Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

New Continuously Variable Transmission

  1. Jan 11, 2013 #1
    https://www.youtube.com/watch?v=nOzRzmfh884


    Unlike other CVTs that use belts or steel balls, this is obviously not friction based therefore it should be more efficient and be able to withstand higher torques without any risk of slippage.

    The main problem that must be solved is the trepidation that appears in the transmission. I think that using 2 such mechanisms in parallel working in opposite directions would cancel out the forces generated by the rod movement and by the bearing oscillation.

    I'm also not sure about the durability of the roller clutch because it would have to take billions of high torque couplings.

    What do you guys think?
     
  2. jcsd
  3. Jan 11, 2014 #2
    Here it is, a rough working mechanism to help you better understand the principle :smile:

     
    Last edited by a moderator: Sep 25, 2014
  4. Jan 18, 2014 #3
    the best cvt right now is the lasted toyota hybrid synergy drive. Two electric motors, one ICE, two planetary gear sets, no clutches or bands, all gears and bearings
     
  5. Jan 18, 2014 #4

    Mech_Engineer

    User Avatar
    Science Advisor
    Gold Member

    I applaud your work but I'm sorry to say this "transmission" won't work for an automotive application. It's an oscillating 4-bar mechanism which would be severely limited in its speed capabilities, vibration concerns, and as you point out the output requires a one way bearing and the output is only being driven half the time.

    It will be a clunky, poor excuse for a transmission with gears and drive shafts. Sorry.
     
  6. Jan 19, 2014 #5
    As far as I know that is not a CVT in itself but rather an integrated power train and it requires a battery to work.
    It is a hybrid drive, not a CVT that you would just connect to an ICE and freely adjust the ratio.

    I know that right now it's not suitable for automotive. But I think it's possible to make it practical if more research is put into it (for which I don't have the money or time, I just wanted to show the potential of such a concept).
    Take the IC engine for example. It is a heavily oscillating device and the first engine didn't have too much power density and it was vibrating like hell. Today they are powerful and much smoother.

    At the end of the first video I said that 2 such devices could be symetrically coupled so that the vibrations would cancel out (like pistons do in an ICE).
    Also, the bearings for the 2 mechanisms would be oriented in different directions then the outputs would be connected through a 1:1 gear system. This way the output would be driven full time, in a wavy manner of course but adding a well designed elastic coupling would smooth the output.
     
  7. Jan 20, 2014 #6

    Baluncore

    User Avatar
    Science Advisor

    A friction drive is not inherently inefficient. It is inefficient only if it does slip.
     
  8. Jan 20, 2014 #7
    Not inherently, but in practice they are (about 70% efficiency for belt drive even when it does not slip). Designing them to transmit high torques makes them really inefficient.
     
  9. Jan 22, 2014 #8
    This isn't a bad idea. However, you really have to see what happens with this under real loads and do a kinematic force analysis. I remember a while back that an Australian named Steve Durnin claimed to have invented a CVT called the D-Drive that didn't use friction, but a complex gearing system. News articles claimed it was the holy grail of power transmission. Turns out some engineers took a look at it later and found out he had created an overdesigned planetary gear system that wouldn't work under any sort of applied load. Afterwards, news articles were printed with headlines that said something like, "So, about that holy grail thing...".

    Not saying that this is the same kind of idea, but to really prove the concept there's got to be a more realistic use case than just spinning some wheels or gears in the air. If this were modified to propel a toy car or something with some real weight on it, I'd take it more seriously.
     
  10. Jan 22, 2014 #9
    True, the whole "D-Drive" mechanism could be reduced to a simple differential where one half is driven by a secondary electric motor which is just a replacement for a friction disc i.e. poor efficiency under useful torque.

    Regarding my invention, unfortunately lego technic lacks the components needed to build an automatic ratio changing system so the idea that I would have to stop the car to change the ratio wasn't that exciting to step up from the static concept. Other than that it would work, it can take reasonable loads.
     
    Last edited: Jan 22, 2014
  11. Jan 22, 2014 #10
  12. Jan 23, 2014 #11
    Thanks for the heads up!
    My video is already more than 1 year old so this technology may actually be a variation of my idea and not the other way. Do you know when those products appeared on the market?
     
  13. Jan 23, 2014 #12
    I'm retired and we used them on paper converting machinery when I was a newly minted engineer . The technology is at least 40+ years old.

    AceEngineer
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: New Continuously Variable Transmission
  1. Gears in transmission (Replies: 5)

  2. Gearbox - transmission (Replies: 3)

Loading...