overbalanced chain drive

[oscar]

 
In order to understand the concept that is presented below, you may want to first have a look at the overbalanced chain drive featured on the web page http://www.besslerwheel.com/murilo/index.html

Before proposing some changes to that original idea I want to first analyze it a little.
On the left side of the mechanism there are more chain-links than on the right side. That makes the left side heavier.
So I will refer to the side with more chain-links as "the downward moving side or heavier side" and to the side with less chain links as "the upward moving side or lighter side".
Since the weights on the heavier side are pressing down onto the lower transport sprocket, that sprocket will start to turn, transporting chain-links to the lighter side.

The main requirement for perpetuity of that motion is the following:
The upper and lower transport sprockets need to be synchronized, so that while the lower transport sprocket turns and 1 chain-link is moved from the downward moving side to the upward moving side, the upper sprocket must do the same in reverse to refill the heavier side.
Both sprockets have to transport the same number of weights during each timespan so that the weight difference between the two sides is maintained.

Murilo Luciano's design on the quoted website aims to achieve this by means of two sprockets which are in contact with the same number of chain links and turn at the same speed/rpm.

To achieve the synchronized rotation of the sprockets, he employs a special chain. Each link of this chain has a little "strut" which is automatically engaged while on the upward moving side. The struts prop up the chain links when they are unfolded and keep them in an extended state. The struts prevent the links on the upward moving side from collapsing/folding.
In this way a tower is built on the upward moving side, with all the weight of the stacked links resting on the lower ones and all together on the lower sprocket.
Since the number of links on the underbalanced side is less than on the downward moving side, the design seems valid.

The struts disengage when the chain links are transported to the heavier side. Consequently they will fold. The height they occupy is now less than when they were in the propped up state.

If this basic idea is valid, the main technical challenge is to achieve reliable operation of the struts, that is of their automatic engagement and disengagement.

Since I had difficulties to envision this mechanism working reliably I have tried to alter the idea, so that no struts are needed.

In the design proposed here (see drawing below) the upper transport sprocket will (hopefully) pull up the extended chain from the underbalanced side at the required synchronized speed.
This is supposed to work like this:

The upper and lower transport sprockets are of the same size but the number of chain links they are in contact with is different, since the links transported by the lower sprocket are folded, whereas they are extended when they pass around the upper sprocket.

So the two transport sprockets need to turn at different speed/rpm, yet in a synchronized manner.
They are synchronized by use of an additional normal chain (bicycle chain) or a timing belt, indicated in magenta.

The bicycle chain drive is meant to be powered by the lower sprocket, driving the upper sprocket at the required speed.

The gear ratio for the bicycle chain drive in the depicted setup has been determined as follows:
The two transport sprockets in the sketched example have the same number of teeth (24).

- The behavior of the lower transport sprocket:
When it turns 1/24th of a full revolution, it transports 1 link to the other side.
In other words: when it goes round once, it will have transported 24 chain-links.

- The behavior of the upper transport sprocket:
In order to transport 1 chain link to the other side, it needs to turn 5/24th, that is ca. 0.2 of its circumference.
In order to transport 24 chain links, it will have to turn 5 times.

So the gear ratio (turns of lower sprocket vs. upper sprocket) needs to be 1:5.
This can be achieved with a lower sprocket of 100 teeth and an upper one having 20).

Known issues:

-issue1:
the proposed design needs guides (guide rails) on the downward moving side to prevent weights/links from being squeezed out to the side due to pressure from others resting above them.

-issue2:
smooth feeding process of down pressing links to the lower sprocket

-issue3:
it must be ensured that the "connection plates" of the chain links fold outward (i.e. away from the machine) when the chain links come off the upper transport sprocket.

I am contemplating to build something along this line as a proof of concept.

If anyone who has experience with mechanics instantly spots that I overlooked something and/or that this can never work because ....,
or if I can save myself some trouble if I take other important facts into account ...
please make yourself heard in this thread.

I have attached the .dxf-file which was used as a base for the image.

[FreeEnergy]

http://www.qsc.cc/gravitymotor2.htm

[Cloxxki]

@Oscar:
Thank you for bringing this idea back under the attention. I already liked its simplicity, but you've taking it to the next level!
On technical obstacle you might save yourself, is the 5:1 ratio. Go 1:4 or 1:6, to be able to pin the weight chain links with a little bicycle chain like bushing, which also fit over the upper weight sprocket's teeth. Now your chain seems to have a floating pivot point. It can certainly be done your way, but perhaps not as practical as you might.

Must say I'm impressed with you heavy side guide rails. I would propose to make the out one taller, to further enhance smooth transition. The inner might even be curved with the sprocket.

I have been looking at using an existing $3 bicycle chain as weight chain. NOt very far yet, but if feasible, it would aid replication.
Imagine we manually fold the chain as it would rest on the heavy side. We could be as modest to make the stack only 2 links wide, but more seems better here, why not 6 links wide?. You'll see the pins will not all 180 degrees. But that's fine. Just weld the "flat" ones shut. Leavint the outer pins to pivots as they're used to. Providing sufficiently welded, the chain would (with tapering guide rails encosing the whole stack from upper sprocket to top of stack and then parellel to botom of stack?) nicely fold as we first designed it to do.
On the light side, a guide box would for the chain in its extended position. The welds are supporting it laterally, it can't do anywhere.
Being a regularly pinned bicycle chain, the action on the upper sprocket will be more awkward. But, the light side box containing the extended links might aid the timing. Most complicated will be the lower sprockets, but if this chain drive is to work, even a welded bicycle chain can be made to work.
Alternative to the welds, someone with plastic fabrication technology to his/her disposal, might be able to product "toy" quality plactic strutsturning a standard $3 bicycle chain into a "welded" one, once this idea works, for international implementation, science fairs, etc.

I can't fault the chain drive, and it baffles me the thing is not better documented (dozens of failed devices is what one would expect, it being so incredibly obvious). FE instinct says "it will jam" but where?

I hope this thread will deserve the attention and support it deserves, or someone will soon offer the reason we are missing as to why this one also will not work.

Regards,

J

[quartz]

Hi all,

See also
http://img38.imageshack.us/i/chane3.gif/
http://img23.imageshack.us/i/chane4.gif/

Regards,

[helicalred]

The difference in the number of links (mass) on either side is not the only consideration. The links on the right must be accelerated to five times the velocity of those on the left. It appears to me that the system would be in equilibrium.

Regards, Bill