2_i 

For power transmission on a bike the chain has been difficult to beat. While generally beating other transmission methods, the transmission efficiency and chain and cog durability have been limited by the fact that relatively few teeth get engaged in the power transmission in any one time. The New Motion Labs put forward a new chain design that uses two pins per tooth and grabs the teeth more effectively and engages simultaneously a larger number of teeth in the power transmission. This is called Dual Engagement and should increase efficiency and lifetime of the components. Unfortunately in its current form the new chain design will not work with derailleurs. However, apparently a chain version specifically aimed at the derailleur systems is in the works.

The shapes of teeth in the cogs and rings would need to be modified for the new type of chain. However, since the cogs need to be frequently changed anyway and the rings occasionally too, it looks like current bikes might be adapted with limited bother. It now remains to be seen whether they get the chain mesh with derailleurs. Next would be getting the new drivetrain to sport competition and demonstrating its advantages in practice.

To me it looks like a real opening.

woodcraft

I don't quite see it- the load is still on the front of the tooth,

but maybe it will be the next disc brake or Di2. It would certainly up the ante for quick links!

dabac

I can’t see how that could work.
There is only one side that sees drive pressure.
I suppose it could kinda-sorta ”work” if you use a tooth profile that jams itself in between the rollers, preloading the interface between the roller and the rear flank. That way, while applying power, the rear flank would lose some preload. That loss of preload might be argued to add to the drive in the same manner some claim that a wheel stands on the spokes.
If that is what they’re going for, I struggle to see how that preload feature could make up for the friction losses of the jammed in teeth engaging and disengaging.
Also, keep in mind that a chain drive in good condition already is something like 98+% efficient. There really isn’t much room for improvement anyhow.

Kovkov

This sounds like something that was tried out already somewhere in 1902
and dismissed.

Skulking

I think the "Dual Engagement" term is a bit misleading. The brief overview seems to give the impression that it somehow transfers force to the back half of the tooth, but looking at the more in depth stuff, that idea disappears and it is all about how the dual engagement chain locks in on both sides of the tooth to eliminate energy loss and friction caused by the movement of the chain on the sprocket as it rotates and the contact point changes. It also claims that the design does a better job of spreading out the load across more teeth on the sprocket.

It's an interesting idea, but they need to develop a version that can be shifted before it can move beyond being feasible for anything except single speed bikes. The current chain is clearly going to be more expensive to manufacture as it is more complex, and they will need to figure out a way to make the technology work without making the chain so wide, or you are going to limit the number of cogs you can run due to the extra spacing requirements between cogs. The most difficult thing for them to do may actually be to convince any of the major component manufacturers to license and produce chains using the technology as they are more of an R&D type company rather than a manufacturer, and it is likely that the drive train would need some sort of reworking to be compatible with this chain.

2_i 

On a tooth by tooth basis, I think that the pin must be loose inside a roller that results in the pin exerting a localized pressure on the ring. In the new design the seat for each tooth may sit tighter in the chain, because its shape matches that of the tooth. Otherwise I suspect that the tolerances in the newly designed chain can be tighter, making it seat tighter in the ring and making the teeth work more in sync. Still it is a compromise between the fit and friction.

On my main bike I need to replace the chain once a year. On a folder, when I replace the chain, I need to replace the cogs too. I find it well too much maintenance. The chain putting pressure on just few teeth just makes no sense. There must be a better solution.

Kovkov

The solution is 2x5 drivetrains maximum, thus chains that bend less and steel cogs.

Andrew R Stewart 

I don't see this as any big new design. As others have said the drive forces are still seen by one face/side of the cog's teeth. Chains with more then one pair of link plates have been around for decades, So to for cogs with more then one tooth per chain pivot (I think of auto valve timing chains I've seen). Shimano patented a single sided chain years ago. Chains with a single set of plates/links with rollers/pins on both sides engaging a cog with two sets of teeth (having a gap between that the chain plates ride in) have been around a long time too.

I guess this is like the attempts to reinvent the wheel. A worthy goal but harder to actually improve on then your investors expect. Andy

2_i 

When I look around, I find designs that are multiplication of the standard, essentially few standard chains fused together, nothing along the way of New Motion Labs. Their potential first round application is industrial, to conveyor belts.

dsaul

Looks like a skip tooth chain without the rollers. They replaced the friction of the roller against the pin with the pin rubbing directly on the tooth. It seems like the tooth will wear in short order and negate the "dual engagement" aspect, if there really was any positive benefit from that in the first place.

unterhausen

When I saw it, it definitely looks familiar. Cycling used to feature a real obsession with efficiency. I wouldn't be surprised if someone invented this before.
The engagement doesn't look too good in their video

70sSanO

The real issue is the cog spacing. People who are pleased with 1x11 or 1x12 wouldn’t give them up for a $100 chain and half the gearing range.

As I glanced through the site it noted e-bike and belt drives. For internal gearing, or single speed, especially with electric assist, there is probably a market. Applying 750 watts to a drivetrain could benefit from it.

John

caloso

From the photo, that chain looks like it weighs a ton and has minimal lateral flexibility. Track sprinters and BMXers, possibly.

79pmooney

Probably, But in 1902, they might not have had some aspect of technology needed to pull it off successfully. (Remember that in 1902 the brightest people were inventing bike stuff, not computers, software, aerospace, etc. We haven't gotten any smarter, Just turned our attentions elsewhere. And technology (drive primarily by wars) has come a long ways.

Total aside, but along that train of thought. When I conceived of the design for the fix gear in my logo, I drew up an "L" shaped dropout to be a very easy dropout to remove a fix gear wheel from, even after screwing on a very big cog that brought the wheel all the way forward. (I love it! Works great!) I took it to the local fix gear guru. He took a look and said it had been done, perhaps in the '30s. 3 years ago I saw a track bike built by a Bay area with a forged dropout he had made. Last week learned that a forum member has an old mixte with that style dropout (two miles from my house and perhaps the dropout I was told about.)

And to come back to now. There is one minor aspect of modern chains that is that little elephant in the room. Reliability. The most critical part of the drivechain is nowhere near as reliable as it was just 40 years ago. Don't believe me? Read the race reports. Yes, we are addicted to having in a million cogs and accepting that a mechanical reliability that would be a nightmare for a mechanic in charge of (say) industrial machinery is OK for us. I know this first hand, I have a 9-speed bike, 7-speed bike and 3 1/8" fix gear bikes. The old roller design 1/8" chains running on aligned cogs operate like a good reliable machine (except for the exposure to dust but that was accounted for in the chain size 120 years ago by those bright engineers). I can see this new chain as running like a dream on my fix gears, Too bad I am invested in so many expensive cogs!

Marcus_Ti

Or single speed/fixie or IGH. I agree, that chain doesn't look at all flexible enough laterally to use in a modern derailleur drivetrain.

Which...if IGH development picked up steam and marketshare wouldn't be that big a thing. Of course..SS chain is already insanely efficient as is.

luns

I see no merit to this design, whichever way I look at it. I just see a bunch of smoke and mirrors.

First, if there are truly two pins making contact per tooth, only one of them is transmitting force in the direction you want. The other one is ultimately working against you, and any force that it's applying means more force for the other pin. That's not a load sharing arrangement, it's an advesarial one. If you have a friend help you push a car, but they're pushing the other end of the car from you, that's not helping! If you just get rid of the pin that's working against you and join up the remaining pins, what you're left with is the conventional chain they're trying to distinguish themselves from!

How much load is taken by the first tooth of a conventional chain/sprocket isn't inherently defined to be any particular value, it's a design parameter that's set by the choice of the tooth angle. It's as high as it is because that's the best practical value for it to be. You could design it so that 1% of the work being done by the chain is transmitted to the first tooth if you wanted. 1% of the remaining chain tension is borne by the next, and 1% of what's left is borne by the next, and so on. However, after going halfway around the sprocket, you'd still have a lot of tension left - all those 1%'s don't add up anywhere close to 100%. What this means is that the "slack" side of the chain needs to have a significant tension on it in order to keep the chain from skipping. Conversely, for a given 'slack' side tension, there is a limit to how much working tension you can have before the chain skips, and that maximum depends directly on how much work each tooth does relative to how much it passes on.

Roughly speaking, for engaging the 6 teeth halfway around a 12 tooth sprocket, and you have each tooth do 60% (I think that's the number they gave for a conventional arrangement) of the work presented to it by the chain, the taut side can support up to 1/0.4^6= 244x of the slack side's tension. If we try to reduce the 60% down to 40%, the maximum tension goes down to 1/0.6^6=21x the slack tension. This is an 11-fold reduction in capacity for a modest 1/3 reduction in leading tooth work! If we go down to 10% work taken by the leading tooth, the capacity is only 1/0.9^6=1.88x or yet another 11-fold reduction!

So where between the 60% and 10% are these guys shooting for? Whatever the value is, you can make a sprocket with tooth shape of appropriate pressure angle to reproduce that with a normal chain. The sprocket at its contact points will be practically identical both ways, the only real difference being in the chain. How is the more complicated chain design any better conventional in this arrangement?

In fact, whatever they shoot for is a value that would be seen by a normal sprocket/chain if you just let the chain wear enough. As the chain wears, it rides higher on sprocket teeth. As it goes up around the curve of the tooth, the pressure angle gets shallower, and the work transmitted by the leading tooth reduces. As that happens, the load capacity of the chain also reduces. Keep going long enough and once you pedal harder than the reduced capacity, your chain skips!

Roller chains have been around a long time and are well understood. I get the sense these people are too blinded by the fancy CAD and simulation tools available nowadays to understand the underlying fundamentals. I wonder whether these people are even aware of the book, Mechanics of the roller chain drive: Based on mathematical studies by R.C. Binder, or if they even understand the concept of GPLD (Geometric Progression Load Distribution) used in every paper I've seen on chain mechanics.

They're good at making pretty pictures, but that's about it as far as I can tell.

sweeks

I am puzzled by this. As a folding bike owner for about 18 years, I've been going through 3 or 4 chains before needing rear sprockets. If you have to replace the cogs with every chain, I wonder if you are keeping track of chain wear. Replacing the chain when it has elongated between 0.75% and 1% should greatly extend the life of the cogs and chainrings.
Also, what does the bike being a folder have to do with chain wear? I am genuinely curious.

2_i 

Well, here I am puzzled. Folders have small wheels and for the same distance must execute more turns meaning that for the same distance the cogs must, on the average, execute more turns. You can put a gear hub, but then you put fewer cogs of which each is utilized more frequently. Otherwise you ride a lot on a big ring in the front and small cogs in the rear. On a full size bike you may change the cassette every second time you change the chain. On a folder this amounts to changing the cogs every time. The Brompton owner manual here says do it every time. Telling that the cogs are worn is a bit tricky, but I actually go through the process of checking rather thoroughly. I take the folder to places where my life can literally depend on it functioning and I cannot afford it failing due to my sloppiness.

tyrion

Twice the engagement, twice the friction. It's simple math.

Andrew R Stewart 

If chain wear and it's resulting failure paths are dangerous then by all means replace that chain far more often the most here do!. I have found it is the periodic maintenance that creates a bike that is dependable and not how expensive or cheap the bike is.

This has little to do with chain dynamics or other engineering principles. It has everything to do with confidence in your bike and it's abilities. Andy

Andrew R Stewart 

As I hit "post" on that last one I thought of one of my earlier moments of clarity of planning and expectation. When I was learning to make a frame the shop owner had a short term GF. She was also making her own frame, one that she planned to ride around the world on (this in 1979). But it was her component choices that did the deed to me. A cotter pinned crankset! When we were drooling for Campy NR or maybe a Stronglight 99 she was thinking what could she fix in a rural African village. My eyes opened to there being more worlds/cultures then the limited one I had experienced at 23 years old.

2_i mentions going where a bike failure might be dangerous. I suggest he/she seeks really old and proven designs and not new and poorly distributed ones. Andy

2_i 

This is the line of my thinking too. The stems of the teeth are more box-like and the pins with rollers rock less between the teeth.

luns

It can't possibly work that way. As the chain wears and its pitch increases, the pitch circle diameter increases and the centre between pin pairs will be higher up the tooth. When the pin pair is higher, the load will bear on the one pin doing the work, and the other one will lose contact with the tooth.

There is a 1-1 correspondence of the engagement between the dual pin contacting a tooth, and a conventional roller fitting into the cut between teeth. The dual pins contact both sides of the tooth only where the conventional roller is resting on the bottom portion of the cut between teeth and not against the sides of the teeth. However, work is not done between chain and sprocket here, this is where the chain is behaving as if it's wrapped around a toothless drum. For any engagement outside of this base circle, the chain only engages one side of the tooth, the same as a conventional chain.

Even worse, I expect there's much more sliding motion between the pin pairs on a tooth as they go through the base circle region than does a normal roller rocking its way through the corresponding bottom of its cut. This is more frictional loss, not less as they claim. When the chain is off the base circle and bearing on the sides of the teeth, the pins look to be a smaller radius than a normal roller. This concentrates stress in the tooth more, not less as they claim.

The emperor has no clothes!

2_i 

From my reading of the literature in the past, most losses are in the low-frequency, subsonic waves pushed into the drivetrain, not friction. These waves are generated when full tension is applied to the teeth, but there is a wavering in there.

luns

The more I think about it, the worse it gets. On a conventional chain, the outer links are solidly joined to the pins and there is no wear between them. The inner links of a bushingless chain rest on the pins on a thickened portion formed from the material that was originally in the hole. It doesn't appear either of the plates of this chain is thickened like a normal chain's inner plates. This means the chain tension is concentrated on a narrower interface, and the higher contact pressure will mean more wear. Also, on this chain, both the inner and outer links both articulate on the pin pair's pivot, with wear from both interfaces contributing to chain stretch. So, I expect this chain will wear faster than conventional, not slower as they claim.