Idea based on Abeling's patent (workaround)

[0c]

I've since given on this idea. I can't possibly do the math or simulation (found wm2d needs my money to look at others', which is just buggers).
My gut is telling me, that in your image and probaly also mine, the wheels will just not reach the speed required for the weight to make it up high enough.
You need serious speed to back back up a wheel. This requires speeds where just a hex shaped roller would just stay put on the inside diameter of the wheel from pure CF. I can't see just falling weight reach these speed when a big wheel is holding them up.
And, the weights entering on top would have the height, but miss the speed, and would need to be brought back up to speed again.

I haven't given up on the idea just yet.

1) The wheel can be started using a starter to get it up to a required speed, so we can have the weights travelling however fast we want.

2) The weight doesn't have to return all the way to the top. It only needs to get above the level of the hub. If it exits the wheel at the rim, it will travelling much faster than the hub. If it has some velocity remaining when it gets to the join position just above the hub, it can actually transfer some of its remaining momentum back to the wheel with the impact when it is caught.

As an example (arbitrary numbers, I haven't done the math), let's say the weight exits at the bottom, travelling 10 m/s. When it climbs up the ramp to just above the hub, it is only travelling at 5 m/s. The spokes near the axle are only moving at 3 m/s. So when the weight impacts the weight catching device, it transfers 2 m/s x it's mass of momentum back to the wheel. Then as the wheel rotates, the weight will move back towards the rim in preparation for the next cycle.

All we need to do is keep the weights impacting the wheel with enough force to overcome rotational losses, wind resistance and bearing friction.

I've also been trying to draw and animate this thing with wm2d. But I'm having some difficulty. Looks like I will need to design some of the pieces in a CAD program and export DXF files. Wm2d just doesn't have the drawing tools to create what I need (how do you split a circle into a half circle arc? Or how do you do a french curve "S" shaped ramp?)

[Cloxxki]

Thanks for your enthusiasm 0c. It's true that giving up, is what the great minds of past centuries have done. At least, we have an environment where it's OK to ponder about free energy devices.
What I am wondering, after the weight gave the wheel that little nudge of spare energy near the hub, this same weight has less than a 1/4 revolution of the wheel to gain 10m/s rimspeed again, from almost zero, and low height. Whether the CF naturally occuraing can provide that, or whether the nudge was greater than the push back down by the wheel, I doubt the weight would be carrying out much positive work on wheel during that phase.

After reading some loose references and ideas, I have a feeling we should really look at the 3rd derivitive of speed (increase in acceleration), to "cheat" time and distance.
Abeling's device has an increased accelerationg going, as seen in Dusty's replication. It looks impressive at least, but to which extent it efficient, will need to be proven. His setup each turn has the weights in the same positions, but the top weight is jammed onto the rim from behind with momentum to spare, and a full rev left to do work.

If your setup could do that, and the height of the wheel is in the proportion to the rim speed, who knows...

[0c]

What I am wondering, after the weight gave the wheel that little nudge of spare energy near the hub, this same weight has less than a 1/4 revolution of the wheel to gain 10m/s rimspeed again, from almost zero, and low height. Whether the CF naturally occuraing can provide that, or whether the nudge was greater than the push back down by the wheel, I doubt the weight would be carrying out much positive work on wheel during that phase.

I don't know if it's enough. I tried to get some interest in modelling the thing, by someone experienced with the tools. One of the guys did a very crude model of my first idea which returned the weights all the way to the top. Nobody else seems interested, and I have been having difficulty getting wm2d to do what I want.

Guess I'll just keep on trying, whenever I have the time.

[Cloxxki]

Referring to your linked drawing, I'd like to suggest the ball leaving the wheel at 4:30 or so. This would eliminate the low-work from the weight in the bottom section of the wheel.
Adding this height back on top should only net an improvement in efficiency.
Similar systems proposed have this same feature (wheel never reaching bottom of the wheel) for this very reason.
Also, I'm thinking the weight doesn't carry out much work until it's caught up with the rim. A feature creating an acceleration (mostly horizontal to require less work?) might aid in this, and ensure long work being done, quite vertically, far away from the rim.

[0c]

Referring to your linked drawing, I'd like to suggest the ball leaving the wheel at 4:30 or so. This would eliminate the low-work from the weight in the bottom section of the wheel.
Adding this height back on top should only net an improvement in efficiency.
Similar systems proposed have this same feature (wheel never reaching bottom of the wheel) for this very reason.
Also, I'm thinking the weight doesn't carry out much work until it's caught up with the rim. A feature creating an acceleration (mostly horizontal to require less work?) might aid in this, and ensure long work being done, quite vertically, far away from the rim.

If I can ever get a reasonably accurate sim going, I can play with those factors and see whether it helps or not.