Excess Torque Hypothesis : Mechanical Sustainable Imbalance Possibilities ( <>> ) ?

[Fletcher]

Wow Wubbly .. that's impressive. And one big scripting effort. Many thanks for taking the time and effort to build it.

I may have to take it for a spin ;7)

[Wubbly]

It wasn't that hard. Unfortunately I don't think WM2D has the horsepower to process all those elements. To test it I would draw a circle in the middle of the chain somewhere, pin it to the background, and run it. The chain fell down onto the circle and you can watch the simulation grind to a halt as it tries to figure out all the interactions of the links.

[Fletcher]

I know. I can handle it tho. What I did was build the chain and place it in the sim. Made all objects NOT Collide. Then selectively (using shift) had the chain collide with the roller wheels etc.

That was quick and relatively painless, and the sim could handle it.

[daxwc]

I have been doing lots of contemplating on your design Fletcher. My analyse is of all the weights in both mirrored systems only 2A and 3A are working to lift 1A. I am not so sure 8A isn’t nullifying 3A.

My quest is how do we now get the other side of one system (4A, 5A, 6A, 7A) to redirect torque to 1A?

[daxwc]

Another direction:
Maybe we would be better off to get rid of the chain. Turn both mirrored systems around and have 2A fight for the same space as 2B. One weight would hit the other and knock the other system off balance so neither system can find balance. Of course you would have to ungear the two systems.

[Fletcher]

I have been doing lots of contemplating on your design Fletcher. My analysis is of all the weights in both mirrored systems only 2A and 3A are working to lift 1A. I am not so sure 8A isn’t nullifying 3A.

My quest. is how do we now get the other side of one system (4A, 5A, 6A, 7A) to redirect torque to 1A?

Hi dax ..

Another direction:
Maybe we would be better off to get rid of the chain. Turn both mirrored systems around and have 2A fight for the same space as 2B. One weight would hit the other and knock the other system off balance so neither system can find balance. Of course you would have to ungear the two systems.

Well, I'm glad it gets some neurons firing in different directions for you dax.

Just to sum up. The Gaffle Jack arrangement I came up with is based on the basic actions of a SB (pantagraph/scissor), re-engineered. It is an alternative in-series inter-connection method between lever-weights (lws), to what B. shows with the MT9 family, which are direct connected by simple ropes or cords.

My sims showed that in-series cord only inter-connections were very weak i.e. the lower hanging lws were unable to raise up 8A or even 1A. Therefore the fan shape (peacock's tail affect) cannot be achieved that he shows for this family. However the Gaffle changes all that, because the leverage factors are changed in a positive way. Now there is good strength from the Jack Effect and the lower hanging lws combine forces and thru the linkage devices lose GPE but can raise up the higher lws so that they can gain GPE. The proviso that NET GPE lost is greater than the GPE gained as it is for all mechanical actions. Unfortunately the single 'Gaffle' wheel had negative torque tendencies i.e. it rotated the wrong way, opposite of the direction the Jack worked.

What I was attempting to do was take this negative torque out of the equation so to speak. By gearing a second system so that torques were nulled or cancelled. Then placing the Prime Mover Chain over one Jack system (which would hold it out to one side) so that overall the system was torque biased in my favour.

Finding that combination of the two carrier wheels lead to many sim experiments. Two inside the same wheel on the same axle. Two separate axles but geared. Turning in the same direction or opposite directions etc.

While the Theory of Torque Cancelling then Adding Back Imbalance Torque from a Chain seems logical it was hard to simulate an effective way to implement it successfully (for me anyway).

Then I got sidetracked for a while. I thought, lets think about this differently.

A Ramelli Balance (geared Roberval) is automatically torque neutral in all positions. Let's start with that and add a Chain. But I couldn't use the 'Gaffle' Jacking system with a Ramelli, just lws and rollers set at an angle.

So I built a table-top mock-up of it in my workshop using wooden gears and a bike chain. At the same time I finally managed to get a sim to work of the chain added.

Now I had a Torque Neutral (Cancelling) Carrier System that kept the lws at down 45 degrees orientation. And over them to one side was hung a circular chain for imbalance.

(I'll have to dig out the sims and put up a pic for you to look at more closely i.e. how I got the chain arranged etc).

Anyhoo .. as expected the sim showed a movement in the direction the chain was offset. But then the sim slowed down and stood still eventually i.e. it did not accelerate.

[I'm dredging my memory banks right about now to recall the hows and why.]

Basically the chain had a residual swaying motion as it moved around. This wasted energy. And the lws were unable to lift the chain to its next hand-off, and it found PQ.

Conclusion : Even with a Ramelli Balanced Wheel where an Off-Set Chain is added it could not accelerate and create sustained imbalance !

IMO this was mechanically the near perfect imbalance proposition. And it did not work !

So the idea/theory of the Ramelli method of torque nulling was discarded, by me !

[I'll see if I can find some pics and a sim].

[daxwc]

Fletcher:

Unfortunately the single 'Gaffle' wheel had negative torque tendencies i.e. it rotated the wrong way, opposite of the direction the Jack worked.

I never realised that or forgotten it. That makes sense the eventual direction you took now. So where is its eventual (natural) keel point? 1A at 3 o’clock?

IMO this was mechanically the near perfect imbalance proposition. And it did not work !
So the idea/theory of the Ramelli method of torque nulling was discarded, by me !

Noted; but the Gaffle still interests me.

Now there is good strength from the Jack Effect and the lower hanging lws combine forces and thru the linkage devices lose GPE but can raise up the higher lws so that they can gain GPE.

So 7A, 6A, 5A, 4A are contributing nothing to the Gaffle leverage itself they are just negatively torqueing the wheel or did I miss something?

[Fletcher]

That's right. The natural keel point is when enough of the lws have deployed (fanned). Since system GPE gained is less than GPE lost the system CoM/CoG is below and to the side of the axle and eventually it rests directly below. Even if just below.

Only when the in-between cord (part of the Gaffle engineering) is taut (full stretch and tension) can each successive down-going lw apply its force (leverage) to the one above. What I referred to as a "Ganged Jack Effect".

When the cord is not taut then there can be no transference of force, and that lw is no longer part of the "gang" doing active work, but it still has a torque contribution to the whole system, albeit negative as described until PQ is reached.

[Fletcher]

Here are some photos I took this morning. They show the table-top sized static mock-up of my Ramelli + OOB Chain system I mentioned yesterday.

[I included the can of CRC in the shot for easy size comparison.]

I built the sim of this build first then used those dimensions for the build in wood. WM is great for the designing process. And worked well when 3D was a factor of the build to be figured out.

I'll search out the original sim(s) and put a pic of it up so you can see at a glance the 3D rendered as 2D, and follow the actions and chain hand-offs to rollers etc.

[Tarsier79]

Looks like a great build Fletcher. Nice work.

[Fletcher]

Thanks. I enjoyed the process in the most part. I also built an Off-Set version for interests sake, with split axles where one axle stub is higher than the other. Took a bit of figuring out and was still too friction..ish. I'll include a basic Off-Set pic (not a pic of the sim(s) coz it takes ages to go thru thousands of them to find them) to get the idea vs the Ramelli approach.

That was about mid last year when I went off on a tangent to kinda build a book-end OOB project. My so called 'near perfect' OOB proposition. Which didn't work (as I suspected).

Anyways .. here is a pic of the Ramelli Balance + OOB Chain sim design. Hunted thru hundreds and hundreds of sims to find one I knew I had somewhere. Whew - found it !

N.B. the red rollers are on ONE side disk, while the blue rollers are on the 45 degree angled levers attached to the gears. The chain hands-off from roller sets.

Included for the purists is the sim. Expect it to take a while to calculate as with the chain included it was right on the limit for my laptop. I can't even face it to try and make an animation (above my pay-grade & too lazy lol).

Anyhoo .. when the sim runs the chain settles down and sways as said earlier (system energy loss). And comes to a stop.

[Fletcher]

P.S. the Ramelli Type sim with chain (and pic) is about as stripped down (simplistic) as I could make it and still remain functional. That is why it looks stylistic rather than an accurate rendition of the real world wooden mock-up with levers and rollers off gears etc, which was based on the sim.

I always make sims as simple as possible and practical to cut down on complexity (less moving/colliding parts) to not overwhelm the laptop. I had to do that in this case because of the added chain. Still took some work.

[daxwc]

Impressive build Fletcher I must say. Seems you are getting better at building with age as well as figuring out the simplest design. Thanks for sharing as well as the results.

[Fletcher]

You're welcome dax .. outta my rainy day file. Only have done about 5 actual builds in 20 odd years. Thousands of sim experiments and a handful of real world experiments of interest.

Getting better at building ? Going small requires a bit more thought and care maybe. That's why I forced myself to do it. To improve the skills (& sits on my shelf in the shed). FYI the gears were not that great. YouTube came in handy and directed me to a gear template program etc. It was just a mock-up so didn't see the need for really good meshing gears. If I had would have had an engineering shop or sign-writer outfit knock them out. They would have the full-noise professional programs with correct pitch etc. Otherwise Kaine and his 3D printer, with some arm twisting lol.

Figuring out the simplest design ? Just about to start another mock-up in a week or two with luck [one and done]. Maybe a bit larger this time but will still take a few months I'm guessing. Absolutely bare minimum bare basics of a principle, as far as I can tell. Then I'll hopefully have some interesting data to play with.

[Fletcher]

No one has asked - maybe shy or still thinking. For those wondering and wanting clarification.

The sim pic of the Rameili + Chain device a post or two ago shows only 5 planetary gear systems with central Sun gear. That was necessary to reduce sim complexity so it would run with a Chain included. Before placing in the Chain Prime Mover into the sim I was usually building Planetary systems with anything from 5 to 12 and usually just 8. So the mock-up build is an 8 planetary gear system with levers and rollers at down-going angle of 45 degrees, which is different from the sim pic of only 5 gears.

As you can imagine the system CoM is meaningless in this design. There are ghost actions happening. When you look at the pic it is clear that the chain is physically right side displaced for CW rotation. It hangs over the blue lever-rollers attached to the planetary gears which maintain their orientation at all times.

The perhaps ghost element is where the wheel 'feels' the forces of the Chain. When it is supported by the blue lever-rollers the Chain weight is 'felt' at the center pivot of the gear [essence of a Roberval Balance] and not out at the roller where it physically is in contact. When the blue roller hands-over the Chain to the sidewall red roller the Chain weight is felt at the red roller where it physically contacts.

So the so called advantage is the small difference between the radius of the inner red rollers and the pivot/axle of the outer planetary gears radius being slightly greater.