There has been some talk of CVTs on CD lately. So how many of you teams are actually designing them for your drive train for the 2006 season?
We are actually working on it, we have a ‘prototype’, if you will, but it’s not working as we would like, but we’re still working on it.
you should talk to with santosh and me weve thrown ideas back and forth for a couple months. i know he has some cool ideas. i however have some other things (hush hush top secret stuff) im working on.
If you get a chance, talk to Matt Krass, I have seen few of his designs. He will be able to give you some good ideas. Even better, If Andy or John get a chance, you can also talk to him about the design, I am sure he will be willing to give you guys ideas. Yes I am talking about John V Neun. 
Never had any plans for a CVT. Too difficult to get working well because it is all friction based. And friction has a much greater potential for slipping than a postive engagement setup like gears. I design for reliability and robustness(notice I did not say simplicity) and to me a CVT just is not it. However, if people do it, and it works, I’d be more than glad to see it.
After the thread where about half of the people are against even basic shifting or anything other than the FIRST drive train, I doubt there will be many going for a CVT ;).
We currently do not have a design in the works, but we did prototype one over the summer of 2002.
Our CVT was very basic and consisted of a big round aluminum disc mounted to the side of Skyway’s bead lock wheels(http://www.skywaywheels.com/products_002.htm), an aluminum frame, drill motors/transmissions, and the best part gum rubber.
We mounted the two wheels and discs to the center of the frame, mounted high traction wheels (gum rubber) to the output of the drill transmissions, and then mounted the drill transmission to a moveable assembly that would allow us to move them farther towards the center of the axis of the wheel for more speed or father away for less speed.
The problem was we did not have enough traction to keep the wheels from spinning on the aluminum discs.
This simple cvt prototype/idea was a basically taken from the little CVT’s in push lawn mowers.
I don’t have a picture of it, I don’t even know if we took any of it… so I have included a Paint drawing of what it looked like.
For those of you who know us by our gum rubber wheels (2003) robot and our gum rubber omni wheels (2005) this little project is where found out all about different kinds of rubber and what worked the best, thus GUM RUBBER!
I’m thinking a rubber coated disk and a knurled aluminum wheel would be better than a smooth aluminum disk and a rubber wheel.
Using belts is definately a good idea. You can put tons of tension through a v-belt and get brilliant traction on pulleys. So… just figure out how to get a pulley that changes size. Not so hard anymore, eh? I’m sure the brilliant young minds of FIRST can figure out how to do it.
Wouldn’t that aluminum wheel + rubber wheel have a fairly large amount of friction losses? I mean, unless the rubber wheel has a very small contact point, then it’s going to slip along pretty much all of its surface cause the aluminum disk will be travelling faster at the outer edge vs the inner edge. As I understand it, that’s why other roller designs use hard rubber designed to contact at a point.
And in answer to the thread, 57 doesn’t look to be using a CVT in the foreseeable future.
The most important thing with CVTs (Ive been researching them allot). One weight is a big issue and you need to carefully choose material. Two durability is going to be a problem. Constant grinding causes the CVT to wear down which is part of the reason they aren’t in cars (the other is reason one and the last reason). The third thing to keep in mind is friction. It doesn’t matter if you can get a great torque ratio if every time you try to push against something the gears are slipping! What you need is a durable, lightweight material, that has great friction which brings us to the last thing to consider, cost. If you want a nice drive train its going to cost you! I don’t doubt that some of you might differ in opinion but this is just my opinion based on my research. I however very much recommend pursuing this venture and figure out the best way to use them. These are very worth while to make and if you want i have some links of good materials and sites that will help you develop one. I’m very interested in teams that build them and i would love to here an ideas you have.
Actually CVT’s are in a lot of cars today. The new Toyota Prius Hybrid uses a CVT transmission, so does the Mitsubishi Colt Plus, and Ford’s Escape Hybrid uses a CVT. I don’t exactally see how constant grinding = CVT wearing down faster. Straight-cut gears in a gearbox grind more than a CVT do.
As for slipping, that depends on how much torque you’re putting through the friction plates/wheels/belts/pullies/etc… Since torque and speed are proportional, one could just gear the area where the cvt can potentially slip to a high rpm, where there’s relatively low torque.
And as for materials – material selection varies with your restrictions, you don’t have to have deep pockets to make a really good FIRST competition drivetrain, it’s all in the design and how you implement that specific design.
Actually, a good number of cars use them now like the Escape hyrbid and I think the Ford 500 too and probably many more.
EDIT: NoodleKnight beat me to it.
Yes they are now and for this refer to my final example. they are just making a comeback. CVTs have been in the car industry for a LOOOONNNNG time but they never gained popularity until recently. Yes i know their are exceptions as their always are!
I hate to debate this every time CVTs come up, but all the cars using the Toyota Hybrid System (this includes the Prius and the Ford Escape) do not have CVTs. The initial Prius advertising did say that the car had a CVT, but after a lawsuit they removed any mention of the term CVT from their advertising. The Hybrid System uses a single speed planetary transmission that is very similar to the ThunderChicken’s CCT.
However, there are quite a few CVT cars on the market now in the US, including the Ford 500, Focus, and Freestyle (say that 5 times fast!), the Audi A4 and A6, Nissan Murano, MINI Cooper, the Saturn VUE, the Honda Insight and Civic, and the Chrysler Sebring and PT Cruiser. In fact, Nissan has said that they want to sell 1 million vehicles equipped with continuously variable transmissions annually by 2007.
Incidently, I came across a really cool LEGO CVT. I don’t know how well it works, but I guess anything’s possible.
THATS AMAZING! the design really is simple. thanks for posting that link ill foward it to some other teams who are researching this.
The gum rubber drive wheels as about 3’’ in diameter and was about 1/2’’ thick.
We probably could have Knurled the aluminum disk, like Sanddrag said, but that would probably eat up the rubber quickly.
Like you said, I think the design would have probably worked better with a drive wheel that had a curved contact point rather than one with a flat contact point 1/2’’ wide on the aluminum disk.
We basically scrapped the prototype because we knew that it would be very difficult to get enough power the the wheels, without the drive wheels slipping.
I am sure we could still make it work.
Thinking back, we could probably have added a lot pressure to the drive wheels, thus we would have needed to add thrust bearings on the wheels.
We also probably could have made a better method of moving the drive wheels in and out on the disk.
Well there are these cone pulleys (STOP WHAT YOU ARE THINKING NOW). These pulelys arent being used as most cone pulley desigins put them to use.
These cone pulelys ride next to eachother as in the flattned tip of the cones almost touch eachother. There is also a V Belt riding in between these cones. The cones have grooves in them that do expand, but so does the V-belt.
There is another pulley at the other end that the belt also wraps around. The single pulley is the one that has the drive motor running it.
When the cones are pushed together, the belt has to ride up the cones becasue it has nowhere to go. as the cones are pushed together, the overall diameter of the touching point of the 2 cones grows and your gear ratio changes.
And that Nissan Murano with the powerplant under its hood, the 3.5 liter engine, with a CVT will be awesome.
1002 the CircuitRunners would love a CVT on next years bot, but that is if we can find one that can handle a lot of torque an one that will be reliable. We however will not build one if the game next year doens’t call for one. Purpose over pimpness is a policy that we are trying to adhere to.
Santosh, are you talking about something like this?
Taken from HowStuffWorks.
If it is, how could you change the diameter of the pulleys and still have a reliable drive train?
These are called variable diameter pulleys. They have been used for many years for variable speed drives on things like drill presses and mills. Team 1114 successfully adapted this concept for use on their robot cart in the fall of 2003. You can buy variable diameter pulleys from a company called hi-lo . The website is www.hi-lo.com
You might also want to look into a design by Jay Tenbrink from 494 a couple of years ago. They successfully designed and build a set of variable diameter pulleys that used a fly weight system to adjust the pulley diameter automatically.
I have spent some time looking over the pictures of their CVT on their website repeatedly, and I have some difficulty in understanding just how the diameter of the pulleys is changed by the weights. From looking at the pictures, I see that the weights fold out when the shaft is spinning faster, but I see that the pulleys remain at the same diameter. Is this a friction based CVT, because I see that the man’s hand in the background is holding onto the shaft tighter in one of the pictures than in the other.
Can someone on team 494 possibly give more information?
http://www.494martians.com/2004/pictures/cvt/cvtcenter.jpg