Author Topic: Engineer tears apart Cervelo quality control esp. BB shell tolerances (video)  (Read 4666 times)

bludger

  • Randonneur and bargain hunter
https://www.youtube.com/watch?v=GWDztuezn0g

I hadn't heard of this guy before but apparently he's an airbus engineer who is also big into bikes, and sells his proprietary handmade aftermarket PF BBs. One thing is for sure he really doesn't like:

Cannondale

Zipp

And now Cervelo

It's a long video (49 mins) but really eye opening for me. I have always been wary of PF BBs, but I didn't know how blatant poor quality seems to be in the 'premium' bike frame market.  :o
YACF touring/audax bargain basement:
https://bit.ly/2Xg8pRD



Ban cars.

Hambini is astonishingly rude about a lot of stuff. Not just frames, but wheels, aero processes, aero testing in windtunnels and in general. It's strange, because he claims to be an empiricist, but puts theory above data (ie how much faster people go for similar watts) when criticising wind tunnel time.

I've not come across a video of his where he does like a piece of bike equipment.

it is a source of considerable irritation to competent  professionals in any field when products are sold to consumers as 'the best', (usually on a tidal wave of hype, half-truths and outright bullshit, much of which is lapped up by an incompetent and unquestioning cycling media, with strong 'windsock' tendencies) when any proper  scrutiny of these 'facts' shows that the claims are wildly overstated (may only apply under certain -highly unlikely- circumstances) or are downright wrong.

I'm not claiming that I have myself reviewed his every utterance or indeed that I am competent so to do. But in areas where those things do apply I have largely agreed with what he has said, (which is a very rare thing).  In addition I read something I didn't agree with on his website and sent him an e-mail; probably it was just a typo or an oversight of some kind but I got back a polite and well-considered response.

I think the world of cycling is all the better for having folk who (for the most part) don't have a  substantial vested interest in such products yet have the interest and technical competence to  review the claims made by manufacturers. 

I recently met an engineer who used to work for Cervelo. They do some clever stuff, no mistake about it, (and he wasn't about to air dirty laundry either  by any means) but they do a lot of things in exactly the same way as other cycle manufacturers do it too, and this often is not without compromise.

One of the greatest fallacies in cycling is that cartridge bearings are 'always better'.  They can be excellent, but in bicycles they are usually nothing of the kind; the usual outcome is that a bearing (of unsuitable, unknown or uncontrolled tolerance and specification) is boshed into a housing (similar) by someone who neither knows nor cares about such things. That it works at all for more than five minutes is a ruddy miracle; that it is usually very far from perfect is almost guaranteed.

It is worth noting that both Campag and Shimano have stuck with cup and cone bearings for their top-of-the-line hubs; this is for the very good reason that they can always be adjusted (by a competent mechanic) to single digit (in microns) clearances/low preloads, which would be very difficult to consistently achieve using cartridge bearings.  Of the current range of bottom bracket 'standards' it is very much a question of 'choose your poison'; none of them are perfect or are ever likely to be.

cheers

Hambini is astonishingly rude about a lot of stuff. Not just frames, but wheels, aero processes, aero testing in windtunnels and in general. It's strange, because he claims to be an empiricist, but puts theory above data (ie how much faster people go for similar watts) when criticising wind tunnel time.

I've not come across a video of his where he does like a piece of bike equipment.

He also seems to go for named individuals in companies, almost to the extent of stalking.  I'm surprised he hasn't faced legal action.  And his use of profanity, whilst mildly amusing in a puerile sort of way, stands at odds with his purportedly empirical analysis - almost as if it were an add-on to get clicks (heavens forbid  ;D).
The sound of one pannier flapping

bludger

  • Randonneur and bargain hunter
I don't see how swearing contradicts expertise. I've met people who swear like mad in my career who were excellent at their jobs. People spend thousands of pounds in bikes and bike stuff. I know some students who work day in day out either studying or money working who use their cash to buy cervelo and zipp products after months of saving, I'd be pretty annoyed if I was wise to their antics too.

I don't think he even swears that much anyway. Perhaps that's my rough construction ways showing.

Quote
I've not come across a video of his where he does like a piece of bike equipment.
He seems to rate Look and Time pretty highly. In a recent vid he's fettling a 2003 Look frame into a racing bike which he reckons will be just as quick as a modern whizz bang.
YACF touring/audax bargain basement:
https://bit.ly/2Xg8pRD



Ban cars.

Gattopardo

  • Lord of the sith
  • Overseaing the building of the death star


One of the greatest fallacies in cycling is that cartridge bearings are 'always better'.  They can be excellent, but in bicycles they are usually nothing of the kind; the usual outcome is that a bearing (of unsuitable, unknown or uncontrolled tolerance and specification) is boshed into a housing (similar) by someone who neither knows nor cares about such things. That it works at all for more than five minutes is a ruddy miracle; that it is usually very far from perfect is almost guaranteed.



I disagree with you in someways.  With current machinery things can be made to very close tolerance, both bearings and housings.  Is this the same with poor quality bearing balls and poorly machined cups and cones.  But here is the saving, assembling bearing and hubs takes less time and less skill (so lower pay) that fettling cup and cones to close tolerance and very difficult to do with a machine.


Re cartridge bearings; a good telltale is when cartridge  bearings suffer a tiny bit of water ingress/damage and become rough. Most often these bearings are (mysteriously) play-free and if not perfectly then acceptably smooth when removed.  How does this happen?  Well there is only one explanation for this, which is that the bearing, when installed, sees a needlessly high (and probably completely uncontrolled) preload. 

This is a very common happenstance in cheap stuff but expensive stuff (which ought to be better) is not immune from it either.  It is certainly annoying if not unacceptable that the ship is spoilt for a ha'penworth of tar in this way. The seeds of imminent destruction are thus built into many (most) bicycle cartridge bearing installations, and the resultant reliability/life expectancy of the bearings is usually a tiny fraction of what ought to be the case as a result.

Oh, and BTW you don't need vast reservoirs of skill to set up adjustable bearings (there are plenty of near-illiterate and/or poorly educated mechanics that are very good at it); by contrast no amount of practical skill will ever make a lot of modern designs any more than randomly reliable, because there are just not enough basic controls exercised in their implementation, and/or it is fundamentally flawed in some way.

cheers

Hambini actually has a video where he's tearing apart cup and cone and cartridge bearings. His conclusion was that the application was crucial as to which bearing you should use, and that for wheels, cartridge bearings were superior to cup and cone bearings. IIRC, he theorised that the reason why Shimano still use cup and cone is because they require lesser tolerances to work effectively and are therefore cheaper to manufacture half decent wheels.
Video here: https://www.youtube.com/watch?v=1WMGrgWVfqA

I have not seen that video but it is very difficult to do a fair test on hubs in particular, not least because whatever steps are taken during manufacturing to control the tolerances of cartridge bearings, all that can be thrown out of the window when

a) the wheels are built and
b) the wheels are installed in the frame.

Spoke loadings on flanges can turn a light interference fit of the bearing into a loose (clearance) fit. This obviously changes the bearing  preload out of all recognition, and it is dependant on the spoke tension and the spoking pattern.  In addition to that, the QR load changes the loading on the bearings too.  Very few hubs which accept cartridge bearings are immune from both effects and many are subject to both of them.

The other thing to bear in mind is that cartridge bearings typically have smaller rolling elements in than cartridge bearings do. This fundamentally alters the rolling resistance coefficient of the bearing, under load.  Often the load on a cartridge bearing is not at all well controlled (because of the effects described above) and it is not at all unusual to find that the majority of the service load is preload, not applied load.

To draw conclusions about the suitability of one bearing type over another it is necessary to make assumptions about the quality of the bearings and the tolerances to which they are installed/adjusted.  I don't think you can say that hubs with cartridge bearings will definitely have a certain drag (not without actually measuring it under true service loads) because this might vary with spoke tension and QR pressure.  The same things are true of cup and cone hubs for course, but you can make the (not unreasonable) assumption that the hubs will be adjusted to allow for installation and wheel build loads and will be installed in a consistent fashion.

The idea that making cup and cone bearings "is cheaper" might be true for lower end hubs but it certainly doesn't explain Dura Ace, Ultegra or Campagnolo cup and cone bearings.

cheers

I may be missing something but isn’t it the job of the fancy metal tube to isolate the bearings from the misalignment of the frame?

I’ve not come across this guy before but “why product X sucks” is a very easy way to get big views on YouTube and there’s a big incentive to ham up your critiques beyond what may or may not be wrong with the product.

bludger

  • Randonneur and bargain hunter
It's true that there is an incentive to ham up product failures but he isn't just a randomer talking shit for clicks to get views to his monetised YouTube channel, the guy literally makes bespoke BBs for stage winning pro tour teams by hand *as a side hobby* outside his lead engineering role for Eurofighter GmbH. That's what has me so interested in the video, 'x is shit' videos are a dime a dozen as you say but stuff by expert engineers really railing on avowedly 'premium' products are pretty unknown in my experience. Indeed I should think there would be a much stronger financial incentive to be a marketing patsy for the cycling industry than furiously weeing inside the tent.
YACF touring/audax bargain basement:
https://bit.ly/2Xg8pRD



Ban cars.

Gattopardo

  • Lord of the sith
  • Overseaing the building of the death star
Re cartridge bearings; a good telltale is when cartridge  bearings suffer a tiny bit of water ingress/damage and become rough. Most often these bearings are (mysteriously) play-free and if not perfectly then acceptably smooth when removed.  How does this happen?  Well there is only one explanation for this, which is that the bearing, when installed, sees a needlessly high (and probably completely uncontrolled) preload. 

This is a very common happenstance in cheap stuff but expensive stuff (which ought to be better) is not immune from it either.  It is certainly annoying if not unacceptable that the ship is spoilt for a ha'penworth of tar in this way. The seeds of imminent destruction are thus built into many (most) bicycle cartridge bearing installations, and the resultant reliability/life expectancy of the bearings is usually a tiny fraction of what ought to be the case as a result.

Oh, and BTW you don't need vast reservoirs of skill to set up adjustable bearings (there are plenty of near-illiterate and/or poorly educated mechanics that are very good at it); by contrast no amount of practical skill will ever make a lot of modern designs any more than randomly reliable, because there are just not enough basic controls exercised in their implementation, and/or it is fundamentally flawed in some way.

cheers

Poor quality bearings are usually the issue, and branded don't seem to be much better.  Recently have found bearing, even the quality ones, seem to be lacking in grease.  Usually replacing the grease solves the problem.  Ducati recently had issues with wheel bearings failing due to no lubrication.

I have not seen that video but it is very difficult to do a fair test on hubs in particular, not least because whatever steps are taken during manufacturing to control the tolerances of cartridge bearings, all that can be thrown out of the window when

a) the wheels are built and
b) the wheels are installed in the frame.

Does that mean you would agree with him that cartridge bearings would be less draggy if we are referring to manufacturer built wheels using through axles?
It might be the case that he's talking theory rather than actual empirical evidence again (this is my quibble with him on aero specific matters). I have very little knowledge of bearing tolerances so I'm just reporting what that video says - maybe it's worth watching it to check that I've not missed something crucial before you throw his conclusions out of the window?

And FWIW, Campag now use cartridge bearings in their top of the line wheelsets - they seem to have 2 different types of ceramic bearings available: https://www.campagnolo.com/UK/en/Technologies/wheels 

I don't think I have thrown anyone's conclusions out of the window but I have pointed out that to draw any conclusions about cartridge bearings requires testing under representative conditions (which includes spoke, QR and service loadings) and/or assumptions about the tolerances of the bearing and its installation.  IME you could test some parts and then get rather different results with others, because the (non adjustable)  tolerances would have to be held within a handful of microns.

I don't think it makes that much difference if you are talking about through axles or not; the flanges still see spoke loads and these will interfere with the fit/preload of the bearing. The skewer will still compress the spacer between the bearings and that will alter the contact angle of the bearing.


Campagnolo have had CULT BB bearings for some time so it isn't a big surprise to see them in hubs too.  The new one is USB (ultra smooth bearing) which is seemingly less smooth than CULT or something.  In USB they mention the bearings have  low friction once and smoothness/low weight  many more times.   Well  cynical little old me says that honey is smooth but not low friction and the bearings can have low friction but when they are installed the friction could increase dramatically.  Low weight is an easy (but almost irrelevant) selling point.

CULT bearings can have low measured friction because they are only lubricated with (thinnish) oil and they have no seals to drag.   Big whup; folk were setting up TT bikes this way decades ago. The main difference is that campag are selling this as being suitable for road use in the longer term but  it isn't.  The bearings have ceramic balls and nitrided stainless steel raceways. Because the bearings are not protected the raceways are soon damaged and the friction in these bearings can increase to more than a standard bearing after a fairly short period of time. The nitrided stainless steel raceways can easily corrode in our delightful climate too. 

The other differences are  that the smaller balls found in such bearings  have inherently higher rolling resistance than bigger balls and there is likely to be appreciable  preload on cartridge bearings.  The net result of all this is that the friction in such bearings might vary from sample to sample once installed and any benefits might be rather temporary, or only apply when the bearing is not loaded.

Cyclists usually get as far as twirling hub spindles between their fingers and spinning wheels; neither of these tests involves the applied service load so don't tell you very much about rolling resistance in service.

If you want low friction in cartridge bearings the best you are likely to manage is to buy bearings with a small internal clearance and then fettle the housings/bearings so that the interference fit doesn't quite remove all the clearance in the bearing. Then remove the inner bearing seals permanently (they serve absolutely no useful purpose), remove the outer seals temporarily (or replace them with shields), use a lubricant that isn't too viscous yet provides low friction/high film strength at the speeds/loads  in question.  The do your race or whatever and afterwards expect to have to replace the seals and lubricant at least, if not the bearings as well if they get contaminated.

You will do better than a CULT bearing with a few miles on it and you never know you might do as well as a high quality cup and cone bearing...

cheers



quixoticgeek

  • Mostly Harmless
I don't see how swearing contradicts expertise.

Try being a woman in tech who swears...

J

--
Beer, bikes, and backpacking
http://b.42q.eu/

quixoticgeek

  • Mostly Harmless
I have always been wary of PF BBs, but I didn't know how blatant poor quality seems to be in the 'premium' bike frame market.  :o

Remember, the design purpose of Press Fit Bottom Brackets was to speed up manufacture and thus save the manufacturer money (like it would ever be passed on to the consumer...)

Given how hated they are, it still amazes me anyone accepts them on a frame...

J
--
Beer, bikes, and backpacking
http://b.42q.eu/

bludger

  • Randonneur and bargain hunter
I can understand PF in the super high performance bike market i.e. pro teams' racing bikes where every saved gram and watt adds up, but for the wider mass market, bleurgh.
YACF touring/audax bargain basement:
https://bit.ly/2Xg8pRD



Ban cars.

Gattopardo

  • Lord of the sith
  • Overseaing the building of the death star
I can understand PF in the super high performance bike market i.e. pro teams' racing bikes where every saved gram and watt adds up, but for the wider mass market, bleurgh.

If the pro teams have and I do therefore I am a pro.


bludger

  • Randonneur and bargain hunter
I mean the guy has a literal PhD in aerodynamics and his work at Eurofighter is the field at the most senior level in the world. I think he has every right to be 'gobby', it must be nauseating watching people getting conned out of their dough by junk claims.
YACF touring/audax bargain basement:
https://bit.ly/2Xg8pRD



Ban cars.

Karla

  • car(e) free
    • Lost Byway - around the world by bike
Yes, because you are amply qualified to judge different people's research.  Clearly.

As long as he's got a literal PhD  ::-).   Still has a bad attitude AFAIC.
The sound of one pannier flapping

bludger

  • Randonneur and bargain hunter
What's the worse attitude - swearing in YouTube videos or conning people out of their dough with fraudulent claims about aerodynamics? Not even talking specifically about Flo Wheels here. The dimpled zipp wheels are a particularly grave example.
YACF touring/audax bargain basement:
https://bit.ly/2Xg8pRD



Ban cars.

Having a PhD and having a job at a fancy company does not disprove him either being wrong or being an arsehole.

(In fact, if anything...)

Karla

  • car(e) free
    • Lost Byway - around the world by bike
There are many people on this very forum with literal PhDs.  Some of them are literal arseholes too ;)