▲ 231 points
back
219 comments
I'll tell ya what I see: no rollers. Pins rubbing (wearing) both sides of the tooth. Including the back side, which will never exert any forward longitudinal force on a chain unless it's made of magnets. I also see a highly suspect FEA diagram on the right, where teeth not even engaged still show non-zero stresses (lighter blue, not the darkest blue), indicating that the color/stress scale has been changed compared to the baseline on the left. So it's not an honest comparison. Add to that the fact that this is a 'better mousetrap' that requires you to first replace all the mice in the house, and I guess I'm a bit (to use one of Bike Snob NYC's favorite words) nonplussed.
I suspect lack of need for rollers is a primary feature.
Conceptually, to me, the new interior bits serve as a sort of individual tooth holder that is free to rotate with the tooth as the tooth orbits the axle. So there's not much friction at the tooth:holder junction as the sprocket turns.
That friction is now occurring at the narrower intersection of a hinged tooth holder and chain link. Being a narrower hinge joint, just the length of a chain link's thickness, it has less surface area for friction, and less room for debris ingress etc.
Conventional chains rotate a pin across the sprocket's thickness, and AIUI use a roller to try keep debris out of and lubricant within that interface to reduce friction. This contraption takes it a step further.
You can actually envision the tooth holder as analogous to the roller in a conventional chain, just with less surface area at the twisting intersection. They've obsoleted the conventional roller.
I don't know anything about FEA, but in reading this article and thinking it through, was the first time it hit me how bad a conventional chain and sprocket is. On a brand-new set with perfect tolerances, you might be able to use many of the front side of many of the teeth, but as SOON as the chain begins to stretch (wear) even fractionally, most of the stress goes to the 'first' teeth in any given power stroke, and the same spot on the chain. I imagine if you're putting, say, 1000w through 2 teeth and 2-4 chain endplates, wear is an order of magnitude higher than if you're able to effectively transfer power through double the teeth and endplates.
Likely the "pins rubbing" you mention doesn't really matter, because the wear from inserting and removing the chain as it cycles around has to be a miniscule fraction of the wear that occurs under load.
If this chain is truly more efficient and/or durable, I would expect them to be selling to two specific markets:
- BMX: This market would be all over anything that would be an improvement in durability, and frankly a percentage would just do it for the different tooth design.
- Auto: Admittedly, the site for the company that is making this new chain mentions automotive in a blog post, but I don't see anything specific to an automotive product. A more durable and efficient timing chain is something I'd expect most of the automakers to care about.
I'm afraid this is kind of a scam too.
The solution clearly loses on weight/benefit ratio to just making a thicker chain.
> I guess I'm a bit (to use one of Bike Snob NYC's favorite words) nonplussed.
JK Rowling taught me that nonplussed is apparently a part of the daily vernacular in the UK.
I really don't get it: what power are they claiming to be transferring on the side of the tooth that isn't engaging with a roller chain?
Also note that this is only for track bikes with a perfectly straight chainline (technically speaking it would also work for internal gear hubs wish are increasingly driven by belts in the upper price range). The shifting chain that is deployed diagonally more often than straight is an entirely different beast, in theory much worse in efficiency and durability. Pretty much all practical durability limits stem from compromises done for that diagonal capability for sitting. Which basically goes ruthlessly against all mechanical engineering rules, like the old adage about how bumblebees were supposedly unable to fly. Yet even they work quite well, given that challenging background they work just awesome and track chains already work better than that.
Internal hubs are amazing and so great to use, until you get a puncture. It’s such a nightmare and I went through so many of those dumb proprietary nuts that strip so easily.
Bikes like the Cannondale Bad Boy have got me thinking, what if the single fork was at the rear? Then you could have an internal hub and change the tube without the extra suffering. Yes, the loading at the read would be significantly greater, but if it were possible it would make internal hubs much better.
Just had an idea [1] about how this "gripper chain" could maybe outdo a roller chain in efficiency:
Maybe the "gripper plates" align themselves perfectly with the tooth at a time when the connection is still not under load yet? A conventional roller won't need much aligning, but it won't do that, it won't settle into place before load is applied. Friction scales with force. If that's how they teased out a minor delta then it's truly very clever (but likely still highly impractical outside the laboratory/indoor track setting)
[1] idea came while taking a shower, were else. Shower after churning that old Campag 12s chain a little, unfortunately only on Zwift these days
I don't think they are claiming any power is transferred by the other side, just that it forces a more uniform fit and distributed the power transfer over more teeth.
Elementary: one roller pushes the tooth, the other one pulls.
Great chain, but the Gates Carbon Drive is likely a superior technology. Europe has largely adopted it, but it is rare in the states. I have a Riese & Müller bike with a Gates Carbon Drive and it is absolutely lovely. Silent, but best of all, no grease on my pants or hands, ever. They handle dirt and dust super well (just dry off and wipe down with a cloth/no gears get jammed with a sealed planetary gear system). It also is said to last thousands of miles, if not 1e4 miles+.
I live in the Netherlands and have traveled Europe very extensively and I have never seen one single belt driven bike... I have a feeling that it'll be a long time before Europe widely adopts them :)
Does look very cool, though.
If you want to shift gears using the carbon drive you have to use an internal gear hub, which has a lot less range than a traditional derailleur and is less efficient. So if you are riding a racing or mountain bike a chain system is probably a better way to go.
Also it looks like you can't shift under load with the internal hubs which is a problem if you are trying to climb hills quickly.
Gates only quote between 95-98% efficiency. You can relatively easily get to 98% with conventional chains using wax and oversizing chainrings and sprockets and you don't need to worry about splitting your frame to replace them...
This is a very niche product but sounds like an interesting approach. The users of it likely won't care if it needs replacing every ride if it means saving watts and winning (olympic) races.
> Europe has largely adopted it
I'm not sure you're using the term "largely adopted" correctly. It implies that something is the majority or norm, not a rare occurrence.
I've got a belt-drive "commuter bike" that doesn't get ridden much these days. It currently has a Gates Carbon Drive which I mostly like, and you're mostly correct. But it came with a Continental belt drive system that was awful. I snapped my first belt on it under load when it was about 6 months old. The damn thing always squeaked no matter how clean i tried to keep it. The rear wheel had to be on perfectly straight or sometimes the belt would slip one way or the other -- and of course, belt tension is very hard to get right. I think the gates system operates at lower tension than my continental did, which is a plus.
I hope the Gates belt is stronger than the Continental one, but I don't know for sure.
I think Ghost ultimately was willing to goodwill replace the continental trash with a complete Gates drive system because a continental belt could not be sourced anymore. But, I still have to worry about the nightmare of axle nuts, belt tension, removing the tiny nut from the hub to disconnect the shifting, and removing the brake caliper (!!!) to fix a flat.
I've destroyed a number of Sturmey Archer planetary drive clutches - the square cross-shaped bit which transfers force from the sprocket holder to the drive - by just using them, leaving me wary of planetary drives. While I've also gone through derailleurs they do seem to handle power better than planetary drives and have the additional advantage of feeling far more like a direct drive. One of the advantages of a bicycle is its simplicity, there is not much between your feet and whatever goes for road you happen to be on. I'll gladly give up 2% of efficiency to keep the system simple, repairable with off-the-shelf parts which can be found anywhere in the world and affordable.
How do you handle gears with the belt? Most of the bike I've seen in the US that use the system are single-speeds. I suppose you could use a internally-geared hub, but a Rohloff is mega-expensive and the Shimano offerings leave a lot to be desired.
The advantage here is that the chain could be made backwards-compatible with existing conventional drivetrains. The Gates drive can't be retrofitted onto older frames.
I have seen maybe one driveshaft based bike, and I’m European, and that was at the school of mechanical engineering.
Any idea of the sustainability of the Gates Carbon Drive system? Looks like it's made of polymer + carbon fiber. On the other hand, traditional chains eat+shed oil and grease.
I disagree both about the adoption rate and the superiority of the tech. It only works if the frame can be opened up and this means the whole bike has to be designed around the belt drive system. It's a slow thing to do and quite expensive, whereas replacing a regular chain takes less than five minutes (it's a bit more messy though).
I have to be honest, when IKEA launched the Sladda, I had high hope that they'd sell a zillion of them so I could pick one up barely used for a pittance in a couple years. Sadly, that seems to have not happened, and we'll have to wait a bit longer for the belt drive to become commonplace :-(
I have a Gates Carbon drive too. It's extremely convenient, but my understanding is that it's lower efficiency compared to a well-mantained chain. Due to the practical challenges of maintaining a steel chain, I think Gates does well and it's a great commuter option.
Every so many years I see a post like this where they claim to have improved the drive chain of bicycles and they really never go anywhere, mainly because they aren't compatible with other kit.
I've built and repaired quite a few push bikes over the years I am still using fairly standard shimano kit. The problem with exotic equipment is that even decent shops might not be able to repair it and it normally requires specialised tools.
I mucked about with Campag and some other exotic kit and it is always a PITA as you always have to go hunting around for parts of ebay after a few years because threads and tapers on things like bottome brackets aren't compatible.
This makes me think about removing every other sprocket tooth and then milling down the attack angle on the forward side so that the two rollers would both seat on either side of one tooth. I think the point of this design is that the link transfers compressive force forward as pressure if both sides are engaged.
The loss in a chain is nearly all in the roller as it engages with the first tooth in the front chain ring. Most of the force is being transferred from this first tooth, and it presses directly on the roller, which is also rolling into its seat position on the tooth. So a better design would have the roller get seated on the teeth before the force gets transferred to it, and then have no rolling until force transfer is done, then roll off gently.
A new chain on a new chainring engages at the bottom of the tooth, right where it seats, and the torque of the chain ring rolls it up the forward tooth ramp until the tangential component of the force pressing into the contact point equals the pedal force. This happens on all points engaged on the sprocket. So you get a lot of rolling, but that’s over a lot of rollers. As the chain wears, it gets longer, so the teeth catch early from the tip, and it rolls all the way down the tooth while the entire chain tension is all on that first roller.
This particular design has no rollers at all. It will still stretch because it has the same pin-plate interface, and it seems that a modest stretch may render it inoperable, but I’m not sure.
This chain is a nice idea but a solution in search of a problem. Right now it is track bike and e-bike only, so not even for hub gears.
The track market is miniscule and we'll catered for. Then the e-bike market does not need a 2 percent gain in chain performance.
The enemy of the chain is dirt. Real world cycling in the 1950s had hub gears and a big case to keep the dirt out. Then derailleur gears aimed at the sports market dropped the chain case. This made sense for cyclists with a team car behind them.
Reduced chain width has significantly changed bicycle drive trains over recent decades. Bicycles have many more usable gears. A manufacturer who pushed a non-standard chain heavily would wound cycle shops. Standardization facilities keeping parts in stock.
Whenever i see things like this, i immediatly think "Audiophile". you know, the $238 "Directionally balanced" Ethernet cables, or the "low oxygen" copper.
Once people take their hobby too far, you will end up with things like this.
For $2K you can buy our (exotic and unnecessarily expensive metal) chain, which reduces friction to 0 and lasts forever.
I ride a fair amount (150-200KM a week in the summers) and am always surprized to see everything on a bike is measured in grams, except the biggest impact, the rider itself?
Sure i shaved 4g on my $400 waterbottle holder.... but when i weight 80KG.. and have another 2kg of water with me, does that 4g really matter?
Side note:
I used to work in a bike shop for about ten years. One time had a guy come in who worked with Kyronics. We all made jokes about freezing dead people - but he said you can apply Kryonics to a wide array of stuff not just dead people.
One of the many things some of the team riders were seriously considering was having their cranks, chains and frames frozen since its supposed to alter the molecular structure to make them stiffer and more durable.
I'm not sure they did it, but it was a topic of constant conversation for weeks afterwards.
Unless I'm mistaken, the efficiency gain is from shortening the length of the pivoting pins to just what's required for connecting the conventional outside chain links to the saddles.
In a conventional chain the pins cross the sprocket teeth and must rotate across that entire width; more frictional loss.
In this design the saddle seats on the tooth, and just the relatively short pins linking the saddle to the outside chain must pivot.
> They’re soon going to be used in top-tier UCI track races.
And that is where this monstrosity will stay, until they find a way to fit it into a 6 speed derailleur system to make it viable for $60 bikes sold in Walmart, and eventually an 11 speed derailleur system.
Chains stretch, thus any sprocket system will inevitably end up concentrating the load on the last link or two of contact.
There are systems that actually could produce less friction, however they can't handle ANY contamination, thus ruling them out.
Your best bet is to look elsewhere. I would try reducing the contact friction as the links engage. To do this, I would recommend using BAM coatings on the chain and sprockets. It has lower friction than Teflon, AND is almost as hard as diamond.
>"This isn’t just some lab experiment, either. They’re soon going to be used in top-tier UCI track races."
As long as it is not an absolute crap pros would use whatever equipment their sponsor tells them to use. Chain durability for pros does not matter and whatever the difference in friction are really negligible.
>"ability to transfer 25% more Torque"
Any decent chain will handle way more torque than cyclist can deliver so what's the point.
>“four times increase in system lifetime”
I believe it when I see it. Also how much more this puppy gonna cost and cyclists would need to change chainrings and sprockets. This much troubles for the sake of chain? Sorry but I have my doubts.
How well can these handle dirt and debris? It looks like a high precision connection which makes me slightly worried that an errant grain of sand or just an accumulation of road grime might jam up the works.
Gears! - But Were Afraid To Ask (MiniLathe) - This Old Tony
Reading through there is something very noticeably absent: any claims of increased efficiency.
If you spread pressure out offer wider areas, you don’t get less overall friction at least in the first order approximation of friction you find in first year physics.
What it looks like is an attempt to make something better that improves metrics which aren’t necessarily coupled to better overall performance.
How does it compare with the rohloff system?
Anyways it seems to be significantly more expensive thus I don't think it will become widespread.
Tolerance high need. Stretch or fracture.. no longer optimal. Current chains.. been working for over a hundred and fifty years and we all know how to fix em and oil em. So for the 2% improvement hero, sure. For me? Less sure. (I ride a clunker i put a Brookes leather saddle on, now it fits my bum cheeks I'm good)
it does look more durable with all the the added metal but how could that make it faster? but either way, I never had durability issues with my bike chains (including motorcycles), specially when you can adjust for "stretching"
As a guy who ignored the chain on his first bike with XTR gears and had to replace whole drivetrain - yep, I would buy this chain. No question asked, as long as they guarantee me that it wont be stretched.
Chain wear is caused by bad lubrication. Both lack of lube and grit in the lube. You can use a bike chain with 25 horsepower if you put it in a sealed case half filled with oil.
There is an easy and straightforward way to make this chain successful. Put in on a team bike that wins. 2% is quite a bit in a peloton. 2% of 2% is Biopace.
Trucks use a solid driveshaft to transmit power to the rear axle. Is there any merit to using that idea for bikes? (with a geared transmission?)
How is the shifting? Sounds a bit too rigid. Smooth shifting is probably more important for performance than those last promilles.
Maan I thought hn had a lot of 1M/year salary software developers, but just look at the amount of bike engineers.
wish belt drives would catch on more.
(yes, I realize the frame needs to be split to remove the wheel, but in the newest designs this doesn't seem like too big a deal)