Blood From Stone

[Sam Peppiatt]

Mr. "V",

Now I know what you mean; it's a great argument! I've read it before meany times before.
But, the good thing about a gravity wheel, is; it can't read. So; it doesn't know it can't work!

Sam Peppiatt

[raj]

AI Gravity Wheel???

Would read and work as his programmer has instructed???

[MrVibrating]

As usual my thinking was a little fuzzy there - turns out, if we're only to perform the MoI inversion on one side of the gravitational interaction alone, then there are of course two variations on this to be tried; moving the masses out then back in whilst dropping, versus moving them in then back out whilst rising..


..so, in that order:


(animation at 200 Hz for speed, figs below from 1 kHz for accuracy)

starting energy = ¼ J

in = 373.7 J

out = 353.91148 J, of which 119.89104 J is extant radial KE, and the remaining 234.02044 J is rotational KE.

353.91148 / 373.7 = 94% efficiency


So the efficiency is close to unity... but short. Why is it short? There are no collisions or friction or any conventional dissipative loss mechanisms here - just a GPE interaction, coupled with an inertial interaction, in an otherwise closed system..


Interesting eh? Could this be a genu-ine non-dissipative loss mechanism? Because we all know by now what a time-reversible non-dissipative loss mechanism does, right? Exactly. That. So let's try the inverse interaction, and see if we can't maybe flip our fortunes there:


(animation at 200 Hz for speed, figs below from 1 kHz for accuracy)

starting energy = ¼ J

in = 365.3367 J

out = 418.8074 J

418.8074 / 365.3367 = 114% efficiency


..so that is jolly nice, if real.



So, time to go back into debugging mode... let's raise accuracies, start also taking the CF work integrals for side-by-side comparison with momentum changes, and cross-checking instantaneous energies with manual calcs. As ever, if it turns out not to be in error then we'll also be zeroing in on the reasons how and why.. negative work by gravity on the actuators remaining a cautious hypothesis only until we can't prove otherwise.

[MrVibrating]

MrV.

Thank you.

So !!! Talking about 'TORQUE" is not Physics talk?
Talking about "TORQUE" is Magical Thinking!

WOW! That's mind-boggling!

Please allow me to ask you another simple clarifying question.
Is Besslerwheel.com forum Besslerwheel PMM forum OR an OU forum???
Kindly do not remind me of Magical thinking. Please!

MrV.
HAIL!!! You are the WISEST OF ALL WISES.

Raj

Torque is just angular force, it's totally meaningless in relation to the search for a PMM as you'd put it, outside of the context of the angle through which it has been applied. Torque times angle is work / energy. Then you could compare that value, with the other force times displacement causing it, of pulling mass in radially against CF force.

That would be 'physics talk'.

Simply insisting that there is torque at some preferential moment is just a useless statement of the obvious.

And yes, if we're willfully disinterested in ever dealing with the actual mechanics of motion then all we have is naive, wishful thinking..

There's real science here, and no good reason for this subject not to be taken more seriously.. but for the stigma of association with the peanut gallery. The scientific method's free to use, we can lift ourselves out of the fog anytime, it's an entirely self-imposed status quo..

[MrVibrating]

Okay Mr.V after thinking about ALL this stuff - your diagrams, Bessler's diagrams, and putting this all together, I slept on it and during REM sleep I came up with diagram.

Now if you focus on the toys of MT138, and people talking about hearing blows on the side of the wheel (I think 8 per revolution), a cross bar that spans across the entire diameter, oscillation, resonance, and put them all together...what might you come up with?

Thinking of Bessler's poem where he talks about an anvil receiving many blows...it is assumed that the little characters are striking anvils in MT138. I thought what if these weren't anvils? What if they are drums? What does a drum stick do when you smack a drum? It recoils and bounces off. So if in this diagram of Bessler's...if the little characters were hitting drums, it would make the cross arms bounce back and forth - given they were given enough force to start with.

So this morning I drew up this diagram that kind of puts it all together. I think perhaps the real magic is once it is spun and the forces are acting together, perhaps an oscillation will kick in. Perhaps 2 or more of these mechanisms tied in with each other would help itself out.

So A, B, C, D are all balanced. When started with an oscillation, A and B fly out and bring C and D in. On the outer periphery is a spring so that when the weights impact, they will be thrust back in (balanced system) and the weights will trade off their positions, acting in pairs. I don't know if they need interior springs so that they don't clack against each other in the middle, but the theory is that C & D will be thrust out and A & B will be brought back in.

Maybe this is old hat, but perhaps Bessler's apparatus was something similar to this?

One other thing that comes to mind is that MT138 drummers have different sized hammers so perhaps one set of weights needs to be lighter than the other to cause an oscillation. In my mind it would make the larger weights want to fly out harder, hit the springs harder, fly back but not all the way and fly back out again. If A & B were left larger and C & D left smaller then change the length of the connecting arms and hook a spring between the arms? I dunno - maybe I'm over-thinking it. (I added my smaller, unlabeled, quickie sketch with the 2nd idea)

silent

They're possibly useful mechanisms mate, i don't really know what's what ATM... need to try make the current gain go away, before wasting time working out how to 'harness' an error..

..as for the toys page - i meant to reference MT 133 / 134 (my bad) - these seem to add some context to the hammer toys - tho are equally mysterious in their own right..

[MrVibrating]

I further was reading last night to see if gravity has a frequency. It's highly debated - 30 to 150hz said one source - another as high as 300hz. Regardless, in this case I have my doubts that it is even anything worth messing with. If Bessler was a clock-maker and had an affinity and familiarity with pendulums, then a mechanism similar to this seems to fit the bill. I think you wouldn't know what it was going to do for sure until you spun it and got it in it's sweet spot.

Now if Bessler was doing a wheel within a wheel, perhaps this mechanism was inside the smaller wheel at which point I'd say the bigger wheel was 1.6 times the size of the inner.

Also, with my 2nd sketch, the larger weights instead of just going straight out, now take a slightly curved path to the outside of the wheel and back.

silent

Gravity has no frequency at classical / human scales (it's the epitome of a 'static field')- if it can be resolved to a fundamental carrier particle (a gauge boson like the hypothetical graviton), then its energy would be referenced in terms of its wavelength at the Planck scale divided by the speed of light, so beyond the tHz range (and far below what constitutes "variation" at our scales - it's still effectively constant).

And because gravity's constant, and mass is also constant, and distance is constant (1 meter is always 1 meter), and gravitational potential energy is simply those three things multiplied together, closed-loop trajectories through static fields yield zero net energy, so looking for a potential gravitational asymmetry is futile, and any notion of such a possibility must be a mirage..

The rate of change of momentum under gravitation is a time-dependent constant, but still a constant at human scales, and not really a 'frequency' in terms of a discrete periodic function.


As for optimal trajectories, that's still a WIP; it's one think to think up interesting interactions to measure, but the real work is getting the measurements..

[MrVibrating]

Here's the gain config again at significantly higher accuracy:

Sim: 'MSD3a' attached (no need for a new anim), incl. Riemann sums.

Raised integrator accuracy from 7 sigfigs up to 16.

Doubled frequency from 1 kHz up to 2 kHz.

starting energy = ¼ J

in = 378.7061 J + 0.25 J = 378.9561 J

out = 418.6985 J

418.6985 / 378.9561 = 1.1048, so 110.5 % OU.

So we've successfully got rid of 4 % of it, but only at the expense of reducing the uncertainty on the remainder by nine orders of magnitude..

Let's take a look at the CF work integral:

418.5473 J

So this would seem to preclude any 'magic' having occurred - all of the rise in RKE has been due to work done against CF force.

However, this appears to be persistently 10 % more work, than the input energy that has been supplied to the system via the actuators.

This obviously seems consistent with the hypothesis of conversion of GKE into PE in the actuators, as the MoI widens whilst under angular acceleration by gravity - basically 'loading a spring', - so supplementing the energy required to push them back in again.

Since this sim is the previous 'dt2.4' version re-purposed, the radial KE's are already double checked; everything here is metered correctly, producing consistent results for all other measurements.

That is, under all other conditions, the CF/CP work integral is precisely equal to the sum of actuator power times time, minus any remnant radial KE, and at perfect unity with the net system change in KE.

So, 'free work by gravity' - in the form of subsidised angular momentum - is now firmly in the frame.

We should be able to verify this by slowing things right down, and going through the interaction one cycle at a time, logging momenta and energies as we go.

We should thus be able to decisively disentangle the distributions of gravitational potential energy into KE and PE, and then re-trace the provenance and accumulation of this apparent advantage..

[Fletcher]

Mr V .. you've already covered it wrt actuators but I'll mention again that any time two masses are flung apart or pulled together an acceleration and deceleration of those masses is required in the same leg. That means a force is applied. And an applied force x displacement is energy consumption which has to be budgeted somewhere.

[ME]

Don't underestimate V's own good insights on his own simulated stuff:

if we're willfully disinterested in ever dealing with the actual mechanics of motion then all we have is naive, wishful thinking..

And the bad ones:

There's real science here, and no good reason for this subject not to be taken more seriously..

Totally caused by not following his own advice... with ugly results:

The scientific method's free to use, we can lift ourselves out of the fog anytime, it's an entirely self-imposed status quo..

(Note: I'm on V's ignore-list for a reason :-)

In the mean time I just keep wondering when V is finally able to derive that kinetic energy equation himself; I'm totally convinced he should be capable, but I just wonder what it gets spun into this time.
Among other assumptions, it should able to directly show the nonsense of his Momentum-'economics'... unless it's part of this "magical thinking"?

[MrVibrating]

So here's the very first cycle, with the same 25 mJ push-start:



..the key to whatever's going on across n cycles is going to be locked up in the details of precisely what's going on in these here 2.919 seconds above.

Since nothing much at all happens in the second half, besides the weight coasting upwards, what really interests us is that first half..

..It's obviously divided into two 90° phases - an output integral as the masses slide outwards under CF force, immediately followed by an input integral as the masses move back in whilst the weight continues from 9 o' clock down to 6 o' clock BDC.

So, taking those two sets of integrals for both input energy and CF work, should reveal how much work gravity did, versus how much was left to us..

The above single cycle's still being simmed at 2 kHz, so we can afford to raise this considerably. Given the circumstances, it would seem reasonable to devote the maximum possible resources to data accuracy, at the expense of running time, so each 90° phase will be simmed at ludicrous frequency. Forget about animations, they'd be 30 mins each..


Results to follow..

[MrVibrating]

Mr V .. you've already covered it wrt actuators but I'll mention again that any time two masses are flung apart or pulled together an acceleration and deceleration of those masses is required in the same leg. That means a force is applied. And an applied force x displacement is energy consumption which has to be budgeted somewhere.

It's metered mate under "Radial ½mV²".


This is obviously already included in WM's own 'kinetic()' calculation, so calculating it independently just allows us to subtract it from that 'net KE' quotient, revealing a remainder precisely equal to the RKE's as calculated from the spontaneous MoI's and angular velocities.

Obviously, it's only input work when accelerating them, but all that radial KE converts back to PE when the masses decelerate - a la 'regenerative braking' (but simply conservation of mechanical energy).. so the net input energy is simply equal to the net input CF work, as a function of the up vs down "time gravitating" asymmetry, from whence cometh the momentum-from-gravity..

In short, it looks like a validation of the hypothesis - we're converting some of the initial GPE into GKE, but also into essentially 'sprung PE', this latter quotient then subsidising the energy required to move the masses back in again. All the while, gravity's adding momentum to the system, so we're basically paying for half of it with its own GKE, and the other half with further input PE via the actuators.. apparently consolidating a time-dependent momentum asymmetry into an effective KE gain. It's friggin' working mate..

..but as ever, gotta assume it's nuts til proof's positive.. working on that right now..

All bases covered, i think..

[silent]

.

[cloud camper]

Mr V .. you've already covered it wrt actuators but I'll mention again that any time two masses are flung apart or pulled together an acceleration and deceleration of those masses is required in the same leg. That means a force is applied. And an applied force x displacement is energy consumption which has to be budgeted somewhere.

Fletch - you keep spoiling Mr V's unicorn spotting.

Mr V has seen his unicorn and is determined to keep it even if no one else sees it!

We have to stick to magical thinking here so any talk of powered actuators is inappropriate!

[MrVibrating]

Okay, one for the sim-heads here:

• on my system, WM will produce a maximum of 32765 simulation frames in a single simulation

• the initial 90° drop has a period of precisely 1.77325 seconds

• 32765 / 1.77325 = 18477.372056957563795291132102072 Hz max frequency

So the enclosed sim is using every. last. byte. of memory that WM can address in a single sim. We could of course break it down into 45° or smaller chunks over the same spread of memory, but we're already many orders above noise..



So, this gives us 32,765 vertical 'slices' of the areas, and thus energies, under the actuator and centrifugal work curves. These are then exported to a text file then copied and pasted into Excel, which calculates the 'Riemann sum' (the equation in column 'C') of the totals.

So first off, actuator power * time comes in at -3.33197 J.

As expected, it's negative - a net output of work, by gravity, effectively 'loaded into a spring'.

Net system KE at this point is 16.43038 J. We began with a 0.25 J push-start, so the net rise thus far is 3.33197 + 16.43038 - 0.25 = 19.51235 J.

The CF work integral at this stage is -3.33232 J - so good agreement with actuator P*t.

In terms of output GPE, 1 kg has dropped by 2 meters, hence taking gravity at 9.80665 m/s², G * m * h = 19.6133 J - so we're to within 1 mJ of unity at this stage. Most of the output GPE's converted to RKE, but a little has been loaded into a spring..

So now we can move on to look at what happens in the next 90°, and how that 3 J of sprung PE pays dividends as it unloads again..

Back after din dins..

[MrVibrating]

Mr V .. you've already covered it wrt actuators but I'll mention again that any time two masses are flung apart or pulled together an acceleration and deceleration of those masses is required in the same leg. That means a force is applied. And an applied force x displacement is energy consumption which has to be budgeted somewhere.

Fletch - you keep spoiling Mr V's unicorn spotting.

Mr V has seen his unicorn and is determined to keep it even if no one else sees it!

We have to stick to magical thinking here so any talk of powered actuators is inappropriate!

WTF you talking about you dickhead? On the 'ignore' list you go.. done wi' yuz.