Sjack Abeling Gravity Wheel and the Worlds first Weight Power Plant

[mondrasek]

@mondrasek,

This will not change the main conclusion but will be good to see a sketch.

I think @eisenficker2000's method is correct because the only thing that matters is the direction of the effective weight. If no track is present the effective weight vector would be the initial weight of the sphere and to obtain the acting vector on the arm we need to multiply it only by the sinus of the angle between the weight vector itself and the arm.

In our case here the effective weight vector is less than the actual weight and this decrease is only determined by the track (the ramp), not the slots. The slots provide only the point of application of the said vectors. The ramp is the piece that does the job of ensuring persistence of the negative torque throughout the whole turn. The method of analysis used is OK.

Recall I turned you attention to that difference between your analysis and that of @eisenficker2000 and mentioned that his is the correct one.

Like I said, I missed it too.  But the method we have been using is not correct.  All but two or three of the weights are in contact with two surfaces:  The ramps and the wheel slots.  Both of those angles must be considered when resolving the forces applied to those surface due to the weight.

For example, if you pick up a 10 lb bowling ball with a flat frictionless plate from the bottom, you must hold it with 10 lbs of force.  If you hold it with two plates in a vee, each plate must apply a force.  The total of these forces is not 10 lbs.  But the vector sum will be.  If the vee is so steep that the flat surfaces are almost directly on the sides of the bowling ball you must exert extreme forces, much greater than 10 lbs each to hold the weight of the 10 lb ball.

But why will this design also keel?

[Omnibus]

@mondrasek,

I already told you that this way of calculating the torques is incorrect. It is non-physical. The correct way to calculate torques is the way @eisenficker2000 has done it. That it is the correct calculation is corroborated by the most important, crucial fact for that wheel, namely, by the fact that the center of mass is persistently situated sideways to the axis of rotation for any position of the wheel. If we are to place vectors frivolously and carry out random calculations we can "prove" this is a non-working device in many other ways. However, every time a calculation is made there is one criterion it must be verified by -- whether or not it conforms with the staying mass-axle discrepancy. If it does not conform, such calculation is out.

[Cloxxki]

Dusty's first generation replication (2 weights only) also seemed to wants to turn backwards at times?
I would contribute this to the position(s) of enhanced vertical speed and/or acceleration. You're putting energy (back) into the weight there, taking it from the wheel. Start in that position from standstill, and the "up" weight will have increased leverage, and thus pull down the counterweight, despite being closer to the axle.
The slots are effectively an infinitely variable gearbox, shifting up and down on given positions of a revolution. Think like this: If I turn the wheel 1 degree, how much vertically do both weights move? And can I compare this to their average distance from the axle during this 1degree rotation?

[mondrasek]

@mondrasek's video is a reply to this http://www.youtube.com/watch?v=9-dT4MZCtYo&feature=channel_page I posted, showing apparent perpetuum mobile. @mondrasek's is at two orders of magnitude lower integration limits and smaller animation step.

Actually this is not the case.

Both videos are to be discarded because they use wm2d. The approximation methods and who knows what else which wm2d applies make it unsuitable for such simulations.

Also incorrect.  The first video shows how WM2D can behave incorrectly, just like any simulation package, if the input and parameters given allow for too large of a calculation error.  The video I posted shows how WM2D does in fact correctly predict the action once the input and parameters have been adjusted to eliminate the source of the large calculation errors and with more accuracy.

[mondrasek]

@mondrasek,

I already told you that this way of calculating the torques is incorrect. It is non-physical. The correct way to calculate torques is the way @eisenficker2000 has done it. That it is the correct calculation is corroborated by the most important, crucial fact for that wheel, namely, by the fact that the center of mass is persistently situated sideways to the axis of rotation for any position of the wheel. If we are to place vectors frivolously and carry out random calculations we can "prove" this is a non-working device in many other ways. However, every time a calculation is made there is one criterion it must be verified by -- whether or not it conforms with the staying mass-axle discrepancy. If it does not conform, such calculation is out.

I am sorry Omnibus, but you are wrong.  The vectors are now placed correctly, not frivolously.

You are accepting incorrect methods because they corroborate what you believe is a condition that requires rotation, ie the CoG of the system being to the right of the axle.  But this does not mean it must rotate:

If the wheel was constructed without ramps and guides it would find equilibrium due to a keeling effect.  When the ramps and guides are initially moved into position they do work on the weights to push them into the orbits we are analyzing.  This adds energy to the system.  It also pushes the CoG off center to the right.  But the system is still in equilibrium, not overbalanced.  The energy put into the system by installing the guides and ramps cannot be extracted unless they are allowed to move back out of the system.  This is exactly how I can push the CoG of a pendulum off center by installing a ramp or guide to do so.  Putting that ramp or guide in place inputs energy into the system.