Constant Variable Tranmission for CF SC's

[325icintn]

Does anyone know if this rather simple technology has been applied to CF SC units? Theoretically, you could get max boost at all rpms.

Some of these pulleys are readily sourced from scooter transmissions for $100 new, and probably next to free from junk yards.

Consider the following from another site:

http://flexistentialist.org/gallery_photos/cvtbasics/2variator.sized.jpg

Lets start with the front pulley of the variator. This is where your roller weights and ramp plate are located. This image is a cross-section of the front pulley, showing the front half of the pulley, the belt, and the rollers sitting against the ramp plate.

Now as the RPM’s of the motor increase, the centrifugal force pushes the roller weights outward (number 1 in picture).

The roller weights push out and onto the angle plates surface. This causes the rear half of the pulley to move toward the front half of the pulley (number 2 in picture).

When the rear half of the pulley pushes to the front pulley, it forces the belt out to a higher gear ratio (number 3 in picture). [Different roller weights produce an increase or decrease in the rate of change of the pulley size].

Faster engine speeds cause the belt to go outward [and reduce the speed of the unit].

http://scooter-center.com/catalog/images/products/big10682-0.jpg

[paul e]

>>Faster engine speeds cause the belt to go outward [and reduce the speed of the unit].<<

Sounds interesting... Are you saying that, at slow engine speeds the blower is spinning faster, and at wot redline, its spinning slower, such that the goal is a constant rate of spin, thereby, delivering as much boost down low as up high?

Im afraid I dont follow the diagram, but Im a bit dyslexic that way.

What would happen at idle? Any examples of this in action?

[325icintn]

>>Faster engine speeds cause the belt to go outward [and reduce the speed of the unit].<<

Sounds interesting... Are you saying that, at slow engine speeds the blower is spinning faster, and at wot redline, its spinning slower, such that the goal is a constant rate of spin, thereby, delivering as much boost down low as up high?

Im afraid I dont follow the diagram, but Im a bit dyslexic that way.

What would happen at idle? Any examples of this in action?

You may be dislexic, but I'm just plain dumb. Anyway, I am saying faster at lower rpm and slower as high rpm, producing a near constant rate of spin.

Ideally, you would want a constant ratio between the rpm's of the engine and the rpms of the sc unit.

[xatlas0]

The CVT described uses a basic mechanical proportional controller. It will have a steady state error since it lacks a mechanical integrator controller. (something I haven't ever seen so I don't think it has been made.) The amount of error, given the output RPM ranges required, can be as high as 10%. The device you are describing is usually called a Watt governor, since it was made by Dr. Watt to control the speed of steam engines.

Think about this, given the propensity for fan blades to violently give way, imagine if it was two or more metal balls bouncing around in your engine bay?

I looked into CVTs when I was designing one for a senior design project. The belted designs require new belts constantly. Plus, designing for a high torque really reduces the life of the belt. You couldn't use a typical multi-rib V belt, since the pulley "width" changes with the RPM. Using a singl V belt requires a huge V belt to get similar surface friction, which also means the sides will wear out very quickly.

A significant problem with a Watt givernor is the transient responce. When you would slow down, the controller would overspin the compressor, causing surge. Similarly, when the user gooses the engine, the mechanical controller will make it overshoot the intended final RPM, making it surge again. The typical transient responce of a Watt governor of the size required would be measured in seconds, which is way, way too slow for a car. A certain amount of mass is required for the controler, which is proportonal to the force required to "open" and "close" the pulley. More mass means worse transient responce, in this case.

[325icintn]

How are these drawbacks compensated or not a problem for the scooter transmissions?

I think kevlar belt technology has taken aware the wear issue.

[xatlas0]

How are these drawbacks compensated or not a problem for the scooter transmissions?

I think kevlar belt technology has taken aware the wear issue.

The drawbacks are present, they just aren't as visible. The engine speeds up, the engine slows down, but the ratio changes slower than the engine can change RPMs. In a scooter, the transmission isn't the only thing dictating the output speed, the engine does too.

Kevlar belts are not as good as suggested. Most of the automotive grade belted CVTs fail (by belt shredding) in less than 40k miles, and that is when they use steel/kevlar belts. Belts suck. They are OK for low power operations, like the scooter, and even then, they fail pretty often. Trying to channel more power and more torque makes the belts be not so good. Look at motorcycles, the belts needs to be replaced once every 10k miles.

You would need to size the belt and system to handle the max draw of the supercharger, so somewhere around 50-100hp with some measure of torque. This will require one of two things: an very, very tight belt, or a torroidal system. A belt would work ok if you felt like replacing it about once a year, more often if you hammer on the car.

More concern than the belt, however, is the mechanical governor. A better option would be to use a series of poppet valves to regulate the ratio by using the pressure difference versus a BOV spring. At least then it would be far less likely to explosively disassemble. I researched this style of system for my CVT project, but it had very similar issues as the mechanical system when it comes down to transients. To make it respond fast enough, it has to be computer controlled.

[325icintn]

thanks for the thoughtful responses.

The cs in your sig still makes me drool.

[Keaton]

the belt would slip, and wouldn't work. the super charger is spinning 10k-20krpms.

if you have ever driven a CVT go-kart or golf kart with a CVT, and have gotten stuck in the sand, or on rock the belt just slips and you don't move a whole lot.
also every CVT system i have seen has 2 pullies that change size

[xatlas0]

thanks for the thoughtful responses.

The cs in your sig still makes me drool.

Thank you. The CS is still "in progress", so to speak. I have yet to drive it and not have something else break. But, hey, that means it will eventually be new! :stickoutt

As for the CVT stuff, I had to learn it for my senior project, which pretty much evolved into designing a torroidal CVT for a Honda CBR600 that both fit in the stock housing and could run in the worst operating conditions for a week. Before settling on the torroidal and while figuring out the controls, we (the group, there were 5 of us) pretty much fleshed out the current state of CVTs as we could find it. Hence, I felt compelled to talk about your proposal.

The CVT we designed was (based on FEA, so who knows how it would actually do... :shifty) able to run for 5 days at ~30k RPMs, transmit 110hp and 60lb-ft of torque continuously. After that it would have disintegrated. It could have lasted longer, but because ofthe way the trans is in a CBR, we were limited to using the engine oil. On the downside, it had to be made of titanium-carbide-cobalt, and would have cost ~20k to make.