Decoupling RKE from GPE, for fun and profit

[Fletcher]

That Rabbi must have taught him a thing or two ;)

" Why did God invent Christians ?

Somebody has to pay retail ! "

The money lender analogy, where the gold standard is gravity and the currency Joules In and Out.

You can borrow from the money lender but have to pay back the principal plus inflationary interest over time (the system energy losses) plus uberineterst (the margin) for the opportunity cost of borrowing it in the first place from a lender of last resort. The payback is you make way more than it cost you all up and that is your retail markup.

Just some pseudo economic ramblin nonsense ...

[MrVibrating]

LOL energy usury, whole new can o'worms there... if loopholes aren't technically illegal, are they ethical?


I really wanna do the double fusee experiment now. It's too simple to relegate to a sim, and would still need doing if the results were positive, so might as well cut straight to the chase..

A small desktop model in the 100mj to 500mj range. Use a laser tacho or hi-speed camera to measure RPM. Fusees would need hand crafting. They'll need a high diameter range - as large as the rotor, at least. 200mg drop weight falling 50cm =100mj, and the flywheel just needs to have a high MoI for its mass - small lead pieces attached to the perimeter of a polystyrene disk would seem sufficient.

Looks like quite intricate work crafting fusees. The geometry i think i can handle, but actually fabricating the things is 90% of the build here.. With well-planned designs and techniques they'll be a couple of weekend's worth of work alone.

Still trying to think of a statorless version too, which i'd probably wanna drop anything else to work on, so no great hurry. The dual-fusee approach seems the most elegant solution, tho. Having the drop weights inside the accelerating frame seems to necessitate complexity, perhaps even compromising efficiency - Bessler's wheels hit an RPM sweet spot, presumably a constraint of available drop time, meaning that the very form of the wheel's gain - RKE - was pitted against diminishing GPE as speed rose towards an equilibrium. So keeping the GPE load outside in the stationary frame eliminates any conflict of requirements - we can spin the wheel as fast as we like, while dropping a weight as slowly and as far as we like, gaining all the way.

FWIW i've searched online for off-the-shelf fusees, without success. Pity.. if they work as intended then in a few years RS Components may have a whole range of 'em... but right now they've none.

Bumped into an old phycisist buddy on the South Bank earlier and ran it by him. He was stumped. Initially insisted that inertia must be velocity-dependent, LOL... basically he couldn't fault it.

[Fletcher]

"Inertia must be velocity-dependent".

In my own thread for pulley systems and storksbills I concluded that if inertia was displacement dependent then it was by default related to Speed Ratio aka velocity, IINM.

https://en.wikipedia.org/wiki/Mechanical_advantage

Ideal Mechanical Advantage:

IMA = Force Out / Force In = Speed In / Speed Out

Let's hope that bastion is about to fall. The very best of luck, and I mean that sincerely.

[daanopperman]

Hi Mr Vibrating ,

If I had to make a Fusee . I would try to use a length of 5 or 6 mm Festo pneumatic hose , cut in half and pinned to a wooden cone with small nails , no waiting for glue to dry . The nailing would be to the larger dia of the cone , with the nail almost to the one side of the tubing . Once the tubing is cut and wrapped around the cone you will have a groove for your string .
Since you have a above average IQ you will be able to decypher my coded drawing .

[daanopperman]

To add ,

Just for you , Fusee MT 93 and 94 .

In Fusee 94 , you can actually " see " the drive mass , as the scoop injects the same mass all the time the " volume " seems to increase closer to the center of the fusee .

[MrVibrating]

"Inertia must be velocity-dependent".

In my own thread for pulley systems and storksbills I concluded that if inertia was displacement dependent then it was by default related to Speed Ratio aka velocity, IINM.

https://en.wikipedia.org/wiki/Mechanical_advantage

Ideal Mechanical Advantage:

IMA = Force Out / Force In = Speed In / Speed Out

Let's hope that bastion is about to fall. The very best of luck, and I mean that sincerely.

No you're absolutely correct, all else being equal, in practical terms inertia is velocity dependent... but only for incidental reasons, it seems.

The bottom line is that 1J can change the velocity of 1kg by 1m/s regardless of whatever the current velocity is (which is only to restate Newton's First), provided one can keep applying input energy from within the accelerating frame.

The fusees and chord provide a means of reaching into the accelerating frame without requiring an N3 break that would otherwise be needed to keep our reaction mass stationary relative to the accelerating mass.

Bessler's alternative approach, it seems, was to simulate the effects of an N3 break to freely keep his reaction mass physically within the accelerating frame.

So there's at least two different solutions to the same problem.

And likely others besides.

An EM approach might be to accelerate a rotor by modulating the phases of low or fixed-frequency AC currents, again, chasing the rotor's static FoR without upping the input energy.

With the right kind of CVT a minimum system is just two interacting flywheels, alternating their roles as input and output loads, and notching up a gear between each cycle.

A heavy pair of toroids, or tubes, depending on the application, and the power density could be off the scale. Nothing else currently on the horizon comes close. Personal jetpacks, flying battleships, etc.. seriously on the table. You can almost pull the numbers outa yer arse - half a megawatt per 100kg / meter^3 seems conservative. Current engineering tech is already capable of rotor speeds in excess of 100kRPM, so max power density might be pushing gW/m^3...

But all we need for now is a few hundred milliwatts, which should be more than enough to close-loop while applying demo loads.


Thanks again for the words of encouragement..!

[MrVibrating]

Hi Mr Vibrating ,

If I had to make a Fusee . I would try to use a length of 5 or 6 mm Festo pneumatic hose , cut in half and pinned to a wooden cone with small nails , no waiting for glue to dry . The nailing would be to the larger dia of the cone , with the nail almost to the one side of the tubing . Once the tubing is cut and wrapped around the cone you will have a groove for your string .
Since you have a above average IQ you will be able to decypher my coded drawing .

I was looking forward to logging in tonight to announce i'd thought of a way to make fusees, and you've just ninja'd me... well played.

It's an excellent idea and thanks for sharing. I have various diameters of tubing already.

The idea i had at work was to build up the basic shape by winding a thick strip of material onto a cylinder - easily repeatable for two identical fusees.

There's a type of thick rubber tape you can buy - i have a reel here - but i could also just make two wide coils of any suitable material and extrude them over cones, following your suggestion...

Very flat cones though - ideally we want much wider radial variation than most conventional fusees require, perhaps spanning the full rotor diameter (or more), over relatively few turns, since a clear gain should be apparent after just 3 cycles. So 4 or 5 winds should give a significant margin. I'll calculate the optimum design when i've decided the bill of materials.

Really, the calcs are just for predictive value.. checking theoretical consistency, but likely just ticking boxes. So i'll build a full mathematical model of the finished design for side-by-side analysis, however so long as there's a constant input power curve with minimal GPE to GKE losses and friction, a given drop height will cause a consistent rotor acceleration across a climbing speed range, and thus a linear rise in input energy for an exponential rise in RKE. You wouldn't need to measure anything if you could pick up twice your drop mass after 4 or 5 cycles.. This is all about the different relative dimensions of the input vs output energies, so empirical values are fairly arbitrary.

LOL i spent the last couple of days fretting over designing and cutting them from solid blocks, seriously considering contacting a CNC shop or 3D printers..

I'm unsure about the MT waterscrews you mention - not sure you quite get the concept here, which is that identical inverted fusees have a constant net cross-section, causing an equal angular acceleration for a given drop increment across a rising speed range, and so a linear rise in input energy for half that value squared in output RKE. A single fusee (or waterscrew) can't do the job, as GPE / MoI displacements would't remain isosynchronous.

That's basically a neat way to summarise the exploit - isosynchronous GPE to MoI workloads are inherently overunity, with RKE diverging by half the square of GPE.

And cheers for the piccy - honestly though i only wish this was half as ingenious as i'm trying to make it out to be.. it's just mechanics 101 - what is a Joule? Why does KE square? I haven't invented an asymmetry; it was always there, and we've been explicitly spelling it out in every physics textbook for over three centuries. 1J = 1kg/m/s of acceleration, and KE squares due to Newton's 3rd law (albeit indirectly). Laws of nature can't be intellectual property. The exploit is already implicit in the standard terms. Once the penny drops, all of this is face-palm obvious.

This weekend i'll try wrapping a lead band around a polystyrene disk or cylinder. That'll have an excellent MoI for its mass.

Once i have a viable rotor i'll model of its MoI, which will determine the ideal height and weight for an input GPE load and the fusee parameters. That's the logical order of proceedings, and so gives me some time to consider fusee options.. i can see this taking a good few weeks..

[MrVibrating]

...come to think of it, if the fusee's the same size as the rotor, then it might as well be the rotor. I mean, it's probably going to have significant mass, especially compared to a polystyrene disc. So it'd make more sense to build the fusees first, then fit a lead rim to one, model its MoI, and the remaining work could be done in a day.

[MrVibrating]

In the other two Kassel illustrations, the rope conveying the output power rises through a square window.

Tenuous link? Perhaps. But what about those pendulums? If the form of the gain is RKE, then why would you block it by regulating speed?

I can see two possibilities - these wheels actually did require regulated speed, but the regulator mechanisms were hidden inside. But Bessler denied anything hanging from his axles. Moreover, it would imply that he'd inverted the asymmetry - preventing RKE from squaring up, which instead an internal mass was free to do.

But i can't see how to get a gain from such an inversion - just losses. Will keep working at it, and hope others try too. There may yet be a gain dynamic there.

Another possibility is that the deliberate, repeated occlusion errors in the pendulum hangers are meant to indicate their futility - hinting that they're not supposed to be there, and that letting the speed rise might be a Good Thing..

Since everything in his wheels turned together, there were no stators and thus no CVT, so a good chance that RKE was not the form of his gain. But if it was nonetheless the same asymmetry - linear vs squaring energies via an effective N3 break (assuming this is the only one on the table)- then RKE would have to be the linear input energy, and the KE of some other internal mass would be allowed to square up. This would be the "prime mover"

So it looks like the more consistent explanation. I feel like i'm inching closer..

Perhaps weight drops apply regulated torque to an accelerating internal flywheel, then use its RKE gain to torque the main wheel? Or perhaps the weights torque the main wheel which torques the flywheel which lifts the weights? But that would require a hanging stator to gear up the rotors. None of these concepts seem particularly consistent with many other clues though...

Perhaps speed regulation was actually just a self-imposed handicap to maintain GPE, a drawback of the "peritrochium" design ethic that epitomised the fabled self-turning wheel everyone, not least he, had always envisioned. And yet he wrote that everything had to go around together in a true PM - a point on which, if i am right, then he is wrong. So then why would he think this?

The implication is that there's an effective N3 break in there somewhere. That's why his internal mechanisms can keep input energy in a linear frame. Meaning RKE is the form of the gain, but it's also limited by the requirements of side-stepping N3, which strongly suggests gravity is part of the means by which that trick is accomplished.

I know, i'm going in circles.. but they're getting tighter.. all the dots are there, they just need joining. Would much rather build a full wheel than a one-shot drop..

[MrVibrating]

Just thought of a really simple system - combining all of these components together - a drop mass, a rotor and a CVT, we end up with a simple yo-yo...

As a yo-yo falls its spool unwinds, narrowing the torque angle as RKE rises. I wonder if anyone's checked the RKE of a long yo-yo? The effect might be very small due to the small torque radius variation and high trans KE, but perhaps it's possible to design a "yo-yo" that falls slowly and converts each unit distance of fall into a consistent MoI acceleration. If the device weighs 1kg and descends 1 meter then full conversion of 9.8J GPE to MoI yields 48J RKE.

Perhaps this is also the solution to the toys page - the meanings of the hieroglyphs all pertaining to the whistling top at the bottom of the page, since the same principle applies generally to any RKE imparted at constant force through a narrowing torque radius... the "extraordinary" conclusion being that ordinary spinning tops may in fact be slightly OU already, if constant force is applied to the ripchord..

[daanopperman]

Mr V ,

The water screw in mt 93 , why did he wound the tube on a cone , he could have just wind the tube onto itself and bring the outlet out at the " pump " center . The scoop of the tube is the " input " energy , constantly adding a amount with each rotation , the water in the tube is the "many " pieces of lead which he refers to . Every time you drop a new mass on the turning wheel I see in the tube as the water represent the mass already on the wheel that have been added to impart velocity to the wheel . The Fusee was invented to compensate for temp. changes in watches , with a tapered spiral cone and a parallel wheel with spirals , that is what I see in mt 93 + 94 .
Your build will be much much more complicated than the idea , for to bring the weights to the fusee as they go around in the wheel . Bessler said the power of the wheel is proportional to the "housing" and the dia of the wheel , so I gather the fusee should be in this housing , close to the rim to impart max RKE .
Keep at it , as Fletcher , we all support you .

[pequaide]

You have named yourself well; you jump from one idea to another.

It is not J = ma it is F = ma. A joule does not cause acceleration. A Joule is the application of one newton (force) for one meter. At the end of one meter the velocity is 1.414 meters per second.

A newton is the quantity of force that causes a one kilogram mass to accelerate to one meter per second after the force is applied for one second. 1 N causes 1 kg to (a) accelerate by 1 m/sec/sec. After one second it is moving one meter per second; after 2 seconds it is 2 m/sec, after 3 seconds it is 3 m/sec, after the application of one newton for 400 seconds the one kilogram is moving 400 meters per second.

After a newton is applies for one meter (1J) the one kilogram is moving 1.414 m/sec, after a newton is applied for 2 meters (2J) the one kilogram is moving 2 m/sec, after 3 meters (3J) it is 2.449 m/sec, after the newton is applied for 400 meters (400J) the one kilogram is moving 28.28 meters per second.

Inertia is the measure of the quantity of mass. The inertia does not change because the mass has velocity. The quantity of mass does not change because the mass is moving or is at rest. F/a = m: if the acceleration is high the inertia is low; if the acceleration is low the inertia is high.

A force of 10 newtons can accelerate 10 kilograms to 20 meters per second after the force is applied for 20 second. From F = ma which is; F = m v/t

A force of 10 newtons can accelerate 1 kilograms to 200 meters per second after the force is applied for 20 second. From F = ma which is; F = m v/t

The energy of 10 kilograms moving 20 meters per second is 2000 joules.

The energy of 1 kilograms moving 200 meters per second is 20,000 joules.

The force can be applied in a circle and then we do not have to chase after the one kilogram.

You might not have to use a fusee but I think you might need to focus on one thing at a time.

[MrVibrating]

@Daan

Cheers mate i think i see your point..

TBH i'm still dwelling on how Bessler's wheels might've used the asymmetry.

Obviously, a pair of axially-centered scissojacks in a cross shape, opening and closing together with pulleys on their ends, form a CVT gear.

You don't even strictly need a pair per gear - a single input jack that contracts while another expands gives the same result as a pair of inverted fusees.

The main difference is that fusees are entirely passive, while jacks require energy - however if their displacements are synched equal and opposite then the cost is kept to a minimum.

That gives us squaring RKE for the output, a CVT to narrow the output torque radius while widening the input torque radius... all that's required is a GPE load that will be minimally compromised by the wheel's acceleration. Or perhaps something other than GPE, such as a flywheel or springs..

Dunno.. there's gotta be a simple way to tie this all together..

[MrVibrating]

You have named yourself well; you jump from one idea to another.

It is not J = ma it is F = ma. A joule does not cause acceleration. A Joule is the application of one newton (force) for one meter. At the end of one meter the velocity is 1.414 meters per second.

A newton is the quantity of force that causes a one kilogram mass to accelerate to one meter per second after the force is applied for one second. 1 N causes 1 kg to (a) accelerate by 1 m/sec/sec. After one second it is moving one meter per second; after 2 seconds it is 2 m/sec, after 3 seconds it is 3 m/sec, after the application of one newton for 400 seconds the one kilogram is moving 400 meters per second.

After a newton is applies for one meter (1J) the one kilogram is moving 1.414 m/sec, after a newton is applied for 2 meters (2J) the one kilogram is moving 2 m/sec, after 3 meters (3J) it is 2.449 m/sec, after the newton is applied for 400 meters (400J) the one kilogram is moving 28.28 meters per second.

Inertia is the measure of the quantity of mass. The inertia does not change because the mass has velocity. The quantity of mass does not change because the mass is moving or is at rest. F/a = m: if the acceleration is high the inertia is low; if the acceleration is low the inertia is high.

A force of 10 newtons can accelerate 10 kilograms to 20 meters per second after the force is applied for 20 second. From F = ma which is; F = m v/t

A force of 10 newtons can accelerate 1 kilograms to 200 meters per second after the force is applied for 20 second. From F = ma which is; F = m v/t

The energy of 10 kilograms moving 20 meters per second is 2000 joules.

The energy of 1 kilograms moving 200 meters per second is 20,000 joules.

The force can be applied in a circle and then we do not have to chase after the one kilogram.

You might not have to use a fusee but I think you might need to focus on one thing at a time.

Yes, excellent point - we can eliminate the fusee by not confusing force with energy. The time saved could be better spent playing GTA, or just punching yourself in the face repeatedly

[MrVibrating]

OK James so i fudged the units, but haven't you just underlined the same general dynamic i'm on about? Successive input Joules normally yield diminishing accelerations - the first Joule in your example gives 1.414 m/s, the second brings that up to 2 m/s, the third to 2.449 m/s etc..

So we come back to my initial question - why does KE square up, and hence is this diminishing return a fundamental limit, or one that can be circumvented? Is it not possible to get the same initial 1.414 m/s / J for each successive Joule by applying such a workaround?

And the point about an effective N3 break still stands - its per cycle input energy would remain constant, while the system's net KE squares up.

So the hypothesis that N3 symmetry is the practical reason why RKE squares up still stands, there's still an asymmetry up for grabs if we can freely or cheaply adapt to the accelerating frame, and a CVT still seems like the best tool for the job...

In short, if the source of the 1 newton is always catching up with the kilogram then it can keep accelerating it by 1.414 m/s per Joule, to 14.14 m/s after 10 J in, for 99.9698 J out in the stationary frame.

Trust me, i'd rather be playing GTA. But this needs thrashing out...


ETA: all your so-called proper maths have done is double the gain... i was happy with 50J out for 10J in... Now it's a hundred. This is your idea of help? At least my special force-energy maths were keeping a lid on things.