Dare I say a plausable multiple roller drive for friction mount...

[bigkat650]

I think a spring has the potential to work-- and the great thing about a spring is that it really doesn't get much simpler or easier... or cheaper. I can consult with one of my good friends who is a marketing manager for a regional spring supplier to see what he thought would work well for the situation.

As far as roller size, I was thinking like 5/8 to 7/8inch for the low gear... and something like 1.5 to 2.0 for the high gear... Which should get just about anyone up any hill... pending on motor size... and still allow for high top-end, with the bonus of lower RPM at a moderate cruising speed. And you could use different roller sizes as well, just a suggestion on sizes.


Ultimately, I have found air pistons that simplify the process... however they are for several hundred dollars, and not really cost efficient. Really if done right it should be a rather sleek set-up, running the lines zip-tied to the bike frame... Obviously the 3-way valve would need to be mounted... my idea is mounting it just back from the handle bars on the top cross bar.

 

[5-7HEAVEN]

I Can See It!

bigkat, I see your idea very clearly, but using 50-pound springs, turnbuckle and lever. I can visualize the engine mounted on a steel plate and central pivot 1/2" bolt which is perpendicular to the vertical mounting bracket. Under the engine plate are two Staton friction rollers suspended by 5/8" bearings and hangers. The rollers' threaded ends screw onto 2" stubs which mount 10-tooth sprockets. These sprockets align with the centrifugal clutch. The engine's 10T sprocket clutch chains to one friction roller, which then chains to the other spindle. Of course, the bearing's pillow blocks mount via elongated holes which allows chain adjustment. Spacers under the engine adjusts tension to the first friction roller. Springs provide 50 lbs. tension onto the tire w/1.25" roller. A strong cable with locking lever controls the shift to the 1.5" friction roller.

I would have considered thid option a few months ago, since I was very interested in shifting for more engine rpm. Instead, I chose the simpler, easier but costlier way by using one engine per wheel, each with different size friction roller.

 

[sparky]

Yes, this is very similar to SimpleSimon's idea a few weeks ago.

I don't think he used CO2, and I sure wouldn't even consider it because I'd be switching at every stop. That means I'd have to change CO2 cartridges every other day, if not every day.

 

[SimpleSimon]

I posted a diagram and an explanation of how to implement just this idea bout a month ago. I believe in K.I.S.S in any mechanical device. All that is needed to make it workable is a compression spring, an after market manual choke cable with a variable locking capability, and a little ingenuity. No need for air, or compressed gas, or valves, or any of that.

See: http://www.motoredbikes.com/showthread.php?t=14302 , post #8 and on.

 

[loquin]

A simpler approach (no power loss, anyway) may be to have a sliding outer tube over an inner, grooved (or splined) shaft.

The inner, grooved shaft is directly driven by the clutch bell.

The sliding outer tube is the 'roller,' and is machined with two diameters, and knurled on the drive surfaces.

In the sketch below, slide the roller to the left, so that the larger diameter portion of the tube is contact with the tire, and you have the high speed. When you slide the roller tio the right, so the smaller diameter portion is in contact with the tire, you have low speed. The drive has to be spring loaded, so that you can go to high speed on the fly. Essentially, you'll have a throwout bearing at each end of the roller.

Ref. the attached sketch

***
If you decided to try the dual roller approach, use a timing belt instead of a chain. It's much less noisy, and would be much lower maintenance, especially since you could use a short, wide belt.

Also, rather than spring loading the entire drive, mount it on a 'rocker'. Tilt it forwards at the top, and the front roller is in contact and the rear roller spins freely. Tilt it back, and the rear roller makes contact, and the front disengages, per the second sketch. You could also have an adjustable stop at each position to be able to set the down pressure individually ...

 

[sparky]

What?!? If the roller was cone shaped... it could be made into a CVT, right?

Imagine that... a friction drive CVT.

 

[loquin]

What?!? If the roller was cone shaped... it could be made into a CVT, right?

Yes... but it would wear the tire unevenly, (and would have very poor efficiency) as the different diameters on the cross section of the roller would be spinning at different linear speeds. While this would be acceptable for a few seconds when shifting, you definitely wouldn't want to run in this mode continuously. You need a 'flat' (fixed diameter) portion, at both the high and low speeds, to be be in contact with the tire.

 

[sparky]

Bummer. At least with your idea, the possibility of a third roller is more likely than the other method.

I wonder how easy it would be to "shift on the fly" with any of these multiple roller designs?? Would compression spring really be all that's needed for Loquin's method?

 

[srdavo]

I've been gone for a while......but.....

...It's great to see the MBc "think tank" at work!!!!

keep the fresh ideas flowing!!

 

[SimpleSimon]

A simpler approach (no power loss, anyway) may be to have a sliding outer tube over an inner, grooved (or splined) shaft.

The inner, grooved shaft is directly driven by the clutch bell.

The sliding outer tube is the 'roller,' and is machined with two diameters, and knurled on the drive surfaces.

In the sketch below, slide the roller to the left, so that the larger diameter portion of the tube is contact with the tire, and you have the high speed. When you slide the roller tio the right, so the smaller diameter portion is in contact with the tire, you have low speed. The drive has to be spring loaded, so that you can go to high speed on the fly. Essentially, you'll have a throwout bearing at each end of the roller.

Ref. the attached sketch

***
If you decided to try the dual roller approach, use a timing belt instead of a chain. It's much less noisy, and would be much lower maintenance, especially since you could use a short, wide belt.

Also, rather than spring loading the entire drive, mount it on a 'rocker'. Tilt it forwards at the top, and the front roller is in contact and the rear roller spins freely. Tilt it back, and the rear roller makes contact, and the front disengages, per the second sketch. You could also have an adjustable stop at each position to be able to set the down pressure individually ...

I like that sliding roller idea quite a lot - no reason it couldn't be a three-speed, depending on tire width. The pivot idea is basically what I was addressing - I suggested a compression spring mostly to simplify control - if default is one roller being pressed against the tire, all you need is a single pull choke type cable to shift to the other. I agree that a timing (toothed) belt is a good alternative.