CVT design- make that diesel spin faster

[klwinter2]

CVT considerations.. I am building a diesel bike.. I am pondering at CVT drive.
The common issue with Diesel applications is the lower operating RPMs.

This post is about HOW and Where it introduce the RPM increase into the cvt system..

ie do we Step the RPM up before or after the CVT?

For my project
Constraints:
1. My final rear end ratio is set at 1 : 4.625 . Shaft drive.
2. Diesel engine rpm is 1/2 the gas engines rpm.
3. Due to the lower rpm of my diesel engine I calculated, I will need to increase/over speed the output to the final drive by 1.28:1.
Ie Engine makes 1 rpm and the trans/cvt output makes 1.28 rpm to the rear end.

Based on How a CVT works.. where is the best place to introduce the Rpm increase?

Should I design my rpm increase into the system BEFORE or After the CVT..
In both cases it would be via a jack shaft and gearing..

As Load sensing and rpm are the fundamental forces that are tuned in a CVT to establish performance characteristics, how do I do it?

Kenneth Winter

[tappy]

Best bet would be to increase the rpm before the CVT.
Most engines run at higher rpm than our diesels so the belts are designed for surface speeds and tension loads typical of those systems. By increasing the design rpm of the CVT you'll be able to use a narrower belt, or the same belt on smaller diameter sheaves. These will both make for a more compact CVT system. You'll probably also find that this way the centrifugal weights & cam design on the front sheaves, and the puck and spring on the rear sheaves are operating closer to their design and you'll have less trouble tuning them and/or with them wearing out.

[klwinter2]

Tappy thanks for the post.

I have a good understanding of how a CVT works.

But tuning for the diesel is confusing.

1. I can see how the CVT responds to RPM.. And thus a difference for the diesel.

Question..
1. Size. What determines the size of Disks to use?
I have tiny ones from scooters and a Massive one from a diesel UTV. It is approx 9" in diameter.
I had planned to use the disk.. But the size is killing me.

2. Do the primary and secondary disk sizes need to be the same size? Logic says no.. By having them different sizes you can manipulate the starting
ratio.. ie where 1 st gear starts..
3. IF the Disk sized are not matched... it would seem that would mean the belt can not transverse the face of the larger disk??

3. Seconday.. I have matching 9" disk from a 450 polaris atv.. I am running out of space.

4. As we all struggle with getting the RPM up.. Could I use a Larger Primary than Secondary to create a taller final ratio,, ie ground speed??

I can not figure out how to capitalize on the Diesels greater low end torque?

THANKS FOR THE INPUT...

Kenneth Winter

[gearhead1951]

If you want to "fool" the cvt into performing like it was behind a gasoline engine instead of a diesel , you are going to have to swap low end grunt for high rpm !

A diesel rated at 20 hp that will get you 45 mph can be made to do 60 mph by making the input to the cvt 1.5 rpm greater than the engine but you will now have the equivalent of a 15 hp engine !

These figure are for example , I don't actually know what the trade off would be but that is the direction that it will go !!

[coachgeo]

Dual CVT?

https://www.youtube.com/watch?v=Tnn3Pdxa4r0

[klwinter2]

I have studied this system.. Interesting.
But I dont have that much space..

I still have not found a good explaination of when to introduce the RPM increase into the system.

It appears the diesel utv and snowmobiles modify the final drive gearing..

Kenneth
East TN..

[gearhead1951]

CVT designed to run at 7000 rpm max + diesel that runs 3600 max = 1<2 step up from engine to CVT !

Simples !

[garbs]

If you can get clutches that are tuned for use in diesel application you can worry about the rpm shift after the cvt. However most likely any off the shelf clutches you get will be designed for gas engine use and therefore higher rpm ranges. These clutches can be modified to function at lower rpms (heavier weights in the primary clutch, weaker spring in the secondary) but id recommend against that as rpm is your friend in order to generate the forces needed to keep enough pressure on the belt. What tappy said was spot on, increase the rpm before the cvt. Cvt sheave and belt size are incrimental with power increase but what it boils down to is torque, how much input in ft/lbs a cvt can handle throughout the designed rpm range before the sheaves lose grip on the belt will determine what size of cvt you need. A cvt spinning faster (rpm increased before it) will see less torque and therefore can be smaller. Gearing up after the cvt will have the cvt seeing more torque and will need to be larger to prevent slip. Every cvt system I have ever seen had the secondary clutch larger than the primary and it worked out that the belt traversed the entire face of the secondary when rpm was maxed out. This will only occur if the clutches are tuned for the used rpm range however. Also having the wrong size belt will prevent this. Cvt sets are designed/tuned as a set. They are tuned to work together, swapping around primary or secondary clutches could work but they will need to be retuned to work together. This is simple in principal but cvt tuning can be wizardry in my opinion especially in high hp applications. For your application what it boils down to is the faster you can have your cvt spinning, the smaller it can be and still handle the torque input without slipping. I had the same issue you are having with my build. Search cx500 Kubota. There are two threads on it. For my build I used a comet 40 series clutch witch based on comets rating of I believe a 16hp engine max (I figured for a gas engine the torque would be about 25 ft/lbs) would not be able to handle the torque output of 45 ft/lbs for my d722 engine. But I have an 8 inch htd pulley which drives a 4 inch pulley, doubling the rpm, and halving the torque. My primary clutch is then driven by the shaft that the 4 inch pulley is on. Theoretically the max torque it will see is 22.5 ft/lbs which I figured it could take without slipping. I did all this in order to use the comet 40 series (fairly small) as opposed to a much larger cvt I would normally need to run which I had no room for.

[garbs]

My system actually spins my primary clutch too fast. With stock tuning it was trying to drive at idle and opening up to max ratio far too fast. In order to fix this I lightened the weights in the primary until I could spin it up to about 2000 rpm on a digital rpm readout drill press without it opening. This way I could idle at about 1000rpm before the primary clutch started grabbing the belt (due to 2:1 ratio)

[garbs]

Also to add. The maximum speed that we can get out of our bikes is based on our power figures. You can generally get so much speed fro so much hp. In my original configuration (d722, now a z482 for ease of maintenance and radiator relocation) I Set a desired achievable max speed that I thought my engine could give and designed my transmission system, taking into account my fixed shaft final drive, to give me that speed near maximum engine rpm. To do this you need to know the final drive ratio of your cvt system. These ratios are available online for all comet cvts. You could get all the rpm issues figure out to use a certain cvt function properly but if it is trying to drive your bike at a speed that you don't have enough power to achieve, it wont work. Start with desired attainable speed at one end and max cruising rpm at the other, then calculate for everything in between to make the numbers match up. My transmission is slightly different as it has another step after the cvt before the shaft drive in the form of a sprocket on the secondary cvt shaft driving over to a sprocket on a stub shaft driving the shaft to the diff at the wheel. For me the calculation included engine rpm at max speed (3600), 2:1 ratio of the htd pulleys, max drive ratio of the cvt, sprocket to sprocket ratio after the cvt, then the fixed final drive ratio. My system is to be honest overly complex but the added step of the chain drive after the cvt allowed me to do my final tuning to get all my numbers in line. I believe I have a 1:1.5 ratio there, slowing the rpm input back down a bit. You should be able to get everything to work fine without this step. Basically my system over compensates for my cvt size (too fast) and the chain drive corrects this. This could have been avoided by overdriving my cvt less, maybe a 1.5:1 ratio instead of 2:1 but I had to use the 8 and 4 inch pulley sizes for different alignment reasons and was stuck with the 2:1 ratio.

[tappy]

Question..
1. Size. What determines the size of Disks to use?
I have tiny ones from scooters and a Massive one from a diesel UTV. It is approx 9" in diameter.
I had planned to use the disk.. But the size is killing me.

The size of the discs is determined by the tension loads that can be transmitted by the belt. This is governed by the belt section, the clamping force, coefficient of friction and wrap angle of the belt around the discs (generally one pulley is wrapped less than 180 deg, and the other more than 180 deg, except when the ratio is 1:1).


2. Do the primary and secondary disk sizes need to be the same size? Logic says no.. By having them different sizes you can manipulate the starting
ratio.. ie where 1 st gear starts..

No, the discs (sheaves) do not need to be the same size. What IS necessary is that at least one pair of sheaves can move axially along the shaft they're mounted on, otherwise at most ratios the belt will be forced to bend slightly and this will ruin it in short order.

3. IF the Disk sized are not matched... it would seem that would mean the belt can not transverse the face of the larger disk??

Nope. The spread of ratios achievable is a combination of the sheave sizes, minimum working radius, belt length and centre distance. There's a bit of maths involved but it's all just basically geometry. If you want me to run some figures or give you the equations I'm happy to oblige given a few days.


3. Seconday.. I have matching 9" disk from a 450 polaris atv.. I am running out of space.
The largest sheave size you'll need is defined by the largest torque you need. Running straight off the crank shaft on a Hatz 2G40, with up to 54Nm of torque, I calculate that the driven sheave need be no more than 215mm diameter. If you run a jack-shaft first to increase the rpm then you won't need such large sheaves, depending on the engine you're running and the torque it puts out.

4. As we all struggle with getting the RPM up.. Could I use a Larger Primary than Secondary to create a taller final ratio,, ie ground speed??
You could yes. If the final drive (chain?) was low enough so you never needed a low "low ratio" then yes this would be perfectly do-able. The larger diameter would also allow the pucks in the front sheaves to be at a larger radius, reducing the force for a given torque, reducing friction, and improving shift action.

I can not figure out how to capitalize on the Diesels greater low end torque?
You don't need to try to capitalise on it. It's just a case of matching the torque-rpm characteristics of a diesel engine to the most commonly available CVT belt technology to achieve a compact system. And you've already hit on how to do that - either use a jack-shaft to step up the rpm the CVT runs at, or use a much larger driving sheave and smaller driven sheaves to raise the belt surface speed and driven shaft speed.

THANKS FOR THE INPUT...
You're welcome.

Kenneth Winter[/quote]