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

[Fletcher]

This thread is a continuation of the discussion from here .. https://www.besslerwheel.com/forum/view ... 357#170357

Below are pics of my Static Table-Top Build of some time ago. I won't go into the background of why I built it as I've discussed why in Wubbly's thread.

The first pic shows the basic MT9 group type build I first built, with 12 roller lever weights (rlws). Each weight representation (they are dowls with black paper circular cutouts pinned to them) can be of any mass you like, as long as they are the same mass. All other parts are assumed to be identically made low density wood and thus of low mass contribution to the wheels CoM/CoG. This reflects the conditions of my sims. Each weight is connected in-series via a weight-to-weight rope connection as B. depicts in his MT9 Groupings. N.B. for this static test scenario I did not attempt to build a SB or Gaffle lever and slot pull/thrust system, especially at that scale. It was purely to make a real-world alternative to sim-world model to compare too and get a feel for.

The rlws can be positioned by hand and friction is such that they stay where put. As does the wheel on the axle. The rlws once deployed away from the rim apply their 'ganged' leverage and form a Lifting or Positioning Jack, imo.

As can be seen on closer inspection the Turning Moments (TM) for each weight can be estimated or calculated for any deployment scenarios with simple grid analysis. The TMs are the sum of the negative and positive horizontal (x) distances from a vertical line down thru the axle (Center of Rotation (CoR)) datum. All TMs do is give a positive or negative position from the datum line where the average rlw would be positioned (i.e. ONE rlw). The CoM/CoG is closer to the datum line by a factor of 12 in this instance and is the average position of where 12 weights would be located. The upshot is that TMs are useful for calculating the torque contribution at any one time of ONE weight equivalent on the system as a whole.

As can be seen although there are rlws out to the right at around 3 o'cl there are also ones hanging down below. The net result is a negative TM and CoM/CoG which is counter-productive at this stage i.e. CCW rotation tendencies - but unavoidable as ME outlined because the system also had to lose CoM.y (system GPE) to create any kind of torque profile, helpful or otherwise.

At this stage I was interested in the Jack's Thrust Generating Capabilities if I imagined it to be a Gaffle and Pulley system linkage already detailed elsewhere.

[Fletcher]

The next pic is the Static Model above with the addition of a nearly massless chain link running around the perimeter and over the rlws as my simulated Prime Mover analogue. It was intended as an approximation of the 'apparatus or structure' the rlws were to be 'contained within'. In reality I used a light weight nylon string which was robust and couldn't stretch much but was still flexible.

It's purpose at that stage was to continue a thought experiment of the likely viability of the 'universal lifting technique' of the 'rope/chain' I had envisioned. Because its circumference was predetermined to my best guess I believed the Jack Effect and rlw at 3 o'cl etc would bulge the rope outwards to the right and simultaneously bring the bottom lws inwards and upwards as long as system GPE recovered was less than system GPE lost as per CoE. Vertical distance lifted and angle of closure would become irrelevant with this technique within reason. This would reposition the bottom rlws closer to the rim and mitigate the system CCW rotation tendencies. Latching them somehow after being hoisted quickly would be optimal.

But it would not likely create sustainable CW TMs even tho it looks good in that position in that pic.

I'll provide the sim equivalent in another post.

[Fletcher]

Lastly in this series are 3 pics.

They are the same model with the rope Prime Mover replaced by a wooden link chain analogue of my making.

One simulates a light mass chain. One a heavier mass chain, and the last a longer length circular heavier mass chain.

The things to note is four-fold imo.

1. the anticipated positions of the rlws for each iteration of chain mass and length. And the effect on the the 'Carrier-Wheel' CoM1.

2. the likely mitigating effect to system CoM/CoG of the Chains mass (CoM2) itself which is right displaced from CoR.

3. the bottom lifting ability of rlws an internal ramp analogue (the chain) could theoretically provide when the chain is shortened appropriately, and IF desired.

4. the Jack's effectiveness to provide side-ways thrust depending on chain mass - the rlws at 3 o'cl must be able to deploy fully so there is likely a trade-off relationship between chain mass and thrusting capability.

The math ain't that easy and I couldn't sim them for crashing out ! I simmed what I could and had to intuit and thought experiment the rest.

A potential 'Preponderance Principle' method was developing in my minds eye.

[murilo]

Fletcher,
not to criticize but just a single TIP:
in this kind of devices I would never use a PAIR number for elements...
You are a good builder!
Best of luck!
M

[MrTim]

A few years ago, I built a 'chain drive' wheel, but it didn't work.

It used 8 grooved counterbalanced wheels (very light, being built out of balsa & poster board.) The points of rotation (PoR) of all the mechanisms were equi-distant from the main axle. At the PoR, a heavy balancing wt hung down, and a bar extended horizontally 90 degrees from the PoR. (i.e. an upside down L structure.) The grooved wheel was attached at the free end of the horizontal bar, and was large enough to enclose the balancing wt within it's circumference (the only way to prevent interference.)
The counterweight mechanism kept each wheel horizontal as the main wheel rotated. At 3 o'clock, the grooved wheel extended the rim out to it's furthest from the PoR, while at 9 o'clock the rim was in closer to the PoR.
(It worked very well up to this point. ;-)
The problem came with putting a chain over the wheels. (I used a strip of paper with plaster wts glued to it.)
The mass of the chain disordered the (upper) wheels (despite the counterbalancing ), along with that noticeable sag at the bottom shown in your pictures above. "Tightening" the chain only produced more disorder.
Needless to say (but it needs to be said ;-) it had no inclination to turn the main wheel at all (even after the chain was lightened (several times))
So (to me) this type of mechanism is a dead end (though it certainly looked possible at the time.)
I have since moved on to more fiendish designs... ;-)

[Fletcher]

Thanx guys .. all good grist for the mill - wouldn't want to throw the baby out with the bath water if an external chain could have potential with some augmentation ?

Ultimately that's the value of discussions, to share ideas and debate them.

Still gotta find a potential Preponderance situation to allow for excess torque conditions and sustain rotation. In this configuration the true CoM.x is almost zero, or close to 'balanced', as discussed earlier.

Here's the sim and pics of the massless pulley system lifting test as promised. Because the pulley length is somewhat foreshortened to cause lifting the No.2 rlw doesn't thrust until the wheel has rotated, which I have done with a physical method of changing mass thru an X datum rather than use a motor to turn it.

[ME]

Very nice build!

(I'll ask softly: isn't your "circular heavier mass chain" just demonstrating the Simon Stevin principle?)

[Fletcher]

Psss .. Yes, that's why we gotta augment it again, imo !

[Fletcher]

Joking aside ME .. I had looked at the MT9 Grouping many times years ago, and simulated many of them. They didn't show much, if any, promise. But they were front and center in MT with some positive comments about them from B. which struck me again as curious.

I decided one day to really have another crack at them. Persist. See if I could string together some sort of coherent narrative (join the dots) in association with the Toys Page items. For instance A could be a chain (MT113). As Oystein once pointed out they also added in Roman numerals to xv or vvv. B looked like a similar arrangement with falling levers at the top section. It could be two systems, one internal and the other external. Items C & D looked like duplicate mechs (albeit in opposite orientations, mirrored like 47, CW v's CCW etc) and it seemed redundant to have them both there, unless there was a purpose for that. E was clearly a SB. 5. was a later added rotating top but there was no pull-rope. It must be in the wheel somewhere from other evidence.

I kept coming back to the opposing C and D items which I couldn't fathom - then it struck me of the similarity to the dual opposing pendulums of the DT drawings. Speed regulators maybe, but why two, and opposing ? They would be balanced, or close too !

So the thought processes went like this - what if we took an ordinary false OOB carrier wheel that had say a side-ways thrust capability albeit with CCW tendencies. Added a Prime Mover chain with CW tendencies, then they would be close to CoM.x balanced. CoE says that !

But if I went from MT11's cue and added a second internal false OOB wheel at say 1/2 radius (I'd have to halve the lengths of the levers and double the masses etc) levers falling to the opposite side, I potentially could unbalance my nice and close to balanced initial dual system by augmenting with a third.

+1 + ( -1 + +1 ) or CW + ( CCW + CW ) .. or could be like this ( +1 + -1 ) + +1 or ( CW + CCW ) + CW

Perhaps mechanically sustained excess torque (imbalance) could be achieved in that simple way ? And that's where I'm at at the minute. I sure can't sim anything easily of that complexity tho if it worked you'd see the potential at a glance.

<>>

I'd like all of us to explore the possibilities and potential of the hypothesis. The door may slam shut but at least we will collectively know more than yesterday.

[Fletcher]

If such a system could be designed I know that we would loose some energy from transitioning weights impacting somewhere etc. They have to accelerate and decelerate (be constrained) somehow. And that means energy losses.

The Stevin's Problem might just be sidestepped for once.

The question is can there still be sufficient torque surplus to overcome that inconvenient truth ?

FWIW I once built something like Mr Tim described, almost the same, I even tried a rudimentary double-up and reverse the carrier system as I've described. What I hadn't done was have a built in ganging Jack (zusammen gehangten) and thrusting system predicated on the humble SB.

[Fletcher]

Good things come in three's ;7) Not a dual system but a triple system, or a trinity ! The AP drawing anyone ?

"they are enclosed in a structure or framework, and co-ordinated in such a way that not only are they prevented from attaining their desired equilibrium or 'point of rest', but they must for ever seek it, thereby developing an impressive velocity which is proportional to their mass and to the dimensions of their housing." - DT pg 191

The difference to this approach to sustaining imbalance is that we actually have a tangible 'framework' around the whole shebang, that being the hanging chain Prime Mover, in-which everything else is enclosed. And it must contribute to the outcome. Without it we have balance.

With a little imagination its quite easy to see the metaphor of a child tapping along a hoop (seen in the street). The rlws at 3 o'cl for a CW wheel tap and push the chain in my hypothesis.

When I get some time I'll put up a simplified pic of the 3 parts and how their torques theoretically can combine to perhaps give a sustainable excess.

[ME]

I sure can't sim anything easily of that complexity tho if it worked you'd see the potential at a glance.

For the image, maybe you can draw a sequence on paper so it can be animated?

[Fletcher]

Check ME .. I'll be working on it over the weekend.

Still using the sim for a simplified layout plan but will probably have to use MS Paint to complete the picture of what the essence of this hypothesis and theory is about.

Hopefully things will be much clearer, one way or another.

I'll get back to you when its done. Juggling other things as well.

[Fletcher]

Rightio .. let's see what are the positives and the negatives are ?

Altho B's. concepts were all going around together on a single axle, this is quite hard to imagine, let alone build I suspect. If my hypothesis is sound then it will require some massaging at the very least to mount on one axle.

So for ease of presentation I've split my hypothesis into 3 entities, on 2 separate axles side by side. To perhaps get a better and clearer picture of what I hoped to achieve.

The first pic is my standard Gaffle Pulley System and 8 roller lever-weight wheel. The rlws are pulled (thrust) outwards at or about 3 o'cl from the Jack Effect of in-series connection. Note that we have to imagine latches holding the rlws until release position as I can not sim them adequately. Also note that the rlws are free to hang down until restored to the rim position on the ascending side.

You will probably gather that the System CoM/CoG (little black and white circle) is to the left and below the axle giving the wheel CCW tendencies. This is not helpful for rlws thrusting outwards on cue.

The second pic is simply the above wheel on its axle mirrored to give another wheel on a separate axle beside it. Notice that both wheels are not connected by any means and their torques are LHS CW and RHS CCW. But both still not helping the deployment of their rlws at the appropriate time and place to be useful for us.

The third pic is the second with the addition of a gear between them. Now their equal and opposite torques are nullled. You can see that the System CoM/CoG lies at the middle and slightly below the horizontal axle line. It has absolutely no tendencies to turn left of right, as their opposite and equal torques are locked in an arm wrestle that neither can win. It is torque balanced and sits there unmoving ! I've called it Torque Nulling but it could also be Torque Cancellation IINM.

N.B. in the sim you can turn the Gear ON and OFF with a 1 or 0 in the Input Box.

[Fletcher]

The next series of 3 pics and sims are the same dual systems connected by a Gear with nulled torque characteristics, IINM.

This time I drive the RHS wheel by 3 different means. The sims run very slowly. First with a velocity motor, the second with a rotational torque applied, and the third with the addition of an ordinary weight (red). This one would act like a pendulum.

The thing to note probably is that in the first two the System CoM/CoG migrates upwards and downwards at the middle station per sector. But all torques are nulled.

Once again the rlws are pinned until release time and place simulating what a chain might do. If I were to build this dual axle test bed I'd probably use gravity activated latches (bit hard) or electro-magnetic switching, or servo's and micro-controllers etc to efficiently time the release and catch. But I get ahead of myself.

The next post and pic will be the same dual axle setup etc with the addition of the Chain Prime Mover (the 3rd entity) I keep banging on about.