Posted by Nick Hall (195.74.122.242) on July 11, 2003 at 11:14:04:
In Reply to: Grim - I must be thick! posted by Nick Hall on July 11, 2003 at 07:27:11:
Grim - thanks for the clarifications.
Thinking about this one, the crucial thing is to track the _aggregate_ centre of mass (ACOM)for the whole thing from the self-start point to where it breaks up.
It is not difficult to make a machine which self starts, completes several revolutions, but which is actually "running down" in the true sense - i.e. with each revolution the ACOM is lower than the previous revolution. In your design this would have meant that the 'switching position' of the masses occured further round the axis with each turn (although with heavy masses, there would not have to be a lot of angular 'slip' - just enough to allow the loss of PE of the ACOM to provide sufficient energy to overcome friction, sound losses etc)
I suspect that in this design it may be very hard to see over a small number of revolutions (and dangerous to look too closely!)
If this is a correct analysis, and if it was made so that it did not fly apart, it would eventually stop because the ACOM can go no lower.
I faced this problem of analysis a couple of times with wheels I had made, so I wrote a quick app (I`m a multimedia programmer) which allows me to video the thing working, then stop frame the video and mark the mass positions.
The programme ultimately gives a table of cartesian co-ordinates for the masses and resultant torque for each degree of revolution over several cycles. With that data you can see whether or not there is an additional input of energy or whether it is effectively running down...
Interesting though...
Thanks
Nick