CVT education please

[Daryl DeArman]

I think I have a general understanding of the premise. I have experience in centrifugal clutch kart racing and many years ago was involved with a Jr. Dragster team which utilized a salisbury drive. But that is it. Never even seen a F500 much less a F500's CVT. Non-existent out here.

I am assuming that the drive pulley has the ability to change effective diameter (where the belt sits) based on rpm, with a spring and weight system with the diameter growing with engine rpm.

I am also assuming that the driven pulley size corresponds to the axle rpm. Effective diameter growing smaller as the road speed increases.

I am assuming that the springs and weights chosen determine when the pulleys begin to change diameter and at what rate.

IF my understanding is incorrect, please educate me.

If my understanding is correct, how much does the front pulley effective size change once the peak tq or whatever engagement rpm is chosen is reached?

What is the difference in effective gear ratio between lowest and tallest gearing?

Thanks for the enlightenment!

[Evl]

The go-to book on all this (and it isn't too thick) is the Aaen one... worth reading. The standard low-to-high ratio is 4:1. I'll let people who know better answer the other questions.

[formulasuper]

Daryl, Evl is correct about the Aaen CVT catalogue/tech manual, F440/500 racers have used it since the beginning of time. However good luck on getting much free clutch tuning info from your competitors; that is 97.5% of the secret to winning in the class. I spent 10 years trying to get it all exactly correct & never did. Gave up & went to shifter cars.

[Daryl DeArman]

yeah, I wasn't trying to learn all the go-fast secrets. Just trying to see if I understood the concept.

Looking at the many posts regarding the possibility of 600cc 4stroke bike motors in F500 and their comparisons I was confused about how the car can stay at MAX HP rpm all the time.

I could see it never dropping below a certain, tuned-for, rpm...but wasn't sure about the range of "gear ratios" the CVT could accommodate that would cover slowest corner to fastest straight. I am sure "high" being 4 times as high as "low" would certainly cover that!

Are my basic assumptions of the theory/operation correct?--not wondering what weights, what springs, what spring preload or any other trickery....

[Evl]

Here's my basic understanding -- which includes no secrets... well, ok 1 secret: buy a car from a guy who no one can pass on the straights, and get him to go over what per-track changes he does when he sells the car. --

The primary clutch controls engine RPM. The weights provide more clamping force as the rpm increases. There is also a spring that pushes against the weights to keep the clamping force from getting to high. As the clamping force increases, the primary pulley gets bigger. In order to keep the belt from snapping, the secondary must get smaller, which means the secondary sheeves get pulled apart making the pulley smaller. The secondary has a helix in it that makes the thing basically a big wedge attached to your rear axle. This makes it torque sensitive, such that the more torque, the more it tries to get bigger, selecting a lower gear. Thats how it works _very_ roughly.

The cool thing is that it everything it working right, you get constant rpm from the engagement rpm until you hit top speed. Kinda fun to look at the rpm-speed graph if you have a data-logger.

[lancer360]

I'm a newbie in F500, but I have been studying CVT's a bit and read the Aeen manual so I will take a stab at it. I'm sure if get anything wrong, someone will correct me.

The CVT on our cars consists of two clutches, the primary or driving clutch and the secondary or driven clutch. Both clutches consist of two halves, one stationary, and one moveable, that are tapered to work with the tapered belt. They differ quite a bit from there. The primary acts initially as a clutch. It starts fully open so there is no squeeze at all on the belt. As the engine RPM's increase, a set of weights with a distinct profile start being pushed outwards by centifigal force. These are resisted by a spring. Once the RPM's reach a sufficient level which is controled by the amount of weight and by the profile of the weight, enough squeeze will be developed to pinch the belt and start the car rolling. From then on, as the RPM's climb additional squeeze will drive the two halves together and cause the belt to ride out, causing the gear ratio to change.

The secondary begins fully closed with the belt all the way out. The secondary is also controll by a spring. The secondary however does not have a set of weights. It is instead controlled by the belt tension controlled by the primary clutch stroking) and the torque applied to it. One half of the sheave is stationary and the other half rides on a slide bearing. It has three or more rollers that ride on a cam. As the torque and the tension in the belt changes, it forces the secondary sheave to open or close based on the spring rate, spring pretension, and cam profile. Often times the cam profiles are multi-angle to get the desired effect.

What makes the CVT's tricky to tune is that unlike a traditional transmission where a change to one gear typically only affects that one gear, one change to a CVT has multiple affects. You may change the weights on the primary to raise or lower your RPM's and find that it had two undesired affects. You then may have to go make changes to the secondary to try and cancel out the undersired changes which of course can have their own undesired changes and so on and so on. Really good clutch tuning can reep rewards in straight line speed and putting the power down getting out of the corners. Since part of the clutch is affected by torque, clutch tuning goes hand in hand with carb tuning to try and get that last few percent gains out of the system to be able run at the top level.

A typicall CVT unsing a Polaris primary and Artic Cat secondary setup starts around 3:1 and can go into over drive of 0.8:1. However, efficiency drops off significantly below 1:1. Controlling the final gear ration on the CVT is done by changing the gear/sprocket ratio between the jack shaft and the rear axle. This final gear ratio depends on the top speed and tire size and fine tuning this is done by monitoring the jack shaft RPM compared to the engine RPM. If data shows the CVT ratio getting significantly into overdrive, then a lower final gear ratio will be needed. Like any car, this has to be balanced against the slow parts of the track. For most tracks with decent straight aways running the 19.5" tires around 3:1 seems to be the most common. Autocross run much higher ratios, sometimes as much as 5:1.

[Jnovak]

I just wanted to clarify 1 item. The actual ratio range of a CVT, in a car & on a race track is about 2.5:1 on the slowest turns & very close to 1:1 on top end.

I define the clutch ratio as engine RPM/clutch RPM.

This is from measured track data & includes the actual slippage of the system.

Thanks ... Jay Novak

[formulasuper]

I just wanted to clarify 1 item. The actual ratio range of a CVT, in a car & on a race track is about 2.5:1 on the slowest turns & very close to 1:1 on top end.

I define the clutch ratio as engine RPM/clutch RPM.

This is from measured track data & includes the actual slippage of the system.

Thanks ... Jay Novak

Jay, by "clutch RPM.", your refering to the jackshaft rpm, correct?

[Jnovak]

Correct Scott, jackshaft RPM.

Thanks ... Jay

[Daryl DeArman]

Thanks a ton. That helps a bunch with the visualization.

One point of clarification sought:

From then on, as the RPM's climb additional squeeze will drive the two halves together and cause the belt to ride out, causing the gear ratio to change.

Do the engine rpms increase or not? Is it theoretically desired to keep them at max HP rpm but not possible to tune the clutches that perfectly? In reality, does the CVT allow the engine to always operate at max hp or just in a much narrower band than a traditional gearbox would ever allow?

[formulasuper]

The engine RPM stays the same while the car's speed increases.

[Jnovak]

The engine RPM stays the same while the car's speed increases.

This actually ONLY HAPPENS if everything is adjusted PERFECTLY. This is one of the very key elements in running at the front in F500.

Thanks ... Jay

[formulasuper]

This actually ONLY HAPPENS if everything is adjusted PERFECTLY. This is one of the very key elements in running at the front in F500.

Thanks ... Jay

Jay is correct & that's why I never did better than 6th at the RunOffs years ago.

[tedsimmons2]

This will give you a visual clue of what is being discussed. This set-up is not "perfect" as it lacked top end speed. But will give you an idea.

[Daryl DeArman]

Thanks for the data Ted...

Helps too!

The "flat" rpm trace from about 10300' to 13575' while the mph trace goes from about 70 to 123 is when the pulleys are doing their size changing, correct?

The upward spike and corresponding settling in RPMs from 10000'-10300' is when the weights and springs in the front pulley are figuring it out?

I can see the advantages of a CVT performance-wise...about the only negative that is apparent to me is the drive train loss over a conventional box.

I guess being "stuck" with 6 gears, the corresponding rpm drops, the area under the HP curve in the usable rpm range and the time lag between shifts is all made up for in the increased efficiency of the 6sp box over the CVT.

If the motor was peaky (2 stroke with aggressive port timing) versus a broader (FI four stroke with variable valve timing and intake tract lengths) then the CVT is more of an advantage.

Will be interesting to see how the F600 vs. F500 comparo shakes out. I hope they are closer in performance at a similar weight than I think they'll be. I'd love to have a F600 in my future--hope it happens.

[lancer360]

If I am remembering the data right from the Aeen handbook, a properly done a modern CVT can be as much as 90% efficient. Which should be in the same ball park as a traditional mechanical gear box. Belt squeeze and slippage is a good portion of the efficiency loss. They are doing some pretty trick stuff now with these clutches to try and keep the belt squeeze to the minimum needed which improves efficiency. More squeeze when the belt is running close to the center and you have minimal surface area and less squeeze as the belt moves out and gains surface area and leverage.