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

Why wouldn't this pepetual motion machine work?

  1. Apr 3, 2006 #1
    The weights are attached to the main wheel by ratchets that only turn anticlockwise. The green thing is a slope that the red wheels roll down.

    It's a variation on another 'perpetual motion machine' that I saw a while ago, but I can't see what's wrong with it! It's driving me mad!

  2. jcsd
  3. Apr 3, 2006 #2


    User Avatar
    Gold Member

    Perpetual motion is nothing special. I can spin a wheel in space and it will continue to spin forever if it's not acted upon by any external force.

    The impossibility is making a machine that runs forever that can perform positive work.

    Edit: I see what you are asking. You're saying that this wheel is said to spin forever despite the effect of friction. I can't really tell what's going on in the picture...some of the weights appear to be fixed while others are free-spinning...
    Last edited: Apr 3, 2006
  4. Apr 3, 2006 #3


    User Avatar
    Gold Member

    First, you'll have to explain how it operates at all. eg. Which way does the wheel turn? One presumes clockwise.

    Note what that means though - from the moment any of the weights land on the green slope until they leaves the green slope they are no longer pulling the wheel clockwise (well, not as much anyway).

    In the position as diagrammed, two weights on the right side (actually three, instantaneously) are supported against gravity by the slope, whereas *three* weights on the left side are being pulled down by gravity. This provides a gross force pulling the wheel COUNTERclockwise, counteracting the clockwise forces.
  5. Apr 3, 2006 #4
    Yeah but the point is the weights in the bottom right section have a much greater moment because they are further out than their opposing numbers.

    Also, I don't think the slope really supports the weights that much. I mean, if it was just one weight and it was on the slope bit, it would still pull the wheel round.

    I guess the ones on the left must add up to the same as the ones on the right, but I just don't really see how...
  6. Apr 3, 2006 #5


    User Avatar
    Science Advisor
    Homework Helper

    I can't make out what it is supposed to do from the picture. I take it the wheel turns clockwise and the green slope is not attached to the wheel?
    Can't tell what the idea behind it is, but I can tell you that increasing the moment of the wheel alone will slow it down.
  7. Apr 3, 2006 #6
    If you calculate the moments in this position, the right side wins even if you say the one on the slope doesn't help:



    Yeah, the slope isn't attached to the wheel.
  8. Apr 3, 2006 #7


    User Avatar

    Staff: Mentor

    Why don't you do the simple trig and calculate exactly how much of the weight is supported by the ramp....
    Fine, but in the instant before the that, the one at the bottom of the ramp is being mostly supported by the ramp and the wheel will want to spin the other way.
  9. Apr 3, 2006 #8


    User Avatar

    Staff: Mentor

    You can probably feed this problem into a spreadsheet without too much trouble, calculating in 1 degree increments what the forces would look like. What you'll find is that sometimes it wants to spin clockwise, sometimes counterclockwise, but the average will be zero net force.
  10. Apr 3, 2006 #9
    I have viewed the gifs, and I see where the flaw is.
    Dave is correct, and I will expand on it.

    What also happens is that as the arm with the weight is "forced" to extend to the right by virtue of the fixed, unattached slope, the slope necessarily produces a "curved" backwards force on the weight which is translated to the top pivot, which is translated to the fixed "spoke", therefore all active forces balance on both sides of the wheel.

    Chris, what you need to understand is that, regardless of any mechanism used, energy is required to extend the weight. If this energy is not externally supplied, then it will take it INTERNALLY, which is the wheels motion, thus the wheel will not rotate at all.
    Now, if the arrangement creates an initially un-balanced system, the wheel will rotate until the system is self-balanced, and one would thus assume that "free-energy" is available if only for that moment.
    Wrong! It takes energy to create an un-balanced system, and any energy "extracted" from the "self-balancing" merely equals(and is often LESS) than that required to set up.
    So, their is no net gain in this design
  11. Apr 3, 2006 #10
    What's the trig to work out how much one on the slope contributes? Not as easy as it seems I don't think...
  12. Apr 3, 2006 #11
    Without the green ramp in place the wheel will turn in 'perpetual motion’ (ref: post #2) assuming ZERO friction resistance to any motion once started.

    After putting the green ramp back, where do you find the extra energy to lift the red balls above their normal path without the ramp? The ramp cannot provide a force or energy.

    Assuming the ramp can only add friction – no hope.
  13. Apr 3, 2006 #12


    User Avatar

    Staff: Mentor

    Draw a free-body-diagram, but off the top of my head, I think its just the cos of the slope times the weight. With the fbd, you can easily get all the resultant forces by drawing triangles and solving the sides.
  14. Apr 5, 2006 #13
    These two things hinder the theory of Perpetual machine to a great extent .
  15. Apr 5, 2006 #14
    Do you remember Newton's laws? It's easy to see in the general case why some arbitrary collection of point masses and rigid solids cannot provide a perpetual source of mechanical work. This example is merely such a collection of point masses and rigid solids. Therefore, it cannot provide a perpetual source of mechanical work. Why you bother going through this all these convoluted intermediate steps is beyond me.

    You should reformulate this problem in terms of the gravitational potential of several point masses, subject to constraints.
    Last edited: Apr 5, 2006
  16. Apr 6, 2006 #15


    User Avatar
    Gold Member

    To both kunalanuk and Rach3:

    I think the poster grants that
    a] we know it can't work, since perpetual motion is pretty much debunked, and
    b] the usual losses through friction can be discounted

    The question is more one of: setting aside the physical constraints, what would stop an ideal machine built like this from working?
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Why wouldn't this pepetual motion machine work?