A suggestion for the unidirectional Gera wheel

[DrWhat]

I can't wait to see the finished product.

[path_finder]

The two plates (of the last shots) are now assembled.
The wheel is on the floor, but not mounted yet on its frame.
The links (electric wires in copper) are shown in white on the shot below.
Note that on this model no strings at all have been used.
The geometry is not the exact the copy of the first animation shown above: the red weights are on the left side (ascending) of the wheel (for KAS).
Nevertheless the position of the yellow weights is conform.

[KAS]

Path-finder,

I can see this build progressing nicely.

You state that the red weights act on the ascent, differently from your original animation.

Could you explain?

I'm a little confused.

Regards

Kas

[path_finder]

Dear KAS,
You said in the previous page:

It appears that if a lever moves any distance on the descent side to form an axial gap, that same gap ensures the balance occurs by increasing the mass on the opposite side.

I take in account all pertinent comments. This is the reason why I made a model with the red acting weights on the left ascending side, in view to confirm/deny your comment. The only way was to build it (there is too much philosophy in this forum). I can give to you the answer: acting the red weights on the ascending side gives a poor result, and in any case it seems to be much more efficient on the descending right side (like previously shown in my animation) like the COG path calculation above was presuming.
BTW the correct model has been built and the corresponding shot is here below.
The links are colored in black for a better view.

[path_finder]

Some data on the experiments made with the above depicted wheel (see last shot):
- The curved planks are a good idea. The torque is very efficient lifting-up the weight smoothly. But it was very difficult to find the exact right position for the red weight. I made also some tests without these planks and apparently I can obtain an equivalent result without them. So far I decided to ignore them for the moment and tested several configurations.
- Working on a previous model with the 'Watt linkage' like worker, I discovered by chance a new configuration giving some surprising good results (an kind of avalanche process). The principle of this new design is given in the two documents hereafter:
The first one is an animation showing the constitution of the worker.
The second one is a shot of one of the four workers built by myself. The pivot C is fixed on the outer rim of the wheel. The first pendulum (with the yellow A weight) has a length of R/2 (if R is the radius of the wheel). The pivot D is located inside the wheel. The central rod passes over the exact center of the wheel (very important feature, minimizing the negative torque of the red weight during the return phase)..

As indicated by the red circle the B red weight can rotate a full 360 grades around the D pivot. Here is the biggest difficulty for the building because at certain moment the B red weight must pass inside the center rod. This is the reason of the special shape for this B weight: it has been cut in two parts. The value of the weight is equal to the A yellow weight, but repartited in two more thicker parts, allowing these two weights to pass inside the central rod in view to make a full 360 grades turn. But the question now is 'how pass the pivot also inside?'. It's obvious that a serious topological contention is present here. The solution consists in removing one of the two rails (marked with the violet cross on the shot).
The last shot shows a wheel with the four workers, but not with the last model as depicted here: so far this model was not able to make the 360 grades turn.
In addition the four yellow weights were locked for the shot.
A shot of the latest wheel incorporating these modified new workers, and some other details will be given very soon.