Blood From Stone

A Bessler, gravity, free-energy free-for-all. Registered users can upload files, conduct polls, and more...

Moderator: scott

Post Reply
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

Post by MrVibrating »

silent wrote:Since I joined this forum a little over a month ago, I've gone through several minor builds and have lived the wheel. I've done something involving the Bessler wheel every single day during this time. I've come to the conclusion like Bessler did that building over-balanced wheels the traditional way is trying to cheat gravity in a way it can't be cheated.

I've had a look at Sjack Abeling's wheel and consider it plausible, however when I saw what I throught looked like air tanks on the wheel in one of the pictures on his website, I decided to not get too excited over it.

I've been trying to think of novel ways to get a weight to travel upwards and tricks to maybe get the wheel to see a weight temporarily as no weight at all using relative speed tricks. I've been monkeying with some ideas involving a capstan and having a rope travel along the perimeter of the wheel. (I cringe due to all the friction involved) and then I've started to think about the hurdy-gurdy instrument and a rosin wheel to vibrate a string causing a standing wave to potentially cause a weight to fly upwards.

The wheel was 12 foot in diameter and rotated 26 rpm. That's approximately one full revolution every 2 seconds or roughly a mile every 5 minutes. Rotating on its own - that's moving. So how in the heck does a weight get in the wheel around 12:00, ride the outside down to 6:00 and then get drawn back up in-time to make the ride again? The more I think about it, *if* separate weights were involved in this manner, it almost seems like there had to be a reservoir of weights that might take several rotations of the wheel to make it back to the top, but this would end up in an eventual deficit because every time a weight takes the ride, another one needs to make it to the top so that throws that thought out of the window.

If the weights were on tethers affixed to levers, perhaps it would make a difference, but the problems of the deficit still remain.

I started experimenting yesterday by taking a ring and sliding 2 ropes through it, then pulling the ends apart. The ring climbs the rope. I've been watching Tim the toy man on youtube because many of his old toys capitalize on little mechanisms and trickery to accomplish their captivating feats.

I've been thinking more about clock mechanisms which completely baffles me. I asked my wife last night, "Don't they have self-winding clocks or weights that reset themselves?" I seem to recall reading something about them, but I can't say for sure. It involves clock works, weights on chains, etc. If you speed up a clock and make things happen faster, then perhaps the spinning wheel is like the hands of a clock, but on a much steeper ratio?

Couple all of this with the thought that if the wheel could do motive work at the axle, then either something was spinning rather fast and was geared down or we are back to weights at the periphery of a wheel or bar. Ah yes, the cross-bar mechanism...

No wonder people haven't been able to figure this out for over 300 years.

silent
It's tough innit? I like the vibrate-y stringy thingy (very Pythagorean) but sounds quite lossy too... A general point about 'pools' - be they a reservoir of weights, levers or whatever, is that Bessler was very explicit that in a "true PMM, everything must, of necessity, go around together - there can be nothing involved in it that remains stationary upon the axle"..

This stipulation would arguably not apply to any successful form of GPE exploit, nor any other kind of hypothetical exploit... besides an effective N3 break. If the name of the game is generating excess KE, then you need to accelerate the entire reference frame - which is what happens when a closed system of interacting masses has non-constant net momentum, which in turn causes KE gains / mechanical OU. It's the reason why statorless operation is so essential.. and that means 'statorless' - nothing can be left behind the angular acceleration of the net system, or else the net momentum cannot increase, and hence KE gains cannot be generated...
User avatar
Fletcher
Addict
Addict
Posts: 8424
Joined: Wed Nov 05, 2003 9:03 am
Location: NZ

re: Blood From Stone

Post by Fletcher »

FWIW .. I generally try to stay clear of elements in the sim that use or output as Power Mr V. Power IINM is a measure of Rate of doing Work so can fluctuate widely. Technically that shouldn't be a problem I think but since it is time dependent then the hertz, or frame, or iteration rate, comes into play for perhaps introducing inaccuracies.

What I can suggest is that you sidestep the sims you are building atm and just build some simple actuator push and/or pull test for example and try to get its power inputs and outputs to reconcile with other fundamental metrics. If they match up then you can be confident that your readings are working correctly in your advanced sims at whatever hertz you get the most reliable result from. IOW's eliminate a possible problem.

In the past I have had trouble analyzing out the Joules accurately, IIRC.

Force x Displacement integrals seem more reliable because they are not time dependent.

You might like to investigate dampened springs (small dampening for more realistic action) or swap out actuators for the 'generic point-to-point' element in 'tools' where you can muck about with the inputs.
User avatar
ME
Addict
Addict
Posts: 3512
Joined: Wed Jun 08, 2005 6:37 pm
Location: Netherlands

re: Blood From Stone

Post by ME »

[Prolly still in silent/ignore mode]
MrVibrating wrote:Anyone else able to take data - or better yet, independently measure this interaction - further input would be appreciated..

..happy to consider any reasonable ideas..
Tried earlier, wasn't appreciated. With a flexing "reasonable", it's hard to take this request too seriously.
Fletcher wrote:Force x Displacement integrals seem more reliable because they are not time dependent.
But they are.

"Force x displacement" is more reliable because that's how WM2D's Kutta-Merson integration works...
It solves Newton's 2nd !!
Hence that earlier claim that one needs to understand where the KE-formula (as a conversion-factor) comes from to appreciate what the simulator is actually doing.

The basic issue goes like this: When Acceleration (Force) depends on Position while Velocity is affected by Acceleration while by definition Velocity changes Position then we have a so-called undefined integral.
That's an interdependent parameter-puzzle with no direct solution and needs to be solved by step-wise refinement. And that's in short what any physics solver does for all the forces it encounters within the given accuracy, at the given time-step (= simulator-frequency).

I understand such technicality is of minor interest to most, but it is just some background-info how this nevertheless clearly demonstrates why "energy-creation" within (at least this) simulator is a non-issue.

[Voice in the desert - out]
Marchello E.
-- May the force lift you up. In case it doesn't, try something else.---
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

Post by MrVibrating »

I'm integrating power times time, so an energy term, rather than 'power' per se.

The freqs i'm running these results at are all very high - maxing out the memory on many of the preceding runs, even for a single cycle or part-cycle.

The instantaneous power is calculated say one thousand times a second at 1 kHz, and then integrated across each of those time stamps - so if the first cycle takes say 2.92 seconds then that's 2,920 discrete measurements of instantaneous power, then integrated across that period via the Riemann function using Excel. So it's the spreadsheet actually doing the energy calculation.

I've tested the P*t metric against F*d with good consistency in the past - better, even, hence why i've settled on it.

For the same reason, i've found it to be the best measure of centrifugal / centripetal work done, too:

• calculating mass times angular velocity squared times the changing radius gives instantaneous CF force

• multiply that by the changing radial velocity gives the rather cool-sounding quantity "centrifugal power"..

• ..and centrifugal power times time gives us the CF/CP work integral

So that's all being calculated on the fly in realtime, not 'sensed' directly from the mechanism; note that in all other cases these two metrics of work done sum to unity..

To within a millijoule or so..

I really can't emphasise that point enough; the only difference between a unity and non-unity result here (KE gain in one direction, KE loss in the other) is performing the inertial interaction in its entirety whilst the weight is still falling... as opposed to moving the masses back in once the weight's rising.

The latter condition measures unity using these same meters. No change to any of them since the beginning.

I'll do what you suggest anyway - can't be too thorough in this situation eh..

FWIW the power's being calculated in terms of force times velocity, rather than say volts times amps or J/s etc., so F*d is still dependent on the same force data, only integrated over displacement instead of velocity.

If i calculate the force independently from first principles - it's basically the above CF force, plus the inertia of the two ½ kg rotor masses - and then integrate that over the radial displacement - it's essentially just gonna mirror the current CF work integral - just with a time-squished curve. But if instead i use WM's internal calculation of force on the actuators, that's another degree of separation i guess, and a cinch to knock up..
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

re: Blood From Stone

Post by MrVibrating »

I'm gonna take Fletch's advice and start comparing the actuator F*d integrals to the other two, but for now, here's those last five-cycles again, this time running at the precise maximum possible frequency, so mapped across the entire memory range:

Image
No need for a pause function.. it simply runs out of memory (if this seems slow, bear in mind it's just a sped-up slideshow of the actual sim).



Act. P*t = 51.22797573 J

CF P*t = 117.5447938 J


So maxing the frequency has simply increased the 'gain'. That's gotta be suspect - you might expect to refine a few more digits but an extra 10 J is no small change..

Let's see what the new integral reports eh..
Attachments
MoI_Swapdrop_3d52.wm2d
(26.56 KiB) Downloaded 47 times
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

re: Blood From Stone

Post by MrVibrating »

Here's a single cycle test of an F*d plot:

Image
@ 1 kHz

F*d = 20.18972451

P*t = 23.47262698

CF work = 23.7045491




...slight teething trouble no doubt; i'm not sure it's zeroed correctly - you'd expect it to track more or less equal distance above and below some median line, so having an upper 'output' half vs a lower 'input' half, so summink don't seem right there..


If it was a torque times angle plot i'd just stick each on their own axis, but with an oscillating action we can't plot against actuator displacement (cos we'd be zigzagging on the x-axis), and likewise if we use radius instead (same measure anyway).

So we have to multiply F*d on one axis, and stick something else on the other - if that's time then we're plotting energy times time = power, tho, surely? It's late and me head's thumpin'..

Will suss it tomoz with any luck..
Attachments
MoI_Swapdrop_3e2.wm2d
(28.2 KiB) Downloaded 39 times
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

Post by MrVibrating »

I don't want to get anybody's hopes up - this is the peanut gallery, caveat emptor etc. etc. - that said, let's just summarise where we're at (because this seems really cool):

• the two metrics that appear to work consistently in all other instances of this interaction, only show the gain condition on this one - when the whole MoI variation takes place during the descent, rather than mixed between ascent and descent

• as seen again above in the most recent example, we can get pretty good agreement between them for a single cycle... yet multiple cycles appear significantly OU..

This is exactly how OU KE develops from an effective N3 break - the very first cycle is buying momentum at the static, ½mV² rate - you're buying precisely the right amount of momentum... just not of equal opposite sign. So its initial cost is a unity result...

..it's what happens after, as the process then repeats, adding more momentum sans counter-momentum, that the KE gains start to apply.

It's getting hard to think of an error that would only apply when the cycle repeats like this... some kind of speed-dependent glitch maybe, as Fletch alludes to.. but then you'd expect raising the time resolution would reduce the error, rather than increasing it..

Let's see how the F*d plots add up.. if we're still in la la land then it might be time to start analysing it one cycle at a time, to try tease out the gain trend and correlate it to momentum changes etc. - see if it's really implementing the 'maffs ov OU' type stuff..
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

Post by MrVibrating »

..given the 2m radius, maybe the x axis should be zeroed at 1m - then we're plotting increasing force above it, and decreasing below it? No that can't be right.. (see, this is why i use P*t)...
User avatar
WaltzCee
Addict
Addict
Posts: 3361
Joined: Sun Dec 09, 2007 9:52 pm
Location: Huntsville, TX
Contact:

re: Blood From Stone

Post by WaltzCee »

[Prolly still in silent/ignore mode]

Some people are taking it all in. Appreciate your efforts. Early on in
the discussion someone made the point:
Yup, smooth mathematical curves in weight motions are never going
to work - this is the very definition of mathematical symmetry.
I think attempting to find a potential difference between various
equations is pissing up a wet rope but it isn't my rope. The discussion
is interesting though with lots of cool pictures. :)
........................¯\_(ツ)_/¯
¯\_(ツ)_/¯ the future is here ¯\_(ツ)_/¯
Advocate of God Almighty, maker of heaven and earth and redeemer of my soul.
Walter Clarkson
© 2023 Walter W. Clarkson, LLC
All rights reserved. Do not even quote me w/o my expressed written consent.
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

Post by MrVibrating »

Will post a half-decent update after me fish supper, but exciting progress...

Fletch's idea's the real clincher now - F*d for the first cycle comes in at 23 J. Net KE's 26 J, so again, 3 J OU on the first cycle.

Like the man says, F*d is a pretty rock solid metric of work in mechanics - fat fingers aside, if this IS real then anyone can easily replicate with real rigs, or just pen and paper, and everything between.. very very simple maths..

It all comes down to the input CF work integral - the first 90° of rotation, where the masses move out from the center, reaching max radius as the weight reaches 9 o'clock horizontal..

..work is input to accelerate the masses against their own inertia...

..and then when they decelerate to a halt at max radius, that workload reverses; their radial KE is converted back to work done to, rather than by, the actuators. This much is just trivia - basic CoE.

But as they're moving out, CF is increasing, as the rotation continues to be accelerated by the falling weight..

..so additional work is being done against the actuators, by that CF..

..GPE is being converted into KE, but then back into PE again - whilst slowing the velocity and hence prolonging the amount of time spent gravitating and thus gaining momentum..

..when the weight reaches horizontal, and the MoI is at max radius, 7 J of PE have been collected from GPE.

16 J of KE have been collected from gravity.

16 + 7 = 23 J.

But net GPE output at this point is only 19 J.

So the system is actually comfortably OU after just 90° into the first cycle.

The extra energy has come from gravity.

Gravity has done more work - in terms of the net of PE and KE - than the transpired GMH.

Full details shortly..
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

Re: re: Blood From Stone

Post by MrVibrating »

WaltzCee wrote:[Prolly still in silent/ignore mode]

Some people are taking it all in. Appreciate your efforts. Early on in
the discussion someone made the point:
Yup, smooth mathematical curves in weight motions are never going
to work - this is the very definition of mathematical symmetry.
I think attempting to find a potential difference between various
equations is pissing up a wet rope but it isn't my rope. The discussion
is interesting though with lots of cool pictures. :)
Cheers mate - it ain't a naughty list, i just don't have time to get sidetracked.


But if OU cannot be formulated then it cannot exist. What is OU, but excess mechanical energy? Thus if it can exist, it can only be derived using the standard energy terms.

Anything else would be pixie dust..
silent
Aficionado
Aficionado
Posts: 803
Joined: Sat Oct 27, 2018 7:50 pm

Post by silent »

.
Last edited by silent on Mon Oct 04, 2021 2:53 pm, edited 1 time in total.
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

re: Blood From Stone

Post by MrVibrating »

Right, here's a back-to-basics version without the acceleration damping - the MoI radius is returned to a direct linear function of angle, hence radial velocity's a direct linear function of angular, spikes be damned.

For the F*d inputs, the instantaneous force acting on each of the two actuators is added together, then multiplied by the current radius.

This 'force times displacement' integral represents the work done by - or to - the actuators, hence their input and output energies.

Since we're already familiar with the other metrics, let's just focus on the new F*d plot in relation to net KE for now..

..specifically, we're going to be looking at the input energy, versus the output energy.


So, with the sim running at max frequency, here's the first 90°:

Image
Monster GIF, but worth it.

First off let's take the GPE:

• Gravity's 9.80665 m/s²

• Radius - and thus height, here - is 2 meters

• The weight is 1 kg

So output GPE thus far is 19.61330 Joules.

Now let's look at the system energy:

• Net KE is 16.42415 J

• Actuator F*d is 7.59980 J


..that's output work there; positive displacement along the CF force vector.


16.42415 + 7.59980 = 24.02395 J of work has been performed on the system, by gravity and CF force.

From 19.61330 J of G*M*H.

The vMoI masses are stationary at this point, at max radius - all their outbound radial KE has been decelerated back into PE - so the only 'cost' in terms of displacement along a force vector is negative CF output work...


So this is what OU looks like folks.. the sum of KE and PE from a given period of gravitation is substantially greater than the static GPE:

24.02395 / 19.61330 = 1.22488x more, to be precise.

So we're well over 20% OU before the weight's even halfway down. Something of an improvement on having to wait three full cycles, eh? ;P


From hereon, the rest of the cycle's fairly pedestrian:


At 180°:

Image


• KE = 65.50661 J

• F*d = 23.65789 J (input work this time)

• GPE = 39.2266 J

65.50661 - 39.2266 = 26.28001 J should remain after re-lifting the weight:

Image


...and we have that, pretty much bang-on.

So, in the final tally (and this is just for the first cycle remember - wait till we get up to speed), we have:

23.65789 - 26.28065 = -2.62276 J excess energy.

On the way up, we only had to input 39.2266 J of GPE.

But on the way down, over that same height, gravity output 39.2266 + 2.62276 = 41.84936 J.


we have an effective GPE asymmetry.



Well bugger me sideways..
Attachments
MSD3e3.zip
(991.22 KiB) Downloaded 58 times
MrVibrating
Addict
Addict
Posts: 2879
Joined: Sat Jul 31, 2010 12:19 am
Location: W3

Post by MrVibrating »

silent wrote:Fletch is usually on the mark. I've been going back and reading old posts and he's one of the old-timers on this list. It's nice when we all try to help each other instead of the opposite (and I've read plenty of that too.)

Thanks for sharing!

silent
Yes his input's been invaluable over the years, plus he never gets snarky (rare 'round here).
silent
Aficionado
Aficionado
Posts: 803
Joined: Sat Oct 27, 2018 7:50 pm

Post by silent »

.
Last edited by silent on Mon Oct 04, 2021 2:53 pm, edited 1 time in total.
Post Reply