Thanks for that. They actually need to be 100 and 50 radius rather than diameter (to give a 150mm crank length), so 200 and 100mm diameter. It seems that once you approach that sort of size, the prices get silly. I’ll have a look at that site though. You’ve made a few geared machines in your time… do you agree with my assumption that nylon gears (even 10mm wide ones) would not be strong enough? They’re way cheaper, but I’m leaning towards over-engineering in the absence of proper calculation.
Nylon is really great stuff as long as there is no applied torque. After that, things get a little bit fluid. Plastics do so like to flow. In this case you’re talking about a one tooth nylon contact surface that is 1cm wide, not a six steel tooth contact surface that is 1cm wide. There is an enormous difference in the yield strength.
Sorry about not being able to tell the difference between a diameter and a radius. You would think I would be able to do that by now. Oh, well, what’s a factor of two here and there?
This is exactly why our dakorman operation is so far over-budget/under-funded/screwed. Well that and the fact he’s been funneling our seed money into a screwball idea for a “Home Particle Collider Kit” that he thinks will be a big hit with the high school science geek set.
On a more serious note,Rob I had some experience using cog gears in my failed geared unicycle experiment.My problem was that those gears mounted on a cantilever shaft will do all they can to twist ,bend or distort until they lose mesh whenever large torque forces are applied,just something you could take in to consideration.
Yes that was one of my concerns. I’m planning to use a thick bar for the body of the crank so I can have two bearings on each shaft, slightly spread out. Hopefully that’ll hold the shafts as parallel as possible (and will be quite heavy, but I’m not worried about that for now). The cogs will be a tight press fit on the shafts (with pin or key to stop rotation), so should stay still - we’ll see! I’m not intending to be riding miles on these - I’m just curious to see what they’re like to ride.
Hope it’s OK to resurrect this - found it after mention of the system came up elsewhere on the forums. Or is there some more recent thread I’ve missed (I’m a newbie to uni and these forums).
Excuse me if I’m repeating something somebody else has suggested, but I’ve not spotted this idea suggested on this thread - maybe on a different more recent one?
It struck me that nobody has considered that we’re not using the optimum tool for the job using conventional spur gears for this system (interestingly, the original design is pretty much exactly the same as in the diagram at http://en.wikipedia.org/wiki/Gear#Sun_and_planet). Spur gears are good when you’re trying to fit gearing into a small space. Not so good when you’re trying to make it cover a larger distance. Of course in the cycling world - even in the unicycle world - we already have a solution for gearing when you want the two axes to be a significant distance apart - the good old chain and sprocket. Now I’ve not calculated any numbers on this, but ISTM the parts on a chain and sprocket drive can be a lot smaller to transmit the same torque than what is required for spur gears, so that instantly solves another issue.
Of course there is an elephant in the room - the original system works because of the reversal in direction in spur gears. You don’t get that with chain and sprocket, and with equal size sprockets you’d be pedalling madly going nowhere (for a very short time before you fell off). You can make the sprocket on the wheel axle larger than the one at the pedal, but all you’d achieve by that is gearing down - a novelty and possibly something practical to make and test the theory (Rob?), but not something anybody wants. Alternatively if you make the sprocket at the pedal larger you could gear up, but you’d have to pedal backwards.
So we need a means of torque reversal - this is about where I’m stuck (and maybe where others have got stuck if they’ve thought of this). Is it possible with modern flexible chain to run it in a figure of 8 somehow? I open the floor for others to take this further. Apologies again if I’m just repeating old stuff.
Oh, and of course you still have the issue of pedal torque!
I’d imagine if you ran a figure-of-8 chain it would have problems under high torque/tension. Would certainly be easier to build though, I agree. Perhaps a pair of smaller gears at one end to reverse the drive, then a chain for most of the length - that would avoid having to use very big gears and may keep the cost down. I was thinking about using a line of small gears (as long as it’s an even number) along the crank, but that would potentially introduce loads of slop. Two big gears would be the simplest way, so there’s only one meshing involved, but big fine-toothed gears are not easy to come by at anything I’m willing to pay for a silly experiment.
Might think about a chain with two gears at one end… you’ve got me back onto thinking about this again, thanks (I think!)
I’ve also thought of a solution to the issue of both cranks falling to the bottom when your feet aren’t on the pedals. If you’re using ISIS then you have space inside the wheel axle to run a connecting rod. Clearly you have the issue of getting past the “crank” bolts, but typical ISIS crank bolts could have a hole drilled through the middle without significantly affecting strength, and the connecting rod doesn’t need to support a lot of torque, so a thin rod should be fine.
That would work - more effort than I’m prepared to go to though. I’m not expecting these things to be even close to practical for everyday riding - after “riding” the lego model with my fingers I think there will be far more problem with pedals trying to flip over under acceleration and braking (or trying to snap your ankles if you strap your feet on) to be worth worrying about little things like keeping the cranks lined up when you dismount.
(I think!)