Multi-KW through bike chain using large sprockets?

[John in CR]

The cyclone rigs seem to have pretty good success using tiny drive sprockets, over 1kw, and regular bike chain. I've got a few hub motor outside the wheel ebike builds in the works. I'll give a Nexus 3 a go on one, but I want to try one as a 3 speed with sprockets (3 front, 3 rear for always a straight chain and little change in length). I only need about 2.5:1 as a gearing range, and I don't need much overall reduction in high gear, so my thought is to stick with larger sprockets to reduce stress on the chain. If my smallest sprocket is 18-20t or so on the front with a big road bike chainring at the wheel for low gear, is it likely to hold up pretty well to 5kw or more?

My alternative is Retro-Direct with 2 chains, but with my long bikes I hesitate to give up the ability to roll backward. I wouldn't even mind packing an extra chain and tool as long as they aren't breaking all the time.

John

 

[mwkeefer]

John,

I don't think stock sprockets will support 5kw but really the issue isn't watts or hp but rather the torque output at the particular gearing point.

With that in mind, what is your bicycle rated for weight wise and second... what is the limit of the weight factor (ie: is it the crank and chain or the frame itself)... I figure that if a bike is rated up to say a 250lb rider than it means the bottom bracket and chainrings are handling roughly 80% duty cycle of torque exceeding 250 ft lbs... the exceeding point would be factored by the mechanical advantage provided by the crankarm length.

I guess what I am thinking... if a normal bike transmission could handle a 250lb rider full tilt, then it could easily handle such power produced by a motor (provided sufficient teeth for engagement) without constant failure and I think 18-20t would be atleast 4-6 teeth of engagement which is about 4-6x the cyclone engagement right?

Recumpence, ggoodrum or miles would be better for answering this question but you will need to know the NO LOAD kV of the motor your running and the AMPs peak you plan on running to calculate the torque output..

The equation is:

1355 / kV = kT (Torque Constant Per Amp)
kT * AMPS Peak = Max Torque at Motor Shaft
kT * Reduction Stage Factor = Max Torque at Output

In the above all the solutions/results are in Inch Ounces... simply divide by 12 then again by 16 for ft lbs of torque.

Hope it helps!

-Mike

-Mike

 

[John in CR]

Thanks Mike,

It kinda helps because it gives me torque, however, not really my answer, though the 60+ftlbs I came up with is interesting. Let's assume 0 reduction from the motor to the wheel. Then a chain driven with 22t sprockets front and rear will have half the force acting on it as using 11t sprockets (or do I need to factor Pi in there for less than 1/6th?), because the larger sprockets create a better lever at the wheel. That's my question, will using large sprockets front and rear allow me to get away with bike chain on a high power rig?

 

[mwkeefer]

Thanks Mike,

It kinda helps because it gives me torque, however, not really my answer, though the 60+ftlbs I came up with is interesting. Let's assume 0 reduction from the motor to the wheel. Then a chain driven with 22t sprockets front and rear will have half the force acting on it as using 11t sprockets (or do I need to factor Pi in there for less than 1/6th?), because the larger sprockets create a better lever at the wheel. That's my question, will using large sprockets front and rear allow me to get away with bike chain on a high power rig?

John,

Though I can't be definitive (and name dropped in the hopes of getting some backup) and once you mention Pie (yea I know) you nearly loose me but lets look at it this way:

1.) Calculate the torque to the crank from your existing 1000w cyclone setup for a baseline we know it can handle (use the no-load kV of the cyclone motor to calculate the torque.. your numbers will be a little off for the planetary reduction losses but it will be good enough for a starting point).

Now lets answer your question the best I can... it doesn't half the force but it does / would distribute it (in theory) over 2x the engagement teeth as the 11t would provide. I looked around the net for ratings on sprockets but other than internal shifting hubs I couldn't find normal sprocket / chain torque ratings (#35 bicycle right?)...

however if we take your existing cyclone torque output and double it, then the stress on the chain would (imho - lack of knowledge, this is instinct) be double... increase the engagement teeth on each end and the stress on the chain returns to the same level, I think sheer forces come into play at the bottom bracket and the rear but I am not sure how to factor for these (side loads) and you already said your going for as straight a chainline as possible.

Hope it helps!

-Mike

 

[John in CR]

I don't think it's the teeth necessarily, because I could file every other tooth, something I may do, off of the 22t sprockets, so they would have 11t just like the real 11t sprockets that are half the diameter. To me it's about the radius of the lever acting on the wheel, so I'm pretty sure Pi isn't involved an there would be half the force pulling on the chain with the large sprocket for the same force applied by the wheel at the ground.

Another thing making me think my idea may work even at pretty high power is the tiny sprockets used on those fancy bikes with the gearboxes enclosed at the crank. The sprockets at the rear on those look like only 11t or so, creating the nice clean thin chainline.

 

[veloman]

I don't think you'd have a problem with quality bike chainrings, at 5kw. To be on the safe side, I'd stay away from full throttle at low speed. The best track cycling sprinters are putting out 2000w at 80-100rpms during a standing start, and their equipment does it without a hickup. If your motor is putting out 5kw at > 500rpm, then torque should be significantly less.

I myself have tried my damnedest at breaking my 9 speed road chain/cogs, going apesheet from a stop in my granny gear on a 15% grade. I have readings of over 150 ft-lbs at the crank from that. (Probably much higher since that reading is 2 seconds in, and at ~10mph, not zero). Never an issue. Bike chains/chainrings can handle a lot more than most people think. As long as the chainline is at least fairly straight and you don't shift!

 

[liveforphysics]

Methods did a bicycle-chain drive Cyclone kit that he pumped 10kw into. It would wear the pin-bushings out on bicycle chain in just a few miles, then shred itself.


Only the difference in radius or diameter (it doesn't matter which you choose to go by, just as long as it's the same) makes for the gear reduction.

John, if you're going to throw that enormous BLDC motor with the mechanical delta/wye switch built-in onto a bicycle via-chain drive, you're going to want to use #219 chain. That motor you've got is capable of more torque than any hubmotor we've seen on a bicycle, and would likely spin a mark-cycle hubmotor backwards in a game of tug-of-war.

If it's the motor I'm thinking, picture something with a rotor width of maybe 3", a massive diameter, a 3/4" shaft, and maybe 4lbs of copper in the thing. It was the biggest BLDC motor I've ever held or seen in person. And it should make a totally wicked E-bike!!!

 

[John in CR]

Veloman,
Thanks, that's the kind of comparative info I was looking for.

Luke,
Maybe I'm just being hard headed about this, because it's a question I've brought up here before without an answer. Please straighten me out if I'm wrong. I had to do a drawing that represents pretty well the comparison I'm looking at with about the gear ratio I'll have. Assume the same motor, speed, and throttle position directly attached to the front sprocket.

What I see in the top drawing is a greater mechanical advantage at the wheel due to the larger sprocket (just like going to a lower gear on a pedal bike), so the force (I hope the correct physics term) applied to the chain is lower for the same acceleration. Because the work done is the same, yes the chain will move faster in terms of links per second, but is that much of an issue? I don't think a direct comparison to a cyclone is a good one due to the tiny drive sprocket they have to use with their much higher Kv motor. I'm going with much lower Kv, and much larger sprockets both front and rear.

BTW, I hope you're right about that 2 speed motor, but I don't see how it's comparable to the Markcycle motor without mine being at a far higher voltage, though my ventilated version may be capable of higher continuous power. As a sealed hub motor in a 10" scooter it propelled a 300lb lead sled plus rider at only about 35mph at 60v, and quite a few people melted theirs down by taking off in WYE instead of Delta. At greater diameter and about double the weight and half the Kv, I'm sure Markcycle's is more motor. With mods, my heat dissipation advantage may give me the edge, but I think mine is more comparable to an Etek or Agni, since some gear reduction and out of wheel use required. Plus I'll need some kind of housing including good clean air intake and exhaust, but I should enjoy near silence, and maybe a power advantage since mine is essentially a brushless outrunner and can go to higher voltage than those brushed motors that would fall apart at 100v or more. I think once I pull this off we'll see a lot of people hacking up large scooter hub motors. It seems so easy I can't believe I'm the first to try. BTW, No I wasn't considering of bike chain for this motor other than a test of the Nexus 3 taking it easy at first.

John

 

[liveforphysics]

You're totally right, in the top drawing, there is half the stress on the chain for the same amount of gearing reduction as the lower one.


I will reply more when I finish up with some stuff at work.

-Luke

 

[John in CR]

Cool,
I wanted to try it because:
1. Bike chain handles the side to side curving to move it to different sprockets, though due to the high power I want the chain straight when settled on a gear (the reason for just 3 speeds). ie Innermost front with innermost rear, etc.
2. Cheap and readily available chains and wide variety of sprockets, and a spare is even compact and easy to carry along. Maybe I get away with something as easy as a road bike chainring on the wheel, and inverted mountain bike chainring on the front. That may not give me the gearing range I want, but I should be able to at least exceed the Delta/WYE difference in order to prove the idea before getting into the details of final gearing or the complications of Delta/WYE switching.

 

[veloman]

I have no experience with high powered motors, just so you know, lol. In Luke's example, 5-10kw sustained may have a much different impact on the drivetrain than 2kw at a higher torque, for 10 seconds. So there might be a wear issue.

I believe one of Matt's first RC recumbent projects used a bike chain as the final drive, and he said it was solid, I believe. In that first video of his, the belt drive was what was slipping. I think he was running a 3kw motor.


If you can, a single speed bmx/track bicycle chain is going to be stronger than a gear 7/8/9 speed chain. But if you are changing gears, then you obviously need the latter.


Another thing, bike chains aren't designed for super high rpms I believe. Using a 219 cart chain is more suited to the higher rpms I would think. This may or may not be a problem if using a bike chain.


Something interesting, I calculated that I actually produce more torque with my legs than most 4 cyclinder car engine cylinders. I just do it at 1/100th of the velocity (and therefore my peak power is 1/70th of the car's peak power).

 

[liveforphysics]

Here is one guy's test with a few KW at high torque through bicycle chain.


http://endless-sphere.com/forums/viewtopic.php?f=2&t=15704&start=45#p256609

 

[John in CR]

Sorry Luke, let me put it in a better context, because my multi-KW title is misleading. I'm not talking about the fun for you and I power of my red bike. I'm talking at most what you felt on my blue bike, which peaks in the 4-5kw range. ie That slow ramp up controller and one of my big hubs still sealed, or a tiny direct drive hub motor ventilated for better power. I'm searching for a cheap but reliable solution for the locals that will still climb a long steep hill, albeit slowly. I might even use it to go on an all afternoon leisurely ride across the mountain. Also, I need a solution to gear slower for kids and still climb steep hills. Sure high turn count X5 motors can do it, but they're simply too expensive especially with the controller and batts needed, but no other hub motors will do it all.

A kid can get up a 20% grade with my big hubs in a 20" wheel, but they're just too fast for kids, so it needs gearing down anyway. Put enough weight on it or a steep enough hill, and a single speed isn't going to cut it without being geared for 15mph top speed or costing thousands. I believe a properly implemented hub motor out of the wheel is the best cheap solution right now that is still near silent and very reliable, and I want to try to come up with something different and much cheaper than Stoke Monkey that gives the motor multi-speed gearing without a geared hub that costs as much as a cheap motor with controller.

I don't know if you were out riding at school commute times, but you'd be surprised at the number of people pedaling their kids to school. A week or so ago I literally had a family I'd never seen or met, sitting on the wall outside our gate waiting for me to return home to ask about my bike...where to buy and then how much for me to build them one. The even more unusual thing was it was the wife asking. Hub motors are a great answer but only for flat terrain. Add in hills of any significance and there's no reliable and economical solution available unless you're a cyclist looking for a bit of assist.

That's for the locals. Now for a bike you or I would get the Silent EV grin, is there a strong chain that can also bend enough side to side for multi-sprocket gear changing? If not then guess what? It time to come up with one. I don't even know the correct terminology to try to google it.

John