bikes. The shape of the WOLF fork and R2.5 seatstays are examples of this, and other examples are soon to follow. Obviously this is a lot of work, and work on our latest round of airfoils (for a soon to be revealed new model) was started about three years ago. But I think the end result will address many of your concerns.” Note the word NEW, as in 2005+. And if my concerns weren’t valid, why would he need to address them? Old Cervelo’s, like the ones many people are riding, ARE NACA profiles. This makes them relevant. And unless you are accusing Gerard of being not so smart, then there was a reason that he chose them in the first place. If Cervelo really is at the forefront of aerodynamic design, which I believe, and which I think you believe as well, then there was a reason they chose NACA profiles for all this time. Granted they have spent time moving away from the NACA profiles to try to find something better for bicycles, where the air is hitting the downtube after passing over the wheel and fork, etc., but their current bikes have the NACA shapes. Which makes the data I posted relevant. Certainly just as relevant as Chet Kyle’s data from however long ago. I was not arguing against the brand new P3C, which didn’t even exist in our minds, with its new airfoil shapes, although I am not taking on faith that it is going to be best in a crosswind instead of a headwind. I was arguing against the bikes that were out there.
I understand that you clearly have a serious problem admitting to be wrong about anything. But I would hope that you can at least recognize the contradictory nature of your own arguments over the couple of threads that this discussion spans. Finally, I would like to put this issue to rest by saying that if you want a bike that is optimized for headwind (low yaw angle) aerodynamics, it is not going to be as good as the yaw angle increases, and that a bike optimized for high-yaw angles aerodynamcics is not going to be as good in low yaw angle aerodynamics.
You’re gonna have to let this one go, my man.