I wouldn't say I don't believe the high yaw data, I think it's accurate for what it measures, but it's less applicable for simulating on the road performance. For me, the issue is that for conditions near the separation yaw angle, the steady state assumption used in the wind tunnel doesn't do as good a job representing conditions in the field as smaller angles. Away from the separation point, as angle or speed changes, you might move a little up or down the drag curve but you always follow the curve. Near the separation point, however, you run into issues of hysteresis. That is, let's say the average wind speed and angle are such that you're just below separation and a gust of wind comes along. The higher actual wind speed will momentarily increase the yaw angle and the flow separates (increasing drag), but then when the gust passes and the wind speed drops back to the average, the flow will not reattach in the same way it separated when the speed increased. In other words, you're following a different curve on the way down than you did on the way up and the average drag will not be the average over the steady-state curve.

It's possible to measure this hysteresis in the wind tunnel and it's happening more and more, but since it depends on the rate of change of the wind, it's difficult to get data that can be generalized.

I understand what tolerance is, but from your comment it makes me think you didn't even watch the video I linked. When paying thousands of dollars for a frame I'd expect the tolerance to be a little better than a kmart special..

If you look at WT data for wheels, bike, or bike+rider there is typically a pretty huge drop in aero drag with yaw. If this were true you'd expect the fastest conditions to be a strong 90deg crosswind for an out-back course. In my experience, this is never the case. It's possible that the tire scrub or misalignment due to dealing with wind is the reason, but that's not been measured AFAIK. I suspect most of the discrepancy is due to flow attachment being much less "clean" in real wind.

Didn't see any video link in the post I quoted...

---

Strava

Super-quick response - might deserve more later, but inherent in a WT test at yaw is lower velocity wind in the bike axis. Many real-world situations to create that would mean a lower bike velocity so you need to consider that when looking at results.

I wonder if we should change tunnel velocity with yaw angle so the bike-axis component of the wind is the same at all angles ... or we just look at cda, not drag or watts :)

---

-------------
Ed O'Malley
www.VeloVetta.com
Founder of VeloVetta Cycling Shoes
Instagram • Facebook

Posted again for you.. This is a disgrace..

https://m.youtube.com/watch?v=pqsRW9zX-SQ

Yes, there are some calculations required. It's been a long time since I looked at it, but as I recall the tunnel data *is* corrected for angle (cos^2?), but a crosswind both ways results in a higher bike axis V^2 component than the no wind situation.

Look at the graph I posted above. 10 deg yaw (+-) is ~2080 on one side and 1980 on the other; avg ~2030. And at zero yaw it's ~2300, so at +or- 10 deg yaw we only have ~88% of the drag. I'll let you calculate a viable V^2 increase from a crosswind that results in 10 deg yaw. ;)

My issue with the manufacturing tolerances and his whole broad statements about brands:

1/ By definition, he is going to be working with frames that are sadly out of whack. To extrapolate from that into a general statement about the manufacturing tolerances of all brands is just bad science. Two words: sample size. Without knowing the percentage of defects, you cannot make any statement about tolerances or QC.

2/ He knows machining tolerances for metal. Does that make him an expert on what tolerances should be for carbon fiber or an expert on carbon fiber manufacture? No.

3/ His whole act is now based on being exceptionally provocative. But that also means nuance gets lost. His earlier videos were very informative and reasoned, and got hardly any views. Ever since he started acting up, the viewership has soared. Fair play to him. But realize that he is now in the "entertainment" business and not so much in the "scientific process" business.

4/ Given the above, as well as some of the other issues (the aero test with no evidence to show it happened and that awful fake legal letter), he has serious credibility issues in my eyes: I take with a grain of salt anything he says which isnt backed up by evidence. When he shows me a poorly made BB shell and shows how off the tolerances are - i believe that. When he uses that n=1 sample to say "Brand X is shite", I dont. When he refers to some mythical NDT with no evidence to show that this actually happened, i dont believe him.

Personally, i find his shtick a little boring and overly repetitive now. But it does appeal to people who confuse a lot of cussing and broad generalizations with "telling it like it is" and "sticking it to the man". I do find "Dicktwitch" somewhat amusing, i have to say.

---

--
Those who are slower than me suck.
Those who are faster than me dope

Tolerance and QC are 2 different things.

No manufacturing process is perfect. And very rarely is every single product checked individually. To simplify a little, manufacturers set up a test procedure to ensure a QC standard which ensures a maximum defect percentage of X% (be it 1%, 0.5%, 0.003% or whatever). This is as much a financial decision as a manufacturing process decision.

So taking just BB tolerances as the only element of quality, in this case:
- manufacturing tolerance would be maximum acceptable variation in BB dimensions from spec (and here, Hambini uses the tolerances of metal machining to evaluate carbon fiber - that is incorrect and obvious to anyone with any basic knowledge of manufacturing processes)
- QC would be the percent of frames that slip through where the BB is out of spec

While i agree that it sucks bigtime to get a frame that is badly out of spec (and that Boardman is pretty bad), you cannot use a single sample to make a statement about the manufacturing tolerance or QC of the product or brand as a whole. Also, there remains the question about what happened to that specific frame before it came to the show - given Hambini's approach, i think a certain healthy skepticism is valid.

For the record, i do think that some of the common BB standards probably require a greater tolerance than may be possible for products built to specific price points/volumes. But that is a broader discussion involving the trade-offs between weight, stiffness/ride quality and also what manufacturers think customers want. But to argue that this means that all manufacturers are just lazy/sloppy and deliberately making poor quality products is specious and lazy thinking.

---

--
Those who are slower than me suck.
Those who are faster than me dope

You nerds who aren't golf nerds forget that dimples on a golf ball aren't just to reduce drag. It also affects the lift of the ball. The ball manufacturers have dimple patterns somewhat based upon guesses at the initial launch velocity.

You've got weekend warriors with clubhead speeds well below 100mph. Then the amateur competitors and pros with speeds above 100mph.

They tailor the dimple design so that you get that more optimized "out, then up, then drop with spin" kind of ball flight. You don't want a parabola ball flight for golf, you wouldn't hit it very far or it wouldn't land very nicely.

Given all that, applying that to a bike wheel is a bit absurd. A golf ball, is after all, a ball. It's not an air foil. The dimple size on the wheels don't even seem to make sense to me given the speed of the wheels versus a 100+ mph golfball.

Not engineering, just common sense.

It was an armchair idea that had great sales potential.

Bit confused by this. Each BB spec has a specified tolerance. Pressfit cups have a wider tolerance spec than pressfit bearings. Whether the BB is a carbon shell or metal shouldn't make a difference.

Edit: found the specs for BB30 vs PF30. BB30 shell diameter can be 41.96 + 0.025 mm. PF30: 46 -0.05 mm. PF30 tolerance twice that of BB30. No mention of the BB material.

So if we do a quick calculation - testing at 30mph in the wind tunnel:

At 0 Yaw, obviously, bike-axis wind is 30mph.
at 15 degrees yaw, bike-axis wind is 30mph(cos(15 degrees)) = 29mph.

That seems like a small difference until you realize that at those conditions, and depending on specifics, that represents about 31 watts less power to maintain speed if that were just a straight headwind (same CdA as 0 yaw)

---

-------------
Ed O'Malley
www.VeloVetta.com
Founder of VeloVetta Cycling Shoes
Instagram • Facebook

There's very little common sense in aerodynamics. Way too hard to make predictions about anything unless guided by experiment and experience.

There's no way I'm watching a whole hour of his pontificating... but from watching just the first few minutes -

He looks at the headset bearing cups and says, oh, I'm just going to assume that these were machined before being inserted and therefore they are not aligned. Yeah... good science.

As for the bottom bracket, let's just assume he is right about what bottom bracket it's made for and his gauges are machined properly. So someone in the world found a bike with the bottom bracket out of spec. Ok. Happens with every kind of product. In this case it really sucks that they could not get a warranty return on that - sounds like it was the fault of the retailer who seems to have gone out of business within days, so they had other crap going on. Should have called Boardman directly.

But think of it this way - alignment of the bearings, whether in the bottom bracket or in the headset, needs to be good enough so that when the spindle or steerer tube is inserted, they can turn freely with just the friction imparted by the bearing itself. Reality is that is almost always the case. How often does one assemble a bike and the bottom bracket or the steer tube have binding that prevents free spinning? It's quite rare.

What can you conclude from this? You can conclude that the tolerances - both geometric and dimension - are appropriate for the application. You do NOT want to over-spec tolerances. Prices can go through the roof really quickly.

I design things in both the bike industry and in the engine industry. I just pulled up a print for a piston I designed. It has dimensional tolerances of 0.01mm. It has a surface finish tolerance of Rz = 4 microns (look that up if you want to know how tough that is to do). The diamters of the press-fit bearing shell openings are allowed to be between 22.483 mm and 22.471 mm while ALSO maintaining both position and circularity within 0.02mm. And then it needs to maintain those geometric tolerances when it's heated up to 700 degrees C (1300 F). If we insisted on bike frames to be made to these tolerances, they'd probably cost $100,000 and they would probably perform worse because they would probably have to be made out of metal. Getting those kinds of tolerances in carbon fiber, even with machining both the frame and the metal inserts after they've been installed - very well might be impossible.

---

-------------
Ed O'Malley
www.VeloVetta.com
Founder of VeloVetta Cycling Shoes
Instagram • Facebook

Sorry, i should have been clear - I wasnt referring to defined acceptable tolerances but ease of attaining them in manufacturing.

Hambini likes to bang on about how "this frame is badly out of tolerance, and so the manufacturer doesnt know what they are doing" - just because it is easy to machine aluminum to specific tolerances doesnt mean that it is easy to manufacture carbon to the same tolerances. So the lack of tolerance doesnt necessarily imply sloppiness/incompetence on the part of the manufacturer, but just could be a general limitation of precision possible with carbon.

That said, this is not to give the bike industry a free pass. It should collectively accept some responsibility for jumping on the pressfit bandwagon if the standard tolerances are hard to achieve in manufacturing - but it could also be that they considered it a tradeoff for lightness/stiffness/aero/whatever. And without knowing actual % of numbers that are out of spec (ie, a sample size larger than the n=1 that Hambini uses for his analysis), it is hard to say to what degree this is a problem.

Also, IF that Boardman frame did indeed come as is, it should definitely have been replaced by warranty. That's either bad service on the part of Boardman, or there is something about the frame's history that we are not privy to.

---

--
Those who are slower than me suck.
Those who are faster than me dope

That's either bad service on the part of Boardman, or there is something about the frame's history that we are not privy to.[/quote]
Quote.

I’m going with something not quite right. He has more than a few convoluted “stories” about how these frames ended up in his hands.

Not really sure how things work in the UK but the majority of issues relating to bike frame warranty denials are related to not original owner, or user error.

I find it weird that someone would just hand over a brand new frame to hambini “to have a go” at fixing it, as opposed to just getting a new frame from the manufacturer.

Maurice

Talk about placing your life into questionable hands...

---

-------------
Ed O'Malley
www.VeloVetta.com
Founder of VeloVetta Cycling Shoes
Instagram • Facebook

The hysteresis thing might be a significant factor - but it might not be either.

I have not seen any transient test data to even determine what the hysteresis of detachment/reattachment on a bike + rider is in different conditions. If you know of such data, I'd love a link. I would think recording drag while rotating the table in most tunnels would lead to a lot of outside influences on the data and make it useless - but maybe that's not right. Nor have I seen any data on how likely it is for a rider to spend a lot of time in conditions in which attachment and detachment is repeatedly happening back and forth. So I think it's interesting academically to think about but whether or not it is a significant factor is unknown. At least to me.

---

-------------
Ed O'Malley
www.VeloVetta.com
Founder of VeloVetta Cycling Shoes
Instagram • Facebook

Regarding the comment that "there's very little common sense in aerodynamics":


Indeed. One such counter-intuitive example: a truck with tailgate up is more aerodynamic than when the tailgate is down. It's been proven in wind tunnel tests multiple times.
https://www.autoblog.com/...e-up-down-mpg-video/

I feel the same way. Given his history of doing so, i wouldnt put it past him to make up stories in order to get eyeballs. About the only credible reason i can think of Boardman not taking a genuinely new frame back could be related to the fact that the shop where it was purchased has shut down - and even then, it would be in egregiously poor form of Boardman to dick over a customer like that. I cannot see any company in this day and age doing so.

It is far more likely that Hambini is using this as an opportunity to knock Boardman, who - going by previous posts - seemed to have earned his ire for some inexplicable reason.

Bit of a shame. I used to enjoy his earlier videos focusing on the technical solutions to BB issues. He certainly knows his stuff there. The cussing doesnt bother me one bit - but when he's obviously cussing just to fill time, I lose interest. What am i, 6 years old, that just hearing cusswords is going to grab my interest? Get on with it already. Add to that gross exaggerations and unsupported conclusions, and he's just a circus sideshow now. Sadly, there are far too many people that conflate his obvious expertise on BBs with knowledge on other areas, and fall for his pseudo-tech babble and unsupported conclusions.

---

--
Those who are slower than me suck.
Those who are faster than me dope

Thanks. No way you can achieve the same tolerance control with carbon vs metal. My understanding is that the material shrinkage is very difficult and expensive to control. Would trust a metal BB insert in a carbon frame much better. Can still get problems but at least it's machinable if you do.

I think Luescher in one of his videos mentions a conversation he had with a Factor engineer. They were considering making frames under contract for a well-known brand and told them that it was simply not possible to make the BB shell to the required tolerance at the desired price. It would cost an extra $20 per frame to do this. The brand walked away and found someone else.

Sorry, I can't recall where I saw the data, but it has been discussed several times. Josh Poertner might have talked about it in an early marginal gains podcast on hysteresis. The data aren't taken while the table is rotating; they do a sweep increasing yaw and measuring at fixed angles then sweep back decreasing the yaw (or vice versa). Different drag values are seen at the same angle depending on the direction of the sweep which makes sense as one is for attached flow and the other with the flow separated.

I don't know how much time a rider spends in this regime either, but the discussion was around why wind tunnel data might be less reliable or usable at high yaw. So the hysteresis might not be significant for the majority of conditions seen on the road, but it likely is when looking specifically at behaviors around the separation point.

James Huang has an excellent interview with the engineer, Rob Gitelis, on his Nerd Alert podcast https://podcasts.apple.com/...8732?i=1000468588411 starting about 17:30 in. They talk about this at length.

Because someone else brought Hambini up, this may be the time to mention that Adam Kerin (Zero Friction Cycling) also had a run-in with Hambini over the latter's data on bottom bracket friction.

The link above is a bit of a long piece on hybrid ceramic vs steel bearings. Kerin does tend to write a bit long-winded (no offense intended). But basically, my read is that

a) Hambini claims to have done a 10,000km test for bearing lifespan and friction in hybrid ceramic vs steel BB bearings. He claims that hybrid ceramic bearings don't last as long as steel.

b) Kerin sees some anomalies in Hambini's data. Kerin also seems to think that some of Hambini's claims don't stack up with known facts. For example, good hybrid ceramic bearings have much longer life than steel in industries where they are also subject to heavy loads. Hambini asserts that in cycling, vibration is what does hybrid ceramics in, because you're causing hard ceramic balls to vibrate against steel races.

c) Similar to the wheel test, Hambini refused to discuss his test protocol.

d) Also, Kerin asserts that a 10,000km test is an enormous undertaking; he thinks that many bearing manufacturers haven't done this sort of thing, and is wondering how Hambini pulled it off.

e) Some of the critical points in Hambini's bearing YT vid are based on Hambini's own calculations. Hambini isn't willing to discuss this either.

Back to the guy who brought Hambini up from the dead. I agree that we ought to be more skeptical about industry claims. I disagree that this makes Hambini's overall toxicity worthwhile.