Blood From Stone

[MrVibrating]

..goin' work now, but last experiments i did last night were to try changing the orbital MoI between hi/lo extremes...

Interesting results - with a very heavy oMoI, kJ of input energy get wasted to heat, and net KE is only 20 J...

..but with minimal oMoI (for the current config), oRKE approaches unity..

[MrVibrating]

So.. no one else noticed yet that it also appears to be a rudimentary propulsion source..?

We could replace gravity with centrifugal G-force, and pull momentum from that instead..

[ME]

Hmm. Ok.
Thanks guys, I would have suggested that middle smiley somewhere below my name…

To all inspired people, don't be stopped by my poking.
The gains claimed were made early, so please convince yourself and DO recreate those things in the shop, no matter how crude you make them!
–though: safety first; prepare for disappointment, second--

The requested collaboration could have kick-started by simply explaining why we should NOT use a spring and have to KEEP using those actuators. I understand the charms of investigating the effects of common physics in a simulator and then puzzling with its maths– I have seriously no problem with that, until:

-- Each and every time this question pops up: >>Where do the gains come from?
-- Each and every time the answer is: >> From motors and actuators!
If pointing that out is trolling, so be it.
-- Other questions arise: >> How else create momentum from gravity? Answer: >> Kick it off the roof....! (lift it first)
-- >> Are there better ways to get rotation from energy >> Use a motor as-is. I would try a brushless DC, a pc-fan for instance. Without any load or friction or blades to move air, it would theoretically keep accelerating (which depends on current)– but reaching its max (because of resistance) after a handful rotation (Who knows how it ramps up?)

Really, how many mixed signals can we handle without hints on the seriousness, or the practical viability of this all?

Anyway, the actual collaboration attempt is with the rest of this forum who seriously try to discover exactly THAT without the fog, assumption, and hidden engines.
I just CONFIRM, to those who are not entirely sure what to make of all this “language of overunity� flooding, that the gains come from simulated motors, and is indeed nonsense in the sense of Perpetual Motion despite the exclamations of “momentum gain� and “overunity� factors.
That’s all.
MrV tells us indeed the same thing.. only silently in the corner but over-eureka-ed by other stuff.
We simply have another area on this forum where deceptions and misrepresentations belong. It’s not my depth where this drama is getting dragged into. I personally don’t care, I did the same simulation stuff as mrV.

So as a simulation exercise, mrV’s stuff is great. But it’s nothing more than that.
In the end we are all (even MrV, I'm convinced!) attempting to investigate the possibilities of a mechanical overunity system as Bessler allegedly created.

"it" in question is OU,

Perhaps I'm looking at “overunity� in the wrong dictionary...
Could someone please share Mr’Vibrating’s definition of "OU" with all of us?
Perhaps supported by a (for the V-knows-how-much-repeated times) summarize in about two sentences how it will be managed?

----
Add. Here's my exact point as a form of TL;DR;
----

Do you think it's worth attempting a build? If so, do you have in mind the mechanism that causes the red weights to move in and out as the wheel spins?

silent

Much too early to waste a moment on any of my BS mate.

Agreed.
V doesn't seem to realize that people take his stuff for serious research, instead of fiddling around a simulator.

As for powering these radial translations - and converting the KE gains back into sprung PE or whatever to cycle them - i currently have no clue; these are engineering issues

No, of course not. Those engineers will refuse for sure, or give up very soon!!

As mentioned, this "engineering issue" is exactly OUR PROBLEM to tackle!!

----
Add.
----
Sorry, that I may have drowned mrV's request (you now know how to block me).
Here it is again, please respond - I have no idea.

So.. no one else noticed yet that it also appears to be a rudimentary propulsion source..?

We could replace gravity with centrifugal G-force, and pull momentum from that instead..

[thx4]

Marcello is a pragmatist and this is essential in this kind of research.
Mvibrating is pursued by his dream that drives him, even if the main thing is elsewhere.
From there to understanding everything!!!!, but what a pleasure to read you:)

[MrVibrating]

Marchello doesn't understand the difference between momentum and KE and this is essential in this kind of research.

He can't read energy terms, and thinks an experiment specifically designed to plot force * velocity * time and power * time is all very well, but how are we supposed to know how much energy it's using?

He's the 'not even wrong' contingent, yet thinks everyone else is beneath him - but how are you supposed to argue with someone who can't even grasp CoM? He thinks he can alter a system's net momentum with a motor. "Counter momentum" is just some buzzword. Good luck to him, but i have no interest in his thoughts on anything.

[MrVibrating]

Slight improvement - 172 in for 72 out. Still shite, but more of a departure between orbital and axial RKE:



The momentum yeild's still decreasing with orbital velocity - maybe this is an inevitable intrinsic limit, and also governed the speed of B's wheels; since the faster the orbit gets, the less time the weight spends gravitating in its axis.. so it's more powerful under load, since this maintains a better momentum yield over time. This would seem consistent with B's claims his systems gained further advantage from attached loads..

..another issue tho could be that the two axes are counter-rotating - if instead of using the motor as a clutch to brake the speed difference away, we actually applied positive torque with it, across the same small angle each cycle, we'd find that we were inputting less momentum per cycle for the same given input energy per cycle, because of that ever-accelerating counter-rotation between the two axes, incurring the ½mV² penalty..

..so even tho we're using the motor as a passive brake, perhaps we have the same issue, only in reverse - the faster this counter-rotation gets between the two axes, the less momentum gets transferred per collision. That's what i'm worried about anyway.. ideally, we wanna maintain the best momentum yield possible for as high a speed as possible..

So i'm doubtful any variation of the current scheme is likely to work.. dunno.


I did try making orbital MoI really low, but when it's too light the orbital axis just wobbles around, even as net momentum builds up. So it has to have some appreciable MoI just to carry any kind of consistent momentum.

..but then, the more MoI, the lower its speed and thus KE for a given momentum.. conflicting requirements.. is there a sweet spot? Increase the collision elasticity? Maybe the motor's a bad way of equalising the speeds, and actual collisions would be better?

Most likely i'm going about implementing it wrong. Bessler's wheels were, by all accounts, heavy, yet if they were banking momentum gains into that high MoI, it's KE value would've been tiny, and far below the energy losses from the collisions..

So, then... what else might you do with a cyclical momentum gain, in order to turn a KE or PE / GPE profit on it?

Was it the main rotating body of B's wheels that were being imbued with the excess KE in the first instance, or was it instead being instantly converted to GPE somehow? Or some other internal KE rise that then loads a spring..

The fastest route to OU i can see is this "two equal inertias and a reactionless acceleration / inelastic collision" sequence, for 50% net loss on the first cycle, unity on the second and 150% on the third.

However if you just accelerated both inertias up to that same speed they'd have after two cycles - the unity threshold - using any conventional means of acceleration, and then perform a single reactionless acceleration & inelastic collision, you instantly get the 150% result in a single stroke..

Is there a way we can implement this trick with real inertias?

[MrVibrating]

Yet another angle of attack - something else you can do with a rise in system momentum - is simply use it as a launchpad for some other, conventional form of acceleration.

Basically, a simple spring-loaded mechanism, whilst being subjected to a reactionless acceleration, should be able to output more KE than the PE it was loaded with, equal to the square of the velocity component of the reactionless momentum rise..

Yet all these tricks are useless, unless you can reduce the cost of rendering the effective N3 break in the first place. It looks like the most recent sim is doing that - the axial CF and RKE is no longer crescendoing upwards with the orbital momentum and KE... but the loss/gain ratios are still on the lossy side..

[MrVibrating]

Well, I'm going to take a punt and say he would've understood about circular (orbital) MOI relationships. For starters he would have observed a rolling hoop being tapped along, or maybe had seen the disk, ring, and sphere experiments race down a slope when all the same mass, and noted the winner. So he would understand the importance of mass displacement in a wheel design to energy input and GPE release.

Where I probably diverge from you is how he might have got to seriously investigating mechanical MOI changes (assuming this is what he did). I think he would have noticed an aberration in an OOB design or concept he was testing. Something that could be of use to cause accumulated wheel momentum, but not caused by sought-after asymmetric torque. And that was the uh-huh moment.

I say this because many of us have played around with forcibly changing radius of equidistant weights and also using CF's etc. No one successfully AFAIK.

And he would have recognized the importance of a semi-open system rather than the bog standard closed system.

All-said-and-done he attributed his insight to a dream from God, but he would wouldn't he. As-an-aside the number 72 (internal angle of the 5 sided pentagon) is important in the Kabbalah, representing the 72 names for God.

I expect he would've tried using inertial torques to accomplish lifts - this still firmly in the realm of chasing a GPE asymmetry..

..equally, he may have been investigating 'speed-dependence', perhaps, as a general topic rather than with a focus on energy or momentum per se - if he were unaware of the real distinction, he might've found a momentum gain there, where anyone today takes for granted that GPE's not speed-dependent so wouldn't even consider investigating further..

Still, i have this quaint notion that you could maybe get there just by thinking in terms of 'motion' itself, what sustains or increases it, and everything else seemingly falling under his definition of "friction".. so in essence, he may simply have been seeking "anti-friction" or "positive friction" - in which case, you have just two such examples to play with; OB torque, and inertial torque. And again, since both are closed systems in and of themselves, combining their effects and studying their inter-reactions is about the only place you can even start looking for interesting non-linearities or whatever..

I dunno mate... but i'll take the '72' thing as a good omen on my last set of sh!t results.. (172 in for 72 out)

[silent]

.

[MrVibrating]

The optimal specs for everything are determined by the maths of OU:

• if you think about it, there's what we might regard as a 'Jacob's ladder' sequence to the energy gain principle

• if the collision wipes out half of the per-cycle input energy, but conserves all of the momentum gained, then net efficiency climbs by 50% per cycle - we hit unity at two cycles, and OU at three (150%). Then 200% after 4, 250% @ 5 etc. etc.

• if the collision wipes out 75% of our input energy, then that efficiency accumulator drops to 25% per cycle, and we hit unity after four cycles, and OU at the fifth (125%). Then 150% at 6, 175% @ 7, 200% @ 8 etc. etc.

• so you can see the pattern - whatever the momentum yield (how much input energy we pay for how much new momentum), provided it is constant per cycle, then net efficiency increases by that amount per cycle - so if we're only 10% efficient (wasting 90% of input energy per cycle) then it takes 10 cycles to reach unity, and 11 to make a 10% excess.

So these are the design criteria for OU, laid down by nature.

If all you want is the blueprints for a design that already encompasses the above considerations, then you'll have to wait a bit longer because we're not there yet - none of rigs here work yet. They are not 'climbing higher on Jacobs ladder', and this is most likely because of crucial design failures.

However beguiling the motions may appear, their input energy / momentum yields are not constant enough across a sufficient range of speed, yet, to produce anything other than a net loss of energy.


That said, they are beguiling insofar as gaining momentum each cycle. Again, CoM is not supposed to be energy-dependent.

If this angular momentum we're generating can be converted to linear... whooosh! Who cares how much energy it costs in that case, right? So the CoM-related implications and potential prospects are indeed fascinating.

Turns out i inadvertently have a day or two off work (bike's clutch went, waiting for new plates) so i intended to get to the bottom of this.

I need to slow it right down and carefully plod through one cycle at a time, noting the momentum gained from gravity each cycle, its energy cost, and its distribution and value post-collision. Then see how this develops over successive cycles..

..again, the conditions for OU are pretty simple - it's just a matter of achieving that constancy of momentum yield across some useful speed range.

The current rigs are firmly on 'Jacob's ladder'... they're just sinking faster than they can climb. So we're doing the biz - we're in the game - we're just currently losing, is all.

Study the ladder, understand how and why it works, then you can think about how to tap in mechanically..



ETA: - just to clarify - the energy cost of the momentum we're buying includes both its KE value post-collision, and the energy dissipated by it. So for example to obtain ½ J per kg-m²-rad/s efficiency, we're inevitably going to be wasting another ½ J per cycle to heat & noise, so the optimal momentum yield is actually 1 J / kg-m²-rad/s - only 50% efficiency - yet if this can be maintained across a speed range encompassing just three consecutive cycles, this is the quickest route to 150% OU.

[MrVibrating]

..just out of curiosity, is that hypothetical 50% inefficient heater-motor really the fastest route to OU?

What conditions would have to be met for OU in just two cycles, for instance?

- this would mean that ½ J bought and payed for 1 kg-m²-rad/s, post collision... this would be the rock-bottom energy-price-of-momentum, per ½mV², except it basically demands that we eliminate the inelastic collision

- yet the purpose of the collision is to equalise the velocities, so that the two inertias are once again stationary relative to one another, whilst moving uniformly relative to the stationary reference frame. Whatever the momentum yield, this stroke is essential for resetting it back to its nominal value..

- to put it another way, if the collision were just partially elastic, conserving more than 50% of the input energy, the two inertias would end up at different speeds post-collision, and that speed difference would accumulate over successive cycles, causing the input energy to likewise scale up as a function of that velocity squared...

- thus a 50% KE loss per cycle is the most efficient means of repeatedly resetting the ½mV² accumulator back to its starting value for each successive cycle, since the loss ratio is constant per cycle, invariant of net speed - a linear rise in costs - for a gain ratio that squares with rising velocity.

This is the root of the 'loss causes gain' dynamic.



However, the key aspect here is that it requires two equal interacting inertias. They could be angular, linear or a combination - angular-angular seems most practical, as the axial distance between them never changes.

But if they're not equal, then we lose that peak efficiency, thus raising the number of cycles to reach and exceed unity.

So for example if the ratio of the inertias is 1:2, we lose two-thirds of our input energy per cycle, so net efficiency only climbs 33% per cycle.

If the ratio is 1:3 (eg: 1 kg @ 1 m/s colliding inelastically with a 3 kg 'static' inertia), we lose three-quarters of our input energy - so for example that acceleration would cost ½ J, but leaving just 125 mJ post-collision. Net efficiency then climbs by 25% per cycle.

If the ratio is 1:4 (yes i'm running quick sims here to check all of this - i'd screw it right up otherwise) - so 1 kg @ 1 m/s landing on a static 4 kg inertia - we now have one fifth of our input energy left as KE post-collision (so 0.5 J in, 100 mJ out). Maintaining this ratio, net efficiency now climbs by 20% per cycle.

These are the maths of OU.

There really isn't that much to it, is there?

If you can grasp the simple dynamics unfolding here, then you're ready to start 'filling in' these parameters with real inertias and velocities.. lifting some kind of practical design from the gain principles, with a clear-headed objective of how much KE should be left after how much input energy has been spent each cycle.. IOW you'll be in the same boat as me, with every advantage over my crappy design skills and disarrayed thinking..

[Georg Künstler]

Hi MrVibrating,
you wrote:

to put it another way, if the collision were just partially elastic, conserving more than 50% of the input energy, the two inertias would end up at different speeds post-collision, and that speed difference would accumulate over successive cycles, causing the input energy to likewise scale up as a function of that velocity squared...

As I have described it many times, i use an indirect impact, this is the same as you described it as an partially elastic collision. So different words for the same function.

From engineering point, this function already exists and is already build.
This partially elastic function allows a weight shot up against gravity.
Like a pole vaulter.
So it fulfills also an second condition to be faster than the gravity acceleration.

But good insights so far.

[MrVibrating]

Enclosed is a cleaned-up version of the 2.5 rig - using a coaxial second rotor to consolidate the momentum gains.

The original 2.5 sim contained errors that were clouding the results.

In this revised version, the sim pauses each time the weight returns to TDC, the moment the clutch engages.

Clicking backwards a few frames will thus show the momenta and energies immediately prior to the collision, and forwarding a few frames shows them immediately afterwards.

The results show that we start out firmly 'on the ladder':

• the first collision equalises the velocities, halving the net energy

This is, on the one hand, perfectly trivial, yet it is also nothing less than the first rung on the OU ladder..

All we need to do to break CoE is maintain these conditions - as closely as possible (ain't gotta be spot-on) - for another two cycles..

..the more successfully we can sustain that initial efficiency, the closer we'll be to 150% at the third cycle.

Obviously, we can afford to slip a little - 110% @ 3 cycs would still be fantastic.

However what this sim also makes clear is that it's incapable of doing this so long as the second rotor is coaxial, which is causing a confluence of two further dynamics:

• being coaxial, their intrinsic angular velocities must rise together; this causes a reduction in the 'time-spent gravitating' on both up and down sides of the interaction, reducing the net momentum yield per-cycle

• at the same time, CF/CP forces are squaring with rising angular velocity, progressively raising the per-cycle cost of the ever-diminishing per-cycle momentum yield


And so now to apply the same rigour to collisions with an orbital MoI..

[MrVibrating]

..the difficulty with this orbital MoI lark, from experiments so far, is that the orbital MoI is comprised of the rest-mass of the axial rotor (the weight plus vMoI masses) - as i was saying to Fletch, this causes flaky results, since that orbital MoI is what we want to dump our axial momentum gains into.

Put simply, we'd ideally want to collide it into a rock-solid stable MoI.

Instead, the MoI we're colliding with is varying with the axial angle of the weight - IOW, the oMoI isn't wholly composed of 'rest mass', since half of it is swinging around in vertical loops..

..and this is with just one axial rotor, counter-balanced in its orbit by an inert deadweight... if that were a second axial rotor working in reciprocation with the first, then the oMoI is gonna be even more twitchy..

The way to mitigate this would seem to be to use as wide a radius as possible for the second rotor. This helps isolate the relative angular velocities - with more time per orbit, the axial mechanisms aren't so rushed, and could presumably complete more than one axial cycle per orbit...

..which would also help with the fact that as orbital radius increases (for aforementioned stability purposes), so does orbital MoI (mass times radius squared)..

..because if you refer back to the OU ladder, the quickest route up it is to use two equal inertias / MoI's.

Yet because MoI=mr², and the orbital MoI is comprised of the rest-mass of the axial rotor/s, the only way to have an oMoI equal to the aMoI is to halve the former's radius in relation to the latter...

So with the axial radius at two meters, we'd need to use an orbital radius of just one meter... this would allow us to have two counter-balancing axial rotors on either side of the orbit, all three rotors (two aMoI's plus the one oMoI) having an identical MoI.

So it's mathematically do-able...

..except we run into this stability issue - with such a light oMoI, when the clutch is applied to distribute the momentum gains, the orbital velocity jerks, which will jerk a second axial rotor or any others attached..

It may still be feasible, perhaps with very careful design and implementation...

But the easier workaround would seem to be to go for a higher-radius oMoI - this means we'll be colliding the faster, lighter aMoI with a slower, more lumbering oMoI, which knocks us right down the OU ladder - remember, as the ratio of the two inertias diverges from 1:1, so the number of successive cycles to unity and beyond increases..

..hence, if we only had one axial cycle per orbital cycle, if the two MoI's are in a 1:4 ratio, we're gonna need ten full orbital cycles before reaching break-even, and eleven+ to make energy..

..but a wider orbital radius also gives more of a window for further axial cycles per orbit!

So if we had say two axial rotors, and each completed multiple cycles per orbit, we could be OU much quicker.

With enough axial rotors, we could be OU before 180° of orbital angle has elapsed..

Gonna make a start now on a nice big oMoI and see how things pan out..

[MrVibrating]

...first, here's a quick demonstration of this orbital MoI instability issue:



..as you can see, the orbital velocity - the component of the orbital momentum that we most need to accumulate - is oscillating.

Worse, it's at its minima the very moment the clutches engage, hence transferring their momentum gains to an MoI that is decelerating rather than coasting or accelerating..

..the net result of this is that we're spending energy performing negative work, decelerating and accelerating the orbit in turn, gradually raising the magnitude of the oscillation, but wasting far more energy than the KE value of the banked momentum.

This is further exacerbated by the fact that the rising angular momentum is manifesting on the axial rotors, as much as the orbital one, thus raising CF/CP workloads and decreasing per-cycle momentum yields.

In conclusion, trying to minimise or match the orbital MoI in relation to the axial MoI's appears to be a false economy.

So next i'll try a much wider radius for the orbital axis, in the hope of stabilising the relative orbital / axial velocities..