My Idea For A CVT Transmission

So I was looking into CVT transmissions, and they seem to me like they are going to be one of the biggest things for cars in the coming years.

For you guys that don't know what a CVT is here are some links
http://en.wikipedia.org/wiki/Continuously_variable_transmission
http://www.howstuffworks.com/cvt.htm

From what I understand the main problem with these are that they cannot handle to much torque, so right now the only smaller cars use them.

So I was thinking of a way to fix that and I came up with a idea. I didnt have much time to draw up a example of one. Here is one I threw togehter in paint.

http://img213.imageshack.us/img213/3840/mycvtwn7.jpg

The bottom black line is the input shaft from the engine and the one on top is the output shaft. The yellow gear is about the sam size as the orange empty one, the yellow gear turns the orange one that turns the solid red cone shaped gear. The green gear slides along the solid red cone gear, when at the top its like being in low gear but if the solid orange is slid to the bottom it is like being in high gear.

So do you think this could work?

How about ther gear pitch?

It'll work if there are no teeth.

why cant there be teeth? you would need to design some crazy ass cone shaped dealy-o though, with some sort of tooth design that would allow there to be progressively more teeth toward the fatter part, but also with a tooth merger design, im thinking sort of like how railroad tracks merge, sort of. otherwise unless there were some sort of huge gripping power between the cone gear and the green gear, something that could withstand more torque than a CVT. and i dont think thats realistic as far as manufacturing cost and complexity, without using teeth.

yea I thought about the cone gear thing and this is what I came up with

http://img221.imageshack.us/img221/6964/conegearopenfd8.jpg

Imagine that was the cone gear if it was unrolled, its not exact, but as exact as I could get it with paint, but as you move up to the bigger part of the cone and the teeth move farter apart a new tooth will fill the gap untill they are to far apart then a new one will pop up.

Its hard to explaine and im not sure if it would work but I just whipped this up in a few mins

A good idea if there are no teeth, but it would not be a CVT with teeth on the cone and output gear. The first part of the cone has 6 teeth, the second has 14 and the last, what 32? it would essentially be a 3 speed transmission with lots of friction and gnashing of the teeth. The output shaft wouldn't care about what part of the cone it's on, only how many teeth are driving it.

using teeth would be impossible
the best alternative would be belts, but that would not be much better.

look up the nu-vinci bike hub.
you design looks pretty similar to what they are making.
(sorry, nobody has new ideas... its just a rule)

Why can't you have teeth but have them get thinner in a way that another tooth can be fit in at each level? So at the tip say you have 5 teeth, then move one unit down, you'd have 6 teeth, but each would be 4/5 as wide as the last one, so the space between them is the same, eventually you'd come to a point where any thinner and it would break, but I don't see why it can't be done... There might be a small gap between the sliding gear's teeth and the cone teeth the farther down you go, but if you make the teeth long enough, they'd still contact.

Oh, and steel, it does matter which part of the cone it's on, because the radius is different, the circumference is different, one turn of the input shaft on the tip of the cone will spin the output shaft less if the sliding gear was on the base of the cone. If you could use friction to turn the gears it would be just as well, except it wouldn't slide, teeth are irrelevant on gears of the same radius.

Why can't you have teeth but have them get thinner in a way that another tooth can be fit in at each level?

To get two gears to mesh requires a distinct number of teeth and a consistent spacing.
You can't do that with a cone.
You can to it with a lot of different gears approximating a cone (like the rear cassette on a mountainbike), but then it's no longer CVT as you have seperate and distinct ratios.

The only ways to acheive CVT are:
Belts.
Cones with a drive donut fed between them.
Hydraulics.
Electrics.
Rollers and ramps. The Nuvinci hub is a variation on that theme.

It should be noted that all of these devices were "invented" hundreds of years ago, it has taken this long for metallurgy and materials science to create the materials to let them survive and computers to be able to control them.

For example, commonrail diesel engines were one of the first injection methods tried, but they didn't have the valve or computer technology to make it work reliably.

think of it this way, if you put two screws of the same type together, they will mesh together.
but if you put a wood screw next to a machine screw, they do not mesh, and you get no useful work.

Wouldn't the shearing off of the gears count as work? :biggrin:

What about using a small rubber tire and an abrasive cone with some sort of tensioner?

What about using a small rubber tire and an abrasive cone with some sort of tensioner?

That works fine, but you're limited with the amount of torque you can put through it.
You need enough force on the wheel to create traction, but too much force distorts the rubber too much and reduces efficiency (and life).

Yeah, I know. I was just kidding... couldn't you hear it in my voice?

Yeah, I know. I was just kidding... couldn't you hear it in my voice?

It's the accent, makes you hard to understand.:grinyes: