Designing for efficiency

[akumabito]

Excellent info, Euromodder, bedankt! Looks like I have some reading up to do now..

[akumabito]

Wow, threads sure do move to the second page quickly around here. But that's a good sign, right? An active community, I like that.

I have given the whole thing a little more thought and I have decided against the smaller engine from the Honda Innova. It's the semi-auto transmission that made me decide against it - to coast or glide means having to shift to neutral first, rather than just holding the clutch. I'm not even sure engine-off coasting is even possible. I'm looking at the manual-clutch 125cc machines now, such as the Honda CBR125.

I hope to be able to make a choice soon (ish) so I can move forward.

[spirilis]

FWIW, re-gearing or tuning the shifting behavior of a CVT isn't all that hard, if your CVT is designed similar to my Yamaha Majesty (400cc maxi-scooter). The shift points are primarily controlled by the Variator Weights (some metal & graphite roller weights wedged in grooves behind the sliding front pulley and a cover behind it) and the rear sheave spring, which governs the amount of force that attempts to keep the rear pulleys pressed together. Most tuning on the Yamaha Majesty's shift points are done by replacing the weights with different-mass weights or changing the variator to a different design (and changing the # of weights installed).

Sounds much easier than swapping gears on a manual transmission IMO. But on the flipside, a belt-driven CVT needs to have its belt swapped every so often (every 12K miles on the Majesty..... at ~$110/belt, that can be a bit expensive, and USUALLY the variator weights themselves are worn down by then so they need replacement too, that's about $60 a set).

[Frank Lee]

The problem with rubber belt CVTs as I see it, is that I've read they MAX OUT at 80% efficiency. Compare that to up to 98% for gearsets, flat rubber belts, and chains.

[Christ]

regarding rubber belt CVTs, I wonder if that 80% figure isn't an average over the course of operation, including during "upshifting" and "downshifting" when the belt is being heated by friction from tension changes...

Unless they're saying that the belt is slipping alot, which I could understand the 80% figure then, however, if that's the case, why not use Kevlar or something? Lasts longer, doesn't slip as much, and has less deflection under load, so that should increase the efficiency of the unit as a whole, should it not?

What about single CVT vs. Dual CVT (two ranged pulleys vs. a fixed pulley and a ranged pulley.)

I haven't seen any data suggesting a difference, though there obviously should be.

[spirilis]

hm, interesting about the 80% figure, although I have a little trouble believing it. The CVT in my scooter is only air cooled. Though who knows, that might not mean anything. FWIW I am using a kevlar belt right now, meaning rubber with kevlar reinforcement bands inside vs. whatever they use in the stock belt (which I'll admit I'm not entirely sure... I just bought a new stock belt for my 24K mile service and it has some kind of reinforcement bands built in)

Not sure what you mean by the single vs dual CVT? I guess by that definition mine is a dual CVT -- front sheave assembly is 1 fixed pulley and 1 ranged pulley (called the Variator) and the rear sheave assembly is 1 fixed pulley (opposite side of the belt from the front) and 1 ranged pulley with a strong spring pushing it towards the fixed pulley.

The belt becomes stretched with more tension as more torque is applied from the engine so it helps prevent slippage. This is accomplished with clever design of the sliding mechanism between the rear sheave ranged pulley as it slides against the collar of the fixed pulley (the channels it slides on are curved so that greater torque application results in the rear pulleys wanting to come together to downshift, despite the fact that the front variator pulley's weights want to push the front pulleys together to upshift with increasing force as engine RPM's increase--result is that when you are twisting the throttle back hard and producing your greatest torque, it downshifts yet the tension on the belt is phenomenally higher than the tension on the belt during normal coasting, so the increased torque throughput is met with a belt that's so tense it won't slip).

Really a brilliantly clever balancing act done by purely passive mechanical means. I grin every time I think about it.

[spirilis]

Well then again, that belt is pretty thick and stiff and each segment has to frequently change shape, so that could be the efficiency loss. It's a lot like a tire in that regard.
Chains have lubrication providing an advantage over that, and a flat (thinner) belt has less material to change shape at any point on the pulley.

[Frank Lee]

^You got it.

If anyone has better info on CVT efficiency, bring it! Believe me, I've looked for it.

[Boreas]

Do you want a reverse gear? or is there a way to work around that with a small motorcycle engine.

[Boreas]

I have a chinese verson of the yamaha scooter drive train in my 4 wheeler. It has a fan built into the front pulley and an air vent into and out of the belt cover. There must be some waste heat generated with this set-up.