Aero data from 19 wheels by Hambini

[tanhalt]

Speaking of GCN, one more to add...how do you square your proclamations of impartiality with the fact that GCN is apparently known (I've heard this from a bike industry journalist) as a "pay to play"-type review site?

Its in no way a secret. You're reading too much into it.

https://www.asa.org.uk/rulings/wahoo-fi ... a17-1.html

I'm not...I'm merely pointing out the hypocrisy of Hambini consorting with (and receiving unpaid exposure from) folks known to be "less than impartial" in their product reviews, all the while accusing others of similar behavior.

So, if I'm "reading too much" into that, then the same is true of the accusations leveled at me.

But, I guess I shouldn't expect that point to be grasped here in the "Hambini Echo Chamber" :-/

[Jugi]

I'm not...I'm merely pointing out the hypocrisy of Hambini consorting with (and receiving unpaid exposure from) folks known to be "less than impartial" in their product reviews, all the while accusing others of similar behavior.

Would you mind mentioning at least one totally impartial ”cycling media”? In my profession the line of impartiality is drawn at ”a free lunch” during a meeting. I’m quite sure none of these cycling media representatives decline a snack during the press events, where new bikes are test ridden and then ”reviewed”.

But, I guess I shouldn't expect that point to be grasped here in the "Hambini Echo Chamber" :-/

You, sir, are starting to sound like there is a personal endeavour behind your posts in this thread. To preserve any credibility, please do tell us if it is so.

[tanhalt]

You, sir, are starting to sound like there is a personal endeavour behind your posts in this thread. To preserve any credibility, please do tell us if it is so.

The ONLY personal endeavor behind my posts is to find out more about this potentially interesting new approach to evaluating wheel performance. All of my questions are to that end, and that end only. If it is a better predictor or real world performance, then I'd like to know that. But, to be fair, the current methods do fairly well, so it'll have to be shown to be quite a significant improvement.

When Hambini states on his blog that "A superior method of analysis is to carry out a transient analysis in a wind tunnel", it is then incumbent on HIM to show in detail why this is so in order to be believed...with the "proof" of his claim being the ability of the results to better model and predict "real world " performance. Show how the current testing methodolgies "miss" power demands that are actually encountered when riding outside, and how the transient analysis better captures that missing demand. Has anyone seen that?

It's that simple, really. Show us how you got there and how/why it's better. Don't just point to credentials and basically say "this is above your level of understanding".

So far, all we've seen are no details of the data gathering setup or a data summation, partial explanations of the test protocol, and NO evaluation showing better predictive ability of real world demand. Instead, everyone is encouraged to take everything he presents (the main thing being a list with "power values" that are basically just a rank order with no predictive ability) at face value with no questions...and anyone who dares to ask even the simplest question is accused of "hammering" him...please...

[CrankAddictsRich]

This si getting so old... I wish Tanhalt could move his vendetta to another thread. Maybe he culd just keep asking Hambini questions in PM, so the rest of us don't have to see it over and over again.

[hambini]

On to more relevant stuff.

I'm considering (depending on demand), altering the speed charts. Does anyone have a preference? At the moment I've got 30 and 50km/h but I've only done that because 30 is the speed that most people in the UK are riding around at with a decent level of performance and 50km/h because that's the speed the wheel manufacturers use.

[Matt28NJ]

On to more relevant stuff.

I'm considering (depending on demand), altering the speed charts. Does anyone have a preference? At the moment I've got 30 and 50km/h but I've only done that because 30 is the speed that most people in the UK are riding around at with a decent level of performance and 50km/h because that's the speed the wheel manufacturers use.

42-44kph, average speed of the races most of us do.

I wonder why some folks on this thread care to the micro level how much drag the wheels are producing relative to each other if they're not going to race them.

[hambini]

maybe I should rephrase the question and ask what the average race power is (for bunch races) and equate the speed from there

Otherwise the speed chosen my not be appropriate due to drafting.

e.g. an individual riding along can do perhaps 30km/h but when racing they might be doing 40km/h because of the drafting. In that sense the 40km/h speed would not be appropriate.

However if it was a time trial then the raw speed is a good measure.

[otoman]

Great idea. A nice, round 300W would hit the fat part of the bell curve for a lot of body types, race types and speeds. I would consider that a good number for a 40k TT, as well as a wattage that many people might average in a breakaway in a crit or RR.

Alternatively 250W is a more realistic average power for a RR or crit. I wonder if we might find the differences more obvious though at the higher (but still realistic for us 9-5'ers) power. Of course that's your area of expertise...

[cunn1n9]

You, sir, are starting to sound like there is a personal endeavour behind your posts in this thread. To preserve any credibility, please do tell us if it is so.

The ONLY personal endeavor behind my posts is to find out more about this potentially interesting new approach to evaluating wheel performance. All of my questions are to that end, and that end only. If it is a better predictor or real world performance, then I'd like to know that. But, to be fair, the current methods do fairly well, so it'll have to be shown to be quite a significant improvement.

When Hambini states on his blog that "A superior method of analysis is to carry out a transient analysis in a wind tunnel", it is then incumbent on HIM to show in detail why this is so in order to be believed...with the "proof" of his claim being the ability of the results to better model and predict "real world " performance. Show how the current testing methodolgies "miss" power demands that are actually encountered when riding outside, and how the transient analysis better captures that missing demand. Has anyone seen that?

It's that simple, really. Show us how you got there and how/why it's better. Don't just point to credentials and basically say "this is above your level of understanding".

So far, all we've seen are no details of the data gathering setup or a data summation, partial explanations of the test protocol, and NO evaluation showing better predictive ability of real world demand. Instead, everyone is encouraged to take everything he presents (the main thing being a list with "power values" that are basically just a rank order with no predictive ability) at face value with no questions...and anyone who dares to ask even the simplest question is accused of "hammering" him...please...

Mate I am no expert in aero but from what I read of what Hambini said is intuitively obvious. The way wheels have been traditionally tested is in a very controlled environment that is idealised. It amounts to a constant air speed and direction and then adjusts the relative angle of the wheel to get the drag at different yaw angles. Highly understandable that it is done this way as it simplifies things and is I am sure a reasonably good approximation. But it does not reflect the real world where the wind gusts and swirls and is affected by objects such as buildings and trees etc. From the sounds of it more advanced aero testing is done on planes that take this stuff into account as when things go wrong there people die. As Hambini is an aero engineer he is bringing this perspective to wheels and we should be grateful for it. His high level results seem to make a lot of sense to me at least.

To “prove” this to you I am sure we would need a degree in aerodynamics as well as the complicated mathematics that goes along with it and probably safe to say that no one here would understand it properly. However I don’t think it takes much to grasp that the wind when we ride does not come in a constant stream but swirls around. This test is trying to get a more realistic drag value for wheels by taking that into account.

Thanks Hambini - most people here I think really appreciate your work and would welcome more of it.



Sent from my iPhone using Tapatalk

[ryanw]

I’d too say around 280-330w.

Most of my races (Cat3) tend to be around 300w AP / 360w NP.

[tanhalt]

You, sir, are starting to sound like there is a personal endeavour behind your posts in this thread. To preserve any credibility, please do tell us if it is so.

The ONLY personal endeavor behind my posts is to find out more about this potentially interesting new approach to evaluating wheel performance. All of my questions are to that end, and that end only. If it is a better predictor or real world performance, then I'd like to know that. But, to be fair, the current methods do fairly well, so it'll have to be shown to be quite a significant improvement.

When Hambini states on his blog that "A superior method of analysis is to carry out a transient analysis in a wind tunnel", it is then incumbent on HIM to show in detail why this is so in order to be believed...with the "proof" of his claim being the ability of the results to better model and predict "real world " performance. Show how the current testing methodolgies "miss" power demands that are actually encountered when riding outside, and how the transient analysis better captures that missing demand. Has anyone seen that?

It's that simple, really. Show us how you got there and how/why it's better. Don't just point to credentials and basically say "this is above your level of understanding".

So far, all we've seen are no details of the data gathering setup or a data summation, partial explanations of the test protocol, and NO evaluation showing better predictive ability of real world demand. Instead, everyone is encouraged to take everything he presents (the main thing being a list with "power values" that are basically just a rank order with no predictive ability) at face value with no questions...and anyone who dares to ask even the simplest question is accused of "hammering" him...please...

Mate I am no expert in aero but from what I read of what Hambini said is intuitively obvious. The way wheels have been traditionally tested is in a very controlled environment that is idealised. It amounts to a constant air speed and direction and then adjusts the relative angle of the wheel to get the drag at different yaw angles. Highly understandable that it is done this way as it simplifies things and is I am sure a reasonably good approximation. But it does not reflect the real world where the wind gusts and swirls and is affected by objects such as buildings and trees etc. From the sounds of it more advanced aero testing is done on planes that take this stuff into account as when things go wrong there people die. As Hambini is an aero engineer he is bringing this perspective to wheels and we should be grateful for it. His high level results seem to make a lot of sense to me at least.

To “prove” this to you I am sure we would need a degree in aerodynamics as well as the complicated mathematics that goes along with it and probably safe to say that no one here would understand it properly. However I don’t think it takes much to grasp that the wind when we ride does not come in a constant stream but swirls around. This test is trying to get a more realistic drag value for wheels by taking that into account.

I agree that the logic and argument sounds interesting. On the flip side however, the "traditional" methods of measurement seem to do a pretty good job at modeling actual power requirements. So the question then is, how much additional "realism" does this complicated methodology bring...and why are his results (both from the data gathering and the resulting test protocol) at odds in certain cases by quite a bit with other measures? To quote someone I know: "Is the juice worth the squeeze?"

Intuition is great. Actual data and physics is better

[Imaking20]

^Dude. Seriously. Let it go.

I actually think the 30kph is a pretty good benchmark. On most of my solo rides I'm doing ~31-33kph so I figure it's close enough. It does seem like using a wattage target, and race scenario, opens up a whooooole bunch of variability as to what wheel would actually be fastest (given the drafting element, effect of intertia if you're doing something like a crit, etc)... or maybe not?

Anywho, 300w target seems like a good number.

[silvalis]

I don't mind 30kph either as I can easily relate to it.

Too bad the effort required to produce a speed/watt curve is prohibitive.

[MagicShite]

On to more relevant stuff.

I'm considering (depending on demand), altering the speed charts. Does anyone have a preference? At the moment I've got 30 and 50km/h but I've only done that because 30 is the speed that most people in the UK are riding around at with a decent level of performance and 50km/h because that's the speed the wheel manufacturers use.

please do 40kmh, much appreciated.

This is the kind of speeds my team expects everyone to pull at...

[ericoschmitt]

On the current discussion: somewhere between 40-44 km/h seems to make sense, since that's a speed where most dedicated amateurs will be able to do a TT or pull a breakaway for a while before going back to drafting. All that 30km/h tells me up to now is that Yoeleos 88 might be the fastest climbing wheelset around, if the trend follows down to and below 20km/h, since the aero would beat weight savings. It also gives me some parameter to maybe get the average between 30 and 50 and hope it predicts results at that 40-44 range. (could Yoeleo 88 and some lightweight climbing wheelset possibly be tested at 15km/h? who knows.)

------------------------------

I don't understand why most people think TAnhalt's questions are a bad thing at all. At the top of this 40th page he makes a very good abstract of what is missing in Hambini's blog so that the test/method can be then fully accepted as the best way of testing wheels, and we can then forget about everything else and build Hambini a statue, as well as removing rim stickers and gluing ones with his name on all rims I ride after that.

"Intuitively" I also think it makes sense, that it's the right way of testing. I really want it to be, and that his wheel-power chart is the ultimate chart, just to make decisions simple next time I buy wheels. But how can we be sure, really? Ok, we can't have pic/video of the tunnel test (I mean, "ok"), but showing the process of data collection wouldn't have hurt. And what about the bike being used? I think it hasn't been mentioned anywhere so far, and the fork does affect wheel results.

Maybe now you think I'm just another guy like Anhalt... But usually I don't accept things at face value, and that goes to my professional area as well. Why would it be different here? I only asked questions because I'm ~really~ interested in this test, and the results of it. If I just didn't care, I'd have stopped following this topic some 35 pages ago. I'm still here in the hopes some of these points are eventually explained, as many other questions have already been, and I found it very helpful.