Team 19 CVT (My own version.)

[Ronald_raygun]

I've been developing my own version of a CVT-though I'll admit, it's not comparable to George's.

The idea is that the dome tilts about 15 degrees to either the left or right.
The dome itself (though not exactly shown in the picture) is basically a giant gear, and each individual wheel has floating sprocket bars.



In this one, the green circle is a gear that meshes with a disk that has floating sprocket bars in it. The grey parts are lead screws to move the disk back and forth.

Any suggestions for improvement?

[Ianworld]

I'm very intrigued by your design since i'm a big fan of CVTs but i'm not quite sure how yours works. I understand that its supposed to chnge the gear ratio as the gear slides back and forth across the large disc. The change should occur by the change in ratios between the two wheels. I follow all of that, I just don't understand what the blue bars are for.

[sanddrag]

They are rubber inserts to help the wheel get traction on the disc. Because (I assume) the wheel is not actually a gear, just a wheel.

[Ronald_raygun]

Originally Posted by sanddrag

They are rubber inserts to help the wheel get traction on the disc. Because (I assume) the wheel is not actually a gear, just a wheel.

If I were to simply make it a wheel, I would rather coat the whole disk with rubber, and make an insane amount of grooves.

NOTE: For future reference, the yellow dome, and green circle are gears. The dome has teeth like a normal gear, it's just spaced in a funky way.

Those blue bars are I call floating sprocket bars ( Note that there is no water involved ) They are made of steel or some other similar material. It's supposed to enable this CVT to have positive drive as opposed to a friction drive.

They are meant to maintain contact with the gear. This allows it to slide back and forth without breaking any teeth.

Here is a cross section of those floating sprocket bars.





With those bars, I can have a positive drive CVT. This design--if it is actually possible--should be more efficient than the other thread about a CVT made by a fellow Team 19 freshman.

Here is something that explains the concept.

[Tom Bottiglieri]

I'm not all too sure the rubber "teeth" will "catch"

Can you please explain how you are sure the rubber will provide enough friction to hold to the dome, even when pushing hundreds of nm's of torque through it?

[Ronald_raygun]

The teeth are made of steel.

Both the dome and wheel are gears. The reason why it doesn't look like that is because my CAD software isn;t as sophisticated.

[Swampdude]

I thought this was all rubber too when I first saw it. Which would entail low torque. But if the intent is to actually mesh steel teeth, you're going to have a variety of misalignments. Those "floating sprockets" would make contact at one angle which would change along the revolution. Also the contact timing would cause the driven gear to stutter the further out you have it. Then as you pull it to center it would start binding due to the angular misalignments. Unless you used pegs/dowels, that might help. But the timing thing would still happen. Also the loading wouldn't get spread out over multiple teeth, like standard gear mesh designs. Singling each tooth to take full load requires large, strong materials.
I don't see how the dome helps as far as the angle causing a different ratio either - if it's radial shape is consistent. Seems to be the same effect as detail 2 as long as you're sliding the "floating sprocket" wheels in/out from center. Also this effect causes the two shafts to counter rotate.
I applaud your thought process. I'm not sure what to offer as far as improvement goes other than trying pegs instead of teeth. Or stick to what we were seeing in the rubber contact, and low torque use.

[Ronald_raygun]

Here is the website of the inventor of the Anderson CVT.

ACVT

The Anderson CVT is a positive drive CVT which can transfer more torque than the usual friction-based CVT.

Those models are based off that CVT. The only difference is that I don't use any sort of chain, belt, etc.

Do you think it's even remotely possible to use the FSBs in another CVT design?

[Swampdude]

Originally Posted by Ronald_raygun

Do you think it's even remotely possible to use the FSBs in another CVT design?

Sure, but that's a big question. One that you should keep working on

It addresses 1/2 of the misalignment problem in his application. I'm a little skeptical on that chain strength and durability. But it does add more torque to the CVT concept.