Here we go again.
Much of what is presented in the video is a knock-down of the validity of current measurement techniques to evaluate aerodynamics of bikes/parts. Egos aside, i think it does raise a valid point... that the real world is not as laminar as we might wish. The optimal design in a lab (or enclosed velodrome) may not be as optimal in the wild. The MIG example is an interesting one. Sacrificing ultimate aero efficiency to better deal with non ideal aero situations. Sounds a lot like the mantra that zipp used to try to make their wheels not just faster, but more stable in cross winds.
BUT... is hambinis protocol more representative of reality? Everything is trying to model reality, the issue is that reality is constantly changing (wind that is), and no two moments are ever the same. So what test best models reality, and which reality are we testing?
So here is what we have for test options:
- eyeball it/armchair logic
- cfd
- wind tunnel (traditional)
- wind tunnel (non laminar airflow)
- outdoor testing (velodrome)
- chung method
- roll-down tests
- i am sure i am missing a few
Which creates “valid” data? Even if hambinis data was more representative of reality, he states a 2.5% error, so i will not use it to compare wheel a to wheel b, because the 2w difference between those wheels are within the margin of error of the test. Compare this to a laminar flow wind tunnel that may be able to theoretically be more precise, but is it as representative of the real world? Pros and cons.
Rider on bs rider off data is the same issue. Rider on is reality, but a rider increases your error. So dummies were created. But i am not shaped like you, and on and on....
The trends this data shows, deeper is more aero, yeah, i buy that. Aero spokes, hide the nippples. Yup. But just as hambini said he could improve a riders aero with a profile picture, i could also improve the aero of a given poor wheel design without an aerospace degree or fancy wind tunnel with louvres (but isn’t that part of how we got here in the first place... people cutting corners in design...). I could probably look at a line of wheels and with a calliper and a few tools to check spoke alignment, could probably align the fastest to slowest wheels with decent accuracy. It is not rocket science.
The trick is not just design, but to make that product to a high quality, at a budget that makes the product/business sustainable. Poor alignment of aero spokes will kill a wheel performance aerodynamically. It is not a design issue, but a manufacturing one. Accountants are always the bane of the engineer. Maybe every wheel maker could make a faster wheel, but at what cost? Most shy from internal nipples due to inconvenience to the buyer. It is a conscious design decision.
In reality, no test protocol is perfect. The trick is to know/accept the limitations. Actually, the chung method is pretty smart imho. Also, i think the aerocoach approach that mixes outdoor velodrome testing with wind tunnel is pretty good to. Why, because both are accessible and reproducible, and involve some element of in vivo real world testing. But in vivo comes with it’s own set of issues (multiple variables you can not control), so there is no one perfect test. The key is knowing the limits of each test.
The ideal test(s)& is one that is accessible (equipment, cost), and easily reproducible. Anything that requires a test protocol that is only accessible to a couple of aerospace companies... is god damn useless to the common bike rider trying to shave a few seconds off their next bike split.
Reduce frontal area
Don’t balloon your tires beyond the rim
Avoid wrinkles in your clothes
Shave your legs
Aero helmet
Deeper rims
Less spokes, and aero spokes.
Hidden spoke nipples
Quality bearings
Optimal tire pressures
Etc
But like most things, aero performance is not a maximize/minimize issue. Wider tires hurt aero but have other benefits. Less spokes reduces rotating drag, but makes for a weaker wheel. Deeper rims suffer more side force. Deeper frames are heavier. Etc. Etc.
It will be interesting to see if future wind tunnels adopt some of what hambini has expressed. But we will be back to... any optimization will be to the test conditions. And is the test condition really your reality? Or their reality? Or her reality? Or the reality of tomorrow (i am talking weather change by the way).
The reality is there will be wheels that do better on calm days, and others that do better in wind, and others that do better in turbulent changing conditions. There will be no best design. Just a best design for a particular situation. Beyond that, aero spokes, as few as you can, as deep a rim as you can get, not too wide a tire, hide the nipples, narrower hubs, etc...
Much of what is presented in the video is a knock-down of the validity of current measurement techniques to evaluate aerodynamics of bikes/parts. Egos aside, i think it does raise a valid point... that the real world is not as laminar as we might wish. The optimal design in a lab (or enclosed velodrome) may not be as optimal in the wild. The MIG example is an interesting one. Sacrificing ultimate aero efficiency to better deal with non ideal aero situations. Sounds a lot like the mantra that zipp used to try to make their wheels not just faster, but more stable in cross winds.
BUT... is hambinis protocol more representative of reality? Everything is trying to model reality, the issue is that reality is constantly changing (wind that is), and no two moments are ever the same. So what test best models reality, and which reality are we testing?
So here is what we have for test options:
- eyeball it/armchair logic
- cfd
- wind tunnel (traditional)
- wind tunnel (non laminar airflow)
- outdoor testing (velodrome)
- chung method
- roll-down tests
- i am sure i am missing a few
Which creates “valid” data? Even if hambinis data was more representative of reality, he states a 2.5% error, so i will not use it to compare wheel a to wheel b, because the 2w difference between those wheels are within the margin of error of the test. Compare this to a laminar flow wind tunnel that may be able to theoretically be more precise, but is it as representative of the real world? Pros and cons.
Rider on bs rider off data is the same issue. Rider on is reality, but a rider increases your error. So dummies were created. But i am not shaped like you, and on and on....
The trends this data shows, deeper is more aero, yeah, i buy that. Aero spokes, hide the nippples. Yup. But just as hambini said he could improve a riders aero with a profile picture, i could also improve the aero of a given poor wheel design without an aerospace degree or fancy wind tunnel with louvres (but isn’t that part of how we got here in the first place... people cutting corners in design...). I could probably look at a line of wheels and with a calliper and a few tools to check spoke alignment, could probably align the fastest to slowest wheels with decent accuracy. It is not rocket science.
The trick is not just design, but to make that product to a high quality, at a budget that makes the product/business sustainable. Poor alignment of aero spokes will kill a wheel performance aerodynamically. It is not a design issue, but a manufacturing one. Accountants are always the bane of the engineer. Maybe every wheel maker could make a faster wheel, but at what cost? Most shy from internal nipples due to inconvenience to the buyer. It is a conscious design decision.
In reality, no test protocol is perfect. The trick is to know/accept the limitations. Actually, the chung method is pretty smart imho. Also, i think the aerocoach approach that mixes outdoor velodrome testing with wind tunnel is pretty good to. Why, because both are accessible and reproducible, and involve some element of in vivo real world testing. But in vivo comes with it’s own set of issues (multiple variables you can not control), so there is no one perfect test. The key is knowing the limits of each test.
The ideal test(s)& is one that is accessible (equipment, cost), and easily reproducible. Anything that requires a test protocol that is only accessible to a couple of aerospace companies... is god damn useless to the common bike rider trying to shave a few seconds off their next bike split.
Reduce frontal area
Don’t balloon your tires beyond the rim
Avoid wrinkles in your clothes
Shave your legs
Aero helmet
Deeper rims
Less spokes, and aero spokes.
Hidden spoke nipples
Quality bearings
Optimal tire pressures
Etc
But like most things, aero performance is not a maximize/minimize issue. Wider tires hurt aero but have other benefits. Less spokes reduces rotating drag, but makes for a weaker wheel. Deeper rims suffer more side force. Deeper frames are heavier. Etc. Etc.
It will be interesting to see if future wind tunnels adopt some of what hambini has expressed. But we will be back to... any optimization will be to the test conditions. And is the test condition really your reality? Or their reality? Or her reality? Or the reality of tomorrow (i am talking weather change by the way).
The reality is there will be wheels that do better on calm days, and others that do better in wind, and others that do better in turbulent changing conditions. There will be no best design. Just a best design for a particular situation. Beyond that, aero spokes, as few as you can, as deep a rim as you can get, not too wide a tire, hide the nipples, narrower hubs, etc...
Last edited by:
Rocket_racing: Aug 21, 19 9:53