I didn’t notice anyone posting this yet, so here it is. Interesting blog post about yaw.
Good post, good info. Some known stuff, but it’s good to see it explained clearly.
Some corollaries of what he writes: (1) Low yaw is more important that high yaw. (2) All other things being equal, narrow parts, frames, wheels, and tires do way better at low yaws than equivalent wider items (i.e., wider may be better for comfort and for your rear end, but it is also slower).
.
Interesting post. Some known stuff, but it’s good to see it explained clearly.
Some corrollaries of what he writes: (1) Low yaw is more important that high yaw. (2) All other things being equal, narrow parts, frames, wheels, and tires do way better at low yaws than equivalent wider items (i.e., wider may be better for comfort and for your rear end, but it is also slower).
.
There’s also that other thing about high-yaw numbers: they don’t account for the lateral tire forces that try to pry your tire off the rim. A couple of noisy tries at measuring this have resulted in Crr numbers of almost 10x what they usually are. We really need to get to the bottom of this effect.
There’s also that other thing about high-yaw numbers: they don’t account for the lateral tire forces that try to pry your tire off the rim. A couple of noisy tries at measuring this have resulted in Crr numbers of almost 10x what they usually are. We really need to get to the bottom of this effect.
That is interesting, if the effect were that large you think we would expect bestbikesplit predictions to fall apart on windy days. Then again we may be inputting CdA values that account for this to make the data match.
though, the CdA vs Yaw curve I input still has CdA that gets better with yaw
There’s also that other thing about high-yaw numbers: they don’t account for the lateral tire forces that try to pry your tire off the rim. A couple of noisy tries at measuring this have resulted in Crr numbers of almost 10x what they usually are. We really need to get to the bottom of this effect.
That is interesting, if the effect were that large you think we would expect bestbikesplit predictions to fall apart on windy days. Then again we may be inputting CdA values that account for this to make the data match.
though, the CdA vs Yaw curve I input still has CdA that gets better with yaw
And it should. Just not as much as the CdA vs Yaw curve would have you think. The effective axial drag vs yaw curve is much flatter than wind tunnel yaw sweeps would indicate.
And it should. Just not as much as the CdA vs Yaw curve would have you think. The effective axial drag vs yaw curve is much flatter than wind tunnel yaw sweeps would indicate.
Right, but if there was a ~10x increase in CRR during the crosswind sections, and I was inputting CdA data to account for that and make the prediction match (as BBS has no CRR vs yaw angle input, haha) I think it would probably slope upward.
Of course whoever was doing the noisy tests was probably doing something exactly like this…hmmm
And it should. Just not as much as the CdA vs Yaw curve would have you think. The effective axial drag vs yaw curve is much flatter than wind tunnel yaw sweeps would indicate.
Right, but if there was a ~10x increase in CRR during the crosswind sections, and I was inputting CdA data to account for that and make the prediction match (as BBS has no CRR vs yaw angle input, haha) I think it would probably slope upward.
Of course whoever was doing the noisy tests was probably doing something exactly like this…hmmm
To my knowledge, there’s no readily available model for lateral forces on bike tires. And if there is, it will take on a very different form for tubulars than for clinchers. There is a drag model for a bike on banked tracks, though. We might be able to re-purpose it, but it would require the lateral force curve: CyA vs yaw .
What if we road a bike on a treadmill…in a wind tunnel?
What if we road a bike on a treadmill…in a wind tunnel?
A rolling-floor wind tunnel, like this one? http://www.gizmag.com/go/8106/
Yeah, that would do it.
And it should. Just not as much as the CdA vs Yaw curve would have you think. The effective axial drag vs yaw curve is much flatter than wind tunnel yaw sweeps would indicate.
Right, but if there was a ~10x increase in CRR during the crosswind sections, and I was inputting CdA data to account for that and make the prediction match (as BBS has no CRR vs yaw angle input, haha) I think it would probably slope upward.
Of course whoever was doing the noisy tests was probably doing something exactly like this…hmmm
To my knowledge, there’s no readily available model for lateral forces on bike tires. And if there is, it will take on a very different form for tubulars than for clinchers. There is a drag model for a bike on banked tracks, though. We might be able to re-purpose it, but it would require the lateral force curve: CyA vs yaw .
You know, I’ve been looking at some data collected from recumbent tricycles and velomobiles. Their Crr goes through the roof on turns.
Interesting post. Some known stuff, but it’s good to see it explained clearly.
Some corrollaries of what he writes: (1) Low yaw is more important that high yaw. (2) All other things being equal, narrow parts, frames, wheels, and tires do way better at low yaws than equivalent wider items (i.e., wider may be better for comfort and for your rear end, but it is also slower).
.
There’s also that other thing about high-yaw numbers: they don’t account for the lateral tire forces that try to pry your tire off the rim. A couple of noisy tries at measuring this have resulted in Crr numbers of almost 10x what they usually are. We really need to get to the bottom of this effect.
Frankly, that’s why I have to question the method of bootstrapping Crr when field testing on a venue with a curved bank…
What’s “the method of bootstrapping Crr”?
Frankly, that’s why I have to question the method of bootstrapping Crr when field testing on a venue with a curved bank…
Well, we (Alphamantis) cover that case because there’s a model for banked tracks. It’s not perfect, but it is predictive.
What’s “the method of bootstrapping Crr”?
Fixing Crr at a set number when testing at different speeds, which some are prone to do.
Ah. Why is that called “bootstrapping?” (to me, “bootstrapping” means something completely different).
I didn’t notice anyone posting this yet, so here it is. Interesting blog post about yaw.
Don’t know who wrote it, but while their general conclusion (i.e., it takes high winds plus low speeds to create large yaw angles), they seem to be blissfully unaware of the fact that companies like Zipp have measured yaw angles “in the real world.” IOW, the pursuit of low drag at higher yaw angles is driven by direct measurements, and not just a marketing ploy as implied.
The effective axial drag vs yaw curve is much flatter than wind tunnel yaw sweeps would indicate.
For wheels, or for the entire bike+rider system?
(I’m not buying it for the latter.)
I didn’t notice anyone posting this yet, so here it is. Interesting blog post about yaw.
I was just goin to post a thread about burying the high yaw myth/trend.
We need to shift the market back to bikes that are long and low with superior low yaw performance.
The effective axial drag vs yaw curve is much flatter than wind tunnel yaw sweeps would indicate.
For wheels, or for the entire bike+rider system?
(I’m not buying it for the latter.)
I’m including the effects of increased wheel drag (an effect not covered by the Crrmg term alone) along with CxA1/2rho*v^2 .
Would you agree that laterally loading the tires will increase the drag? And that this is an effect not covered by the CxA vs yaw curve?
I’m including the effects of increased wheel drag (an effect not covered by the Crrmg term alone) along with CxA1/2rho*v^2 .
Sorry, until you edited this post I thought you meant the aero component alone.
Would you agree that laterally loading the tires will increase the drag?
Since by drag you mean aero + rolling resistance, yes.
And that this is an effect not covered by the CxA vs yaw curve?
Yes, but then again, neither is, e.g., drivetrain friction.
Of course, the question is, how large is the magnitude of any such effect? Rolling resistance is generally rather small relative to wind resistance, and the amount of lateral loading the occurs is quite small relative to the gravitational force vector. Thus, any tire scrub resulting from a crosswind would have to increase the Crr quite dramatically* to negate the reduction in axial wind resistance.
*I’m not sure data from tricycles or the velodrome really apply here, since 1) unlike a tricycle, the tires of a bicycle both pivot and lean, and 2) on the track, lateral loading while riding at less than the design speed can be quite large, i.e., much larger than that due to a crosswind.