I remember reading the data from Cervelo for the new S5 where they did the zone aero dynamics and broke down each segment and part and come to the conclusion that the handlebar made up 30% of the aero drag. Cervelo says their handlebar saves 4.4w at 25 mph. I believe Zipp found a traditional round tube section creates 0.74 N of drag and takes 7.5w of effort to overcome at 30 mph.
So my question is regarding the 30%. I realize it makes sense with it being out front and hitting “clean” air. What I don’t understand is that 30% part. If we say the handlebar takes around 7.5w of effort to over come the resistance at 30 mph, how can that make up 30% of the aero drag if we assume that 7.5w for a round tube handlebar is correct? 7.5w/.3 = 25w for the whole bike?
I’m sure I’m not understanding this completely correctly and leaving something out.
I can’t comment on the detail of your figures, but surely the biggest component of aero drag will be the rider’s body. The frame drag will be small compared to that. With that in mind, 25w for the frame seems reasonable, maybe a bit on the low side for 30mph, but the rider drag is going to be huge at that speed, 100’s of watts.
I can’t comment on the detail of your figures, but surely the biggest component of aero drag will be the rider’s body. The frame drag will be small compared to that. With that in mind, 25w for the frame seems reasonable, maybe a bit on the low side for 30mph, but the rider drag is going to be huge at that speed, 100’s of watts.
I can’t comment on the detail of your figures, but surely the biggest component of aero drag will be the rider’s body. The frame drag will be small compared to that. With that in mind, 25w for the frame seems reasonable, maybe a bit on the low side for 30mph, but the rider drag is going to be huge at that speed, 100’s of watts.
rider drag is tared out from the percentage in that analysis
I remember reading the data from Cervelo for the new S5 where they did the zone aero dynamics and broke down each segment and part and come to the conclusion that the handlebar made up 30% of the aero drag. Cervelo says their handlebar saves 4.4w at 25 mph. I believe Zipp found a traditional round tube section creates 0.74 N of drag and takes 7.5w of effort to overcome at 30 mph.
So my question is regarding the 30%. I realize it makes sense with it being out front and hitting “clean” air. What I don’t understand is that 30% part. If we say the handlebar takes around 7.5w of effort to over come the resistance at 30 mph, how can that make up 30% of the aero drag if we assume that 7.5w for a round tube handlebar is correct? 7.5w/.3 = 25w for the whole bike?
I’m sure I’m not understanding this completely correctly and leaving something out.
at 30 mph, the whole bike is more than 25w by itself, unless it is a 2D bike, especially a road bike(especially meaning it probably is a hair slower than a TT bike by itself)
I found my TT bike to be ~92w(aero) by itself at 0(without pedals)
.
the % table is from a CFD study where a rider is present, but whose drag is not included in the percentages.
Thus, the drag on the frame is probably quite a bit less, than a frame with no rider.
Also the frame was an S5, not an S3
I realize the frame was an S5, I said so much in the first post of the thread. I couldn’t find the S5 graph when I looked online. Went back and found the averaged drag graph.
Looks like 75-80w for the the road bike. Lets go on the high side for this estimate; 80w x 16% (what Cervelo claims for the frame) and it comes out to 12.8w. Take 80w x 30% (what Cervelo claims for the handlebars) and it comes out to 24w. Zipp said it was only ~7.5w for the handlebars.
Right, but the wind tunnel tests of a bike with no rider on it
Is different than the question of how much drag is on the frame, when a rider IS on the bike.
The CFD study has a rider present in the airflow, and is then measuring how much drag force is on each component.
So you can imagine there may be much less drag on the bike, with the rider being in the way of a bunch of it.
Looks like 75-80w for the the road bike. Lets go on the high side for this estimate; 80w x 16% (what Cervelo claims for the frame) and it comes out to 12.8w. Take 80w x 30% (what Cervelo claims for the handlebars) and it comes out to 24w. Zipp said it was only ~7.5w for the handlebars.
Ok, thanks for explaining that regarding the CFD with rider tired vs. wind tunnel with no rider on it. Got it.
Then that raises an interesting question, what would the CFD with the rider tared show as a difference between an aero road frame vs. a non-aero road frame. I assume that would show something different than putting a rider on those two frames in a wind tunnel and looking at the difference. Right?
To me the Cervelo calculations sounds a bit smoke and mirror is the context of actual handle bar gains/losses. I would guess reason they add and then subtract the rider is that overemphasizes the real aero benefits of the bars. A good portion of the air which flows over the bars must in real life then interacts with the rider such that the drag of the bars and riders are not actually decoupled. Effectively exposed frontal surface is just transferred from the bars to the rider but by subtracting the rider this effect isn’t apparent from the data making Cervelo’s bars look really good. If you want to make a more fair comparison you need to know the exact rider values which were subtracted i.e. how many of the watts saved by the bar were added to the rider.
Lets assume it is 75w for the bike. I assume that is a complete build bike. How much of that is the frame? How much is wheels? How much is drive train? I don’t have an intuitive feel, but I think the wheels take up a large percentage of the 75w.
Lets assume it is 75w for the bike. I assume that is a complete build bike. How much of that is the frame? How much is wheels? How much is drive train? I don’t have an intuitive feel, but I think the wheels take up a large percentage of the 75w.
According to Cervelo: frame is 16%, front wheel is 16%, drivetrain is 9%, rear wheel 5%…etc. There is a breakdown in this chart.
I would never be quick to dismiss cervelo’s claims. Pretty much every time they give numbers out it has been pretty close to what people independently verify later.
To me the Cervelo calculations sounds a bit smoke and mirror is the context of actual handle bar gains/losses. I would guess reason they add and then subtract the rider is that overemphasizes the real aero benefits of the bars. A good portion of the air which flows over the bars must in real life then interacts with the rider such that the drag of the bars and riders are not actually decoupled.
The interaction of the rider with the bars is still there.
The CFD is reporting how much drag is on the bars, with the rider present.
And the drag that the CFD computes for the frame or rear brake or rear wheel is the drag that happens when the rider is present as well.
How much of this drag can be removed for each component, is of course an open question, due to these interactions and other concerns. But it does give one a first order approximation of where your lowest hanging fruit would be, for reducing the drag of your bike further.
Nothing in it is surprising either. Stuff near the front matters more.
I’m not dismissing their claim, its probably true, I’m just point out that the claim is essentially useless information. Cervelo addressed the question how many watts does changing the bar save save for the BIKE with a rider on it and what percentage of the total BIKE drag does that represent. This is a very different question from how many watts does changing the bar save for the RIDER+BIKE and what percentage of the total BIKE drag does that represent (Normalizing out against the rider on the original bars).
What Cervelo did was the equivalent of measuring the difference between the force of air exerted on two pieces of paper of different sizes held by a person in front of fan by measuring the total force the fan exerted on the person+paper and then person and then subtracting the person+paper. Why measure force in such an overcomplicated way? The bigger piece of paper, or standard bar, will obvious be shown to exerts more drag on the system, and what we really want to how the shape of the paper effects the drag of the paper+person. The answer is because Cervelo wants to sell bars and knows the changing bars has a pretty minimal effect if you include the increase in rider drag caused by the decrease in bar drag. If you measure enough different parameters you can make real data that looks like you want in to look people just need to understand what they are looking at.
knows the changing bars has a pretty minimal effect if you include the increase in rider drag caused by the decrease in bar drag.
Well no, I don’t think available data supports that claim at all.
Also what I think the list of %s shows is that bars are the lowest hanging fruit, for improving drag on an S5, and I imagine that is accurate. Do you think optimizing any of the other components is likely to reduce drag more? Which, and why?
