Motor doping thread

[ScienceIsCool]

Re: Re:

When people say hub, they always think of the shell between the flanges. But there's a relatively big volume occupied by the freehub mechanism... I'd be shocked if a three person team couldn't have a functioning prototype in 6 months. This is something a university student team could do as a school project.

John Swanson

Ok, it's time we put our money where our mouths are.
I found a cross section of a Shimano rear hub from som patent link, scaled it to the known width (130mm) and went to work.
I figured you need a quick release. I reduced it to 4mm from the standard 5.
The axle OD I set at 8mm, maybe 7 would work too.
For the hub shell I figured 1mm wall thickness is not impossible, same as 0.5 air gap betweeb hub shell and rotor.
Wheel bearings are spread out as far as possible in the model.
The free hub doesn't need bearings, it can just slide on the hubshell. Mind you, there is no clutch yet.
I put a rotor with 1mm shell thickness on the axle/stator.

Do you see how the cage and windings would fit on the axle while the rotor holds the magnets on the inside?
No matter how rare your magnets are , it's very far from 'easy'.

And how would you route the power leads to the inside of the rotor+bearings?

What about the 1:30 gear box, is that doable too? It must use two stages, no?
Got an idea for a fully functional freehub clutch that fits in within the standard Shimano profile but doesn't steal space from the motor?
How would you make sure this mechanical masterpiece never breaks down in a public, sensational way?

Maybe you envisoned something completely different. Looking like a standard racing wheel hub of course.

Your call.

There's some low-hanging fruit that can give you more design freedom. Make the axle out of tungsten (twice the Young's modulus as steel) and you can reduce the diameter by a couple of millimeters. Same for the quick release. Use a planetary gear out of nylon and that'll be both very small and quiet. If you want, the magnets can actually form the shell with epoxy filling the gaps. Machine the surface and wrap with a thin layer of carbon if necessary. I'm sure you could also push the boundaries with the bearings (more balls in a smaller race) and live with replacing them more often.

That's just off the top of my head.

From your diagram and my mods to the design, the total internal volume for the motor is roughly pi(27mm^2-6mm^2) x 50mm. That equals a bit more than 100,000 mm^3. What does an off the shelf motor need? First Google hit is Transmotec. http://download.transmotec.com/eng/dc-motors/no-gear/round-type/Transmotec-DC-Motors-0.1W-200W-eng-F.pdf They fit a 60 Watt motor in a package 46 mm x 30 mm diameter. That's 130,000 mm^3. Pretty close considering they use extraneous hardware that we wouldn't include.

And that's for a 24 VDC motor which we would never do. We'd want to go very high voltage and lower speed which should give a boost in power to our design.

So there you go. Zero engineering, back of the envelope says that it's possible.

John Swanson

 

[sniper]

Respect. That's a phenomenal exchange of posts.
I wanted to say sexy, but, whatever, einfach hammermäßig.

 

[Nicko.]

I eked out 25mm ID, not 27. With your numbers (you forgot to use the radius btw), the avalilable volume is only 70% of your example motor.
Your example motor has 52W, not 60. With your 'volume equivalent scaling', we end up with ~36W.
This power is achieved by rotating at, sit down for this one, 22000rpm. That's 22 thousand rpm.
The wheel rotates at 400rpm (50kph). Your gearbox needs to be 55:1
A planetary gear crammed in between a fixed axle of 6mm (your number) and an outer shell of 30mm (Shimano) may, might give you 4:1.
Three of those planetary gears gives you 64:1. We're home!

...not!

How would you arrange more than two planetary gears in 11mm of axial space? Or do you have some other solution for the gearbox?
Back of the envelop engineering is all I'm asking, but it must be sound and realistic.

edit: 36W, not 42. even worse...

 

[ScienceIsCool]

Re:

I eked out 25mm ID, not 27. With your numbers (you forgot to use the radius btw), the avalilable volume is only 70% of your example motor.
Your example motor has 52W, not 60. With your 'volume equivalent scaling', we end up with ~36W.
This power is achieved by rotating at, sit down for this one, 22000rpm. That's 22 thousand rpm.
The wheel rotates at 400rpm (50kph). Your gearbox needs to be 55:1
A planetary gear crammed in between a fixed axle of 6mm (your number) and an outer shell of 30mm (Shimano) may, might give you 4:1.
Three of those planetary gears gives you 64:1. We're home!

...not!

How would you arrange more than two planetary gears in 11mm of axial space? Or do you have some other solution for the gearbox?
Back of the envelop engineering is all I'm asking, but it must be sound and realistic.

edit: 36W, not 42. even worse...

Why in the world would you use the first motor I found? It's merely an example of what's possible and sets volumetric power expectations to within a factor of 50% or so. Truth is that you wouldn't want a BLDC motor at all, you'd want a stepper. High torque and you can build it so that it's operating at peak power at ~1000 rpm. I'm not going to run the numbers, but I bet you could get peak at substantially lower rpm. Lots of power/torque/efficiency curves out there for anyone who wants to look.

Here's a picture of a nylon planetary gear. Notice how thin it is. The reduction ratio is calculated as 1+R/S, where is the number of teeth on the ring, and S is the number of teeth on the sun. These numbers will be proportional to the circumference and hence diameter. If the sun is 6 mm and the ring is 24 mm, you'd get a reduction of 5:1. Work hard and I bet you could squeeze a 6:1 ratio That would put the hub at 150-200 rpm at peak.

Circumference of a bike wheel is just over 2m. At 40 km/hr, that give a hub speed of 300 rpm. So we've got *plenty* of space for a reduction gear.

John Swanson

 

[Teddy Boom]

Re:

This series of posts is really great, and much appreciated.

edit: 36W, not 42. even worse...

36W would have an absolutely enormous effect in race conditions though, right? Even 10W would be very significant?

Also, 1000RPM really is getting fairly low, for ideal power to weight or power to volume ratio, isn't it? Certainly 22,000 RPM is an exaggeration in the other direction, but..

 

[ScienceIsCool]

Re: Re:

This series of posts is really great, and much appreciated.

edit: 36W, not 42. even worse...

36W would have an absolutely enormous effect in race conditions though, right? Even 10W would be very significant?

Also, 1000RPM really is getting fairly low, for ideal power to weight or power to volume ratio, isn't it? Certainly 22,000 RPM is an exaggeration in the other direction, but..

It depends entirely on the motor type. I'd expect a small brushless DC motor to run at mid-boggling speeds, but very low torque. Stepper motors give you a lot more heft as you approach stall, but require a driver and are all round more expensive. And yeah, I think even 40 Watts in a perfectly hidden, silent hub would be worth a *lot* of money to certain people.

John Swanson

 

[Nicko.]

Re: Re:

Why in the world would you use the first motor I found? It's merely an example of what's possible and sets volumetric power expectations to within a factor of 50% or so. Truth is that you wouldn't want a BLDC motor at all, you'd want a stepper. High torque and you can build it so that it's operating at peak power at ~1000 rpm. I'm not going to run the numbers, but I bet you could get peak at substantially lower rpm. Lots of power/torque/efficiency curves out there for anyone who wants to look.

Here's a picture of a nylon planetary gear. Notice how thin it is. The reduction ratio is calculated as 1+R/S, where is the number of teeth on the ring, and S is the number of teeth on the sun. These numbers will be proportional to the circumference and hence diameter. If the sun is 6 mm and the ring is 24 mm, you'd get a reduction of 5:1. Work hard and I bet you could squeeze a 6:1 ratio That would put the hub at 150-200 rpm at peak.

Circumference of a bike wheel is just over 2m. At 40 km/hr, that give a hub speed of 300 rpm. So we've got *plenty* of space for a reduction gear.

John Swanson

You say "this brushed DC motor". I say "not that one".
You say "why did you pick that one". I say "pick one that is sound and realistic".
You say "stepper". I say "wut?".

Is there a way to make a drive system of a stepper motor follow the ever varying rotational speed of a cyclocross wheel? Without losing a huge amount of steps?
Mind you, a lost step in a stepper costs you three energy units. Not only do you lose the energy put into that step, the motor will do work in the opposite direction while skipping back that one step.

Of all possible motor drive methods for a racing bicycle, the traditional stepper motor must be the worst when it comes to motion control and efficiency?

Planetary gear:
- if we accept a 6mm axle, the sun gear must be at least 9mm, no?
- it's extremely impractical to lock the ring to the axle and drive the hubshell from the planet
- with a locked planet and the 27mm ring in the shell, the ratio is 1:3
- expect the hub motor to operate a lot in the 500-800 rpm range with a single stage reduction: I call BS engineering

There are a lot of brushed and brushless DC motors in the 22-26mm diameter range that put out 40-60W.
All of them operate at 10000-60000rpm. How could a much less efficient stepper motor put out that power at 10-50 times lower rpm without burning up?

Your engineering-from-the-hip is not sound or realistic IMO. Not yet at least.

 

[Nicko.]

Re: Re:

Risk, reward and detection

Crank motor:
- with a crank motor you can power away from competitors, uphill or to create a gap
- the outcome can be observed by 'bone idle wankers'
- the motor is ~impossible to hide, should one look for it (i.e. remove the crank)

Hub motor:
- the hub motor can at best provide additional energy, leaving the rider fresher for the defining moments of the race
- the effect of the hub motor is completely invisible unless the ryder is separated from the bike
- the motor can be detected by thermal or electromagnetic sensors
- the hub needs to be dismantled to expose the motor to the eye
- batteries may be hidden in the rim and current may be routed through the spokes

Rim motor:
- ~same as hub motor
- harder to detect (thermal) or expose to the eye (e.g. electronics and batteries inside the rim, only magnets on/in frame/fork)
- easy to detect if you want and have the skills

very good

Wasn't sure anyone would pick it up...

 

[Tienus]

- the effect of the hub motor is completely invisible unless the ryder is separated from the bike

There is actually a lot of speculation about hub motors.
Van Aert
Ryder
Izaguirre
https://www.youtube.com/watch?v=GO-NBqya6x0

I saw this one a while ago, it looks a bit odd.
This only has the replay (11:45):
https://youtu.be/iHf1vUM7g0Y?t=11m45s
Full in bad quality (34:41):
https://youtu.be/P3mQ_RGJ2go?t=34m41s

 

[DanielSong39]

You only need ~50-100 watts for ~10-15 minutes (or less) for an attack to be decisive. Surely achieving that kind of output is not that hard.

 

[Teddy Boom]

Re:

You only need ~50-100 watts for ~10-15 minutes (or less) for an attack to be decisive. Surely achieving that kind of output is not that hard.

For the attack proper, 30s effort type thing, these guys are putting out ~1000W (I think Greipel has clocked 1600W on a few sprints, but that will be even less than 30s). The motor would have to be quite powerful to have an influence there. Once there is a gap though, if you are already close to the strongest in the race, I think even 20W for 10-15 minutes could burn out the chasers, depending on the details of the race situation.

 

[Tom the Engine]

Re:

Izaguirre
https://www.youtube.com/watch?v=GO-NBqya6x0

I saw this one a while ago, it looks a bit odd.
This only has the replay (11:45):
https://youtu.be/iHf1vUM7g0Y?t=11m45s
Full in bad quality (34:41):
https://youtu.be/P3mQ_RGJ2go?t=34m41s

You can scratch the two Movistar videos in my opinion.

In the second case (Malori at PR) we don't see the whole incident. We only see him lying there and his rear wheel is comfortably in the air. The wheel doesn't do anything crazy, it just keeps spinning a bit. And the way the bike is lying on the pedal and on his leg it's very possible that the rear wheel never touched the ground after the fall (the front wheel probably did at some point as the handlebars are turnable). So a wheel that keeps spinning under these circumstances is not necessarily something strange in my book.

In the first video (Izaguirre) there are two points to consider in my view: First, his rear wheel immediately goes into the air when he falls down, then just quickly touches the ground but visibly ends up in the air again without touching anything. Secondly, the disk of his rear wheel has a texture on it whose reflections in the sun create the misleading visual impression of fast movement when the camera passes by. But if you watch carefully the moment he gets up and pushes the wheel towards the ground and the guardrail you can see that the wheel is actually rotating only very slowly. So the rear wheel clearly has lost the vast majority of its rotational velocity as is to be expected.

What makes the Hesjedal case that seems similar so different (read: more dodgy) in my view is that Hesjedals rear wheel clearly touches the ground for quite some time. Then the whole bike comes to a standstill (with the rear wheel apparantly still having contact to the ground! But I'd love to see the front view on this.). He clicks out of the pedal and it almost looks as if the wheel accelerates again. And even if it doesn't really accelerate, at least there is a surprisingly large amount of rotational energy left in the rear wheel. Considering that the friction coefficient of the tires on asphalt is high by design I certainly would not have expected this. It's far from proof, could still be a bit of freak circumstances, but on the edge for sure.
Izaguirres case is not even remotely as strange to me.

 

[ScienceIsCool]

Re: Re:

Why in the world would you use the first motor I found? It's merely an example of what's possible and sets volumetric power expectations to within a factor of 50% or so. Truth is that you wouldn't want a BLDC motor at all, you'd want a stepper. High torque and you can build it so that it's operating at peak power at ~1000 rpm. I'm not going to run the numbers, but I bet you could get peak at substantially lower rpm. Lots of power/torque/efficiency curves out there for anyone who wants to look.

Here's a picture of a nylon planetary gear. Notice how thin it is. The reduction ratio is calculated as 1+R/S, where is the number of teeth on the ring, and S is the number of teeth on the sun. These numbers will be proportional to the circumference and hence diameter. If the sun is 6 mm and the ring is 24 mm, you'd get a reduction of 5:1. Work hard and I bet you could squeeze a 6:1 ratio That would put the hub at 150-200 rpm at peak.

Circumference of a bike wheel is just over 2m. At 40 km/hr, that give a hub speed of 300 rpm. So we've got *plenty* of space for a reduction gear.

John Swanson

You say "this brushed DC motor". I say "not that one".
You say "why did you pick that one". I say "pick one that is sound and realistic".
You say "stepper". I say "wut?".

Is there a way to make a drive system of a stepper motor follow the ever varying rotational speed of a cyclocross wheel? Without losing a huge amount of steps?
Mind you, a lost step in a stepper costs you three energy units. Not only do you lose the energy put into that step, the motor will do work in the opposite direction while skipping back that one step.

Of all possible motor drive methods for a racing bicycle, the traditional stepper motor must be the worst when it comes to motion control and efficiency?

Planetary gear:
- if we accept a 6mm axle, the sun gear must be at least 9mm, no?
- it's extremely impractical to lock the ring to the axle and drive the hubshell from the planet
- with a locked planet and the 27mm ring in the shell, the ratio is 1:3
- expect the hub motor to operate a lot in the 500-800 rpm range with a single stage reduction: I call BS engineering

There are a lot of brushed and brushless DC motors in the 22-26mm diameter range that put out 40-60W.
All of them operate at 10000-60000rpm. How could a much less efficient stepper motor put out that power at 10-50 times lower rpm without burning up?

Your engineering-from-the-hip is not sound or realistic IMO. Not yet at least.

First: you've heard of clutch bearing, yes?
Second: Why does the sun gear have to protrude from the axle?
Third: A 3:1 reduction actually fits nicely with the peak power curves we can expect.
Fourth: What would a professional cyclist pay to get even 20 Watts added to their FTP?
Fifth: Skateboard enthusiasts are demonstrably doing this exact thing in their garage with off the shelf parts
Sixth: I've played along, but your hub dimensions are a wildly artificial constraint. Anybody with a brain would pick a relatively popular hub with the largest volume possible.
Seventh: You've allowed zero room for innovation from someone possessing motivation and talent.

Ten minutes, Google, and zero engineering effort says this is wildly plausible.

John Swanson

PS - I never did tap into that volume provided by the freehub.

 

[Random Direction]

OK, any mechanical engineering cyclists with a machinist habit want to build us the motors that nicko and John are blowing my mind with? Shouldn't be too hard for a weekend project.

 

[Tom the Engine]

Just wanted to add a couple of thoughts to this hub motor discussion that are somewhat complementary to the technical details.

1) I don't have much doubt that it is possible and that people have worked on it and probably also succeeded to some extent. That said, possible doesn't mean trivial, and a lot of stuff would be possible but doesn't exist regardless. The sheer possibility does not automatically imply reality.

2) It's one thing to develop such a system that works as intended and is even kinda invisible enough at first sight for it to be interesting for an after-work hero. It's a different thing to create a system to cheat in pro races literally in front of hundreds of thousands of eyes - and many of those eyes looking for motors.
A working system is not enough. You certainly have to go on and considerably improve it in terms of design and specification (like weight, reliability and, above all, stealthiness!) before it can fly under the radar in the pro peloton. So who might be willing to put in this extra effort for the sole purpose of making it next to foolproof for cheating on the big stage?

3) I highly doubt that pro teams are capable to develop such motors internally. (And although I'm generally opposed to drawing conclusions for motors just from comparisons to doping I'll add that I'm not aware that the pro-cycling world ever considerably invested in research or experimentation related to doping. They just used what was available to them.)
So if there are motors in the peloton then they're probably from third party suppliers.

4) There are weekend-engineers all over the planet for sure who try to build it in their garage. But these guys are probably happy when they end up with a working system. And don't invest more time and money for the next steps that are very tedious and not much fun to work on. Because fun is their main incentive I guess. Fun and getting recognition for their achievements, at least from their peer group. Like these guys who put motors in their skate boards and then share it in forums to receive their well-deserved applause.
But I guess we can scratch that here. You don't get any recognition to begin with when you can't talk about it because it has to be top-secret. In addition when you're focusing on making a system most suitable for explicit cheating you're crossing a line that probably leaves only one main incentive to do it: Money.

5) So people developing such motors for the pro peloton don't want a pat on the shoulder or have their expenses covered. They certainly want to get paid, probably big time. Now I start wondering about business models. I'd speculate that some sort of exclusivity agreement with riders or a team from the pro peloton is not the best they could do - by far. There is a lot of money involved in the sport but how much of it really is at the teams or riders disposal?
I guess the real and sustainable money for a motor supplier is with the very wealthy recreational riders. Yes I could imagine a quite secretive business model with only a few selected customers (pros and wealthy people) that for example works on invite-only or recommendation-only to keep the prices as high as possible. But I could also imagine that trying to reach a larger customer base might be preferrable nonetheless in the long run.

6) So unless a cheating pro has managed to tie the fate of his motor supplier very closely to his own fate ("you go down if I go down"), what would the motivation for the supplier be to keep his mouth shut? And I'm not talking about revealing his customers in the peloton - there's no point in doing that - but about keeping his product such a well guarded secret. I see none! It's not illegal to build and sell such systems. Even if there's a legal case on sporting fraud in France or Italy because a rider was caught in a race, the supplier will not be charged with legal responsibilities if he was just minimally cautious about it.

7) Doubling down on objectives. John Swanson stated that Varjas is stupid because he talks about it. Why? It's the exact opposite. Don't forget: His main objective is not to successfully cheat in pro races. It's about making as much money as possible by selling his products. What he does makes perfectly sense to me. There is literally no such thing as bad publicity that harms his business (except people claiming his products are crap or he cannot actually build them).
When a pro rider gets caught that (ironically) is good news for anybody who wants to sell motors to the public. And when it's a successful rider that's even better. Think about it for a second!
(And even if many people were disappointed and turned their backs on cycling because of it - doing damage to the cycling business as a whole - a motor supplier will still have a huge relative increase in orders.)

8) To wrap up the economic argument: The claim that there is a very sophisticated system around in the peloton that nobody in the public has seen so far implies one of the following three scenarios: (a) It's an inside job by a team or (the environment of) riders. (b) A third party supplier got paid a sh*tload of money for an exclusivity agreement and the cheaters found a way to enforce it in such a way that he loses this money if anything about it leaks out (like: a cheating team taking the supplier to court about it.. Would love to see this.). (c) A third party supplier sells to both pros and regular people, but he keeps his public customer base extremely exclusive on purpose (seeking no publicity whatsoever) and somehow manages that no one of his customers talks about it either - for whatever reason.
All of these scenarios are certainly possible. But you have to keep the objectives and incentives of all involved players in mind. You can't just automatically assume that the main goal of everybody is to cheat in pro races. It's not. Everyone optimizes according to the metric that is most relevant for oneself.

==================

My bottom line is: It's rather pointless to post links to blurry videos that allegedly demonstrate a motor but don't really show anything. Or patent applications from the 90s that may or may not work. Or motorized skateboards that can be heard before they're seen. Or some Chinese crap that probably will start burning after 200 meters of use.

There are really only two things that are interesting: A reference to an existing system that could convincingly be used in the pro peloton. Or evidence of any kind that a rider has a motor in his bike during a race.

 

[Nicko.]

Re: Re:

First: you've heard of clutch bearing, yes? Yes of course. But they are not infinitely thin. I'd expect you'd need 1mm extra radial space for it

Second: Why does the sun gear have to protrude from the axle? The sun gear rotates around the fixed axle, the cannot occupie the same space...

Third: A 3:1 reduction actually fits nicely with the peak power curves we can expect. I fail to expect. What motor design gives these power curves?

Fourth: What would a professional cyclist pay to get even 20 Watts added to their FTP? That's 'why', not 'how'. Nice try...

Fifth: Skateboard enthusiasts are demonstrably doing this exact thing in their garage with off the shelf parts They don't need a gearbox and they have 4x the volume available.

Sixth: I've played along, but your hub dimensions are a wildly artificial constraint. Anybody with a brain would pick a relatively popular hub with the largest volume possible.I picked the discussed, rear-wheel-swapping PR-winning bike, saw a Shimano C50 wheelset, got to work. Was that incorrect? Which controversial bike hub should we pick?

Seventh: You've allowed zero room for innovation from someone possessing motivation and talent. On the contrary. I really tried to stretch the mechanical restraints and asked for sound, realistic ideas w/o picking on details

Ten minutes, Google, and zero engineering effort says this is wildly plausible. I havn't seen it yet.

John Swanson

PS - I never did tap into that volume provided by the freehub. Your clutch bearing needs half of it. You can put your electronics in the other half.

Reference for item 6:

You see the hub shell of that C50? It's a little bit thicker than your thumb...

Look, I get your genereal point. I believe there was a wheel motor in the case of Ryder. Don't now anything about that hub though.
I can grasp the extremely difficult constraints of hub motors in a C50 hub or a CX race. You have not shown a sound and realistic solution for those scenarios.

 

[Nicko.]

Re:

You only need ~50-100 watts for ~10-15 minutes (or less) for an attack to be decisive. Surely achieving that kind of output is not that hard.

Show your work.
You can stand on the shoulders of my work...

 

[sniper]

@nicko , that's 95% sure not the rear wheel he rode the finale of PR with.
Unless he decided to clean "Daddy's dirty bike [which] should not be cleaned".
Zoom in on it and spot the differences between front and rear wheel.

Which doesn't mean the hub shell of his PR winning bike was bigger, just that it may have been.

 

[Tom the Engine]

Re:

that's 95% sure not the rear wheel he rode the finale of PR with.

Honestly, I struggle a lot to grasp how you can be 95% (or even close to that number) confident in this.

Zoom in on it and spot the differences between front and rear wheel.

You're completely missing the point. Differences between front and rear wheel are irrelevant.
What you should compare instead is the left side of his rear wheel in the finale with this picture of his bike after the race you're referring to. That's the only relevant comparison if you want to find evidence that the wheel was changed (or cleaned) after the race.

I didn't do an exhaustive search but so far I couldn't find a single picture that shows that the left side of his rear wheel was ever dirty. Can you? I'd be happy to look at it.
I can only find pictures of the finale where the left side of the bike looks very similar to what we see on the image that makes you 95% confident..

For example:
http://www.ridemedia.com.au/wp-content/uploads/2016/04/HAYMAN-Mathew801p-YS1.jpg

Or:

 

[sniper]

Best marks for missing the point yourself there.
But let's not go there again, as all the things you ask for have been posted already, and it's not crucial to the current discussion.

Bottomline, and this is *the point*: the rear hub with which he won PR may well have been larger than on the pic Nicko posted.

 

[Tom the Engine]

Re:

as all the things you ask for have been posted already

Can you please refresh my memory on this? All I can really find is you repeating your line with cleaning daddies bike over and over again.

and it's not crucial to the current discussion.

Are you kidding me?
The whole point of your argument about the possibly larger rear hub is based on your assumption that the wheel was changed after the race (with stated 95% confidence). And as far as I can tell this assumption is solely based on your "observation" that the rear wheel looks clean on that picture. But the two pictures I posted don't support your hypothesis at all - quite the contrary.

 

[sniper]

It *may* have been changed. 95% or 50% or even 25% sure doesn't matter.
It has to be taken into account.

 

[sniper]

Here you go:
http://www.cyclingnews.com/news/meet-the-aero-bike-that-won-paris-roubaix-mathew-haymans-scott-foil-gallery/
Unfortunately there isn't one photo in that series that gives us a clear shot of his rear hub.

 

[Mr.38%]

Blow it up to 500% (or better use PS). It's a 9000 hub, low flanges, rather tiny shell.

This pic is used in Scott's add campain:

Tiny shell, right after finish:


Tiny 9000 shell:

200% is enough to spot a regular hub: http://www.ridemedia.com.au/wp-content/uploads/2016/04/HAYMAN-Mathew801p-YS1.jpg


I cannot spot anything else than DA 9000 hubs in C50s on any pic I found on the webz. These are just a few that stand out because they show the hub and most were taken right on the Roubaix track.

The alleged CK hub on Froomes wheels, given they exist, would be interesting, though.

 

[Tom the Engine]

Re:

Here you go:
http://www.cyclingnews.com/news/meet-the-aero-bike-that-won-paris-roubaix-mathew-haymans-scott-foil-gallery/

I'm sure you know the difference between left and right yourself.
I repeat: Show me a picture of the left side of his bike that indicates that it's very unlikely to be the same wheel. Otherwise you can trash your wheel-change hypothesis right now (unless you have other arguments for it except this allegedly missing dirt).

95% or 50% or even 25% sure doesn't matter.

Adorable.

It has to be taken into account.

So why not go a step further and "take into account" that it could be a different bike altogether?
No, that's really not how it works. We cannot start taking random things into account for no good reason. Evidence comes first and then stuff that could follow from the evidence is taken into account. When the evidence collapses the conclusions that have been drawn from it have to be revoked, too. Even if one badly wished they were true.