Plying CF as pseudo-inertia to scam N3

[daanopperman]

Hi guy's

Cc ,

Your gif ,
If you have your mechanism in a a horizontal position , in a balanced state ( all pair's closed ) and you open the lh side by a force , was it not for gravity and therefore ob , the wheel would not turn ! .

That to me is the indication that this is the mechanism Mr V wants for his non backtorqe reaction less drive .

If I had a pivoted bar , with a pulley on either end , ( also in a horizontal orientation ) and 2 equal wights tethered around the pulley's on each end , I can move the weights up and down either end with out any back torque to the system .On the lh side of the bar , it would not matter if I had the lh side weight pushed down or the rh side weight pushed up , the weights in this case is independent of the system it is connected to ,

I do not think you will have any joy with a asymmetric break , for N 3rd you can not have a action without a reaction , that is a fundamental property of putting a mass into motion , you cannot accelerated a mass without a action , not inside or outside of a gravitational field .
If you use gravity or a magnetic field to increase the velocity of one of a pair of masses , that field will be the stator in your wheel for one of the masses , and if there is no
communication between the 2 masses , there cannot be a reaction between them .

KE =1\2mv² , but I think that is where the 2 masses is in the same frame of reference .
If I have 2 masses on a bus , and send one mass off at the expence of the other , so that one mass is brought to a standstill , that equation is true , but relative to the bus , that is not . Bus and 2 weights V = 10 m\sec , after separation of the 2 masses , one mass is 0 m\sec , the other is 20m\sec , but relativ to the bus , the mass is doing only 10m\sec . The faster the bus goes , does not make the mass go faster after separation .

I think the only way this idea could work is if the accelerated mass is part of the rotating mass ( the wheel ) in the first place , so that the wheel have to looses mass , and later collect t it again .

[cloud camper]

Hey Daan - I will get back to you but for now, this is a very related development!

http://www.sciencealert.com/leaked-nasa ... -does-work

[MrVibrating]

dax .. MrV is using an actuator element to provide the separation accelerations in his WM sim concept.

First the two equal mass weights (the vertical mech) were released from a standing start in a gravity field and at the same instant the actuator is 'fired'.

The top mass of the two has an equal but opposite force applied to it as 'g', effectively cancelling out the two opposing forces i.e. out-of phase forces, no displacement (stationary), no change in mv, no work done, no GPE change, no KE change.

The second lower mass gets an in-phase force boost reinforcement ('g' + actuator acceleration force) i.e. gets all the displacement in the same time interval and all the mv, and all the KE.

Effectively, the concept is to use a stored force to cause a momentum decoupling and thereby increase the KE of the lower mass to do Work.

Coming up with a complete mech to reset initiation energy levels and close the loop etc is the technical task.

The latest sim has 'the mech' dropping in a gravity field so that both masses already have velocity and mv and KE's. To show that the energy cost of firing the 'actuator jack' has no greater cost to the system than if they were stationary at start example.

FWIW you may remember discussions in the past using the horizontal example. About 2 mass blocks on a frictionless tray of a moving truck and being separated by force. What were the mv's and KE's when the tray was stationary and when the tray was at constant speed at force initiation ? This lead to a discussion with jim_mich about relative KE's of the 2 blocks re ability to do Work. At that time the discussion was focused on the rising KE's from a moving start rather than MrV's momentum conservation decoupling. JMO's.

I've just demonstrated the decoupling from GPE, which is a zero-sum deal.

No net work is done by gravity. Mass goes down, mass goes up. GPE-in = GPE-out.

It doesn't even do any work causing the momentum asymmetry - it's an entirely passive effect, simply artificially modifying the effective values of 'resistance to acceleration'.

However i've also just demonstrated the decoupling between the internal energy (required to replenish the internal PE store, if one is required (doesn't seem strictly necessary - a CF workload could directly couple RKE to the internal interaction)), and the RKE of the net system.

Again, the pivotal detail is the energy cost vs value of momentum!

The KE value of a given distribution of mV is an invariant function of 1/2mV^2. Nothing here has any affect upon this. This normal KE standard is actually a benefit to us, because it squares with velocity. We all know this already, so no further elaboration is required.

But what i have just demonstrated above is that conversely, the energy cost of buying that momentum can be swindled, and held constant, irrespective of speed, when purchased sans counter-momentum..

And so the same input energy causes the same rise in momentum, at whatever the velocity.

The cost of our discount momentum here is set by gravity. It's an optimum rate of exactly 9.81 Joules per kg-m/s.

Whatever the speed.


And because the KE or RKE value of the accumulating momentum is a function of speed, there's some threshold of velocity beyond which the system has more RKE than the PE spent internally.


So what i've laid out is a full thermodynamic solution to OU.

Input and output energies have completely different dimensions. Output KE is time-dependent, squaring with velocity, whilst input PE evolves linearly, with essentially the same dimensions as momentum...

..as such, we're inputting not so much "energy" - which, by definition squares with velocity - but simply inertia times velocity... ie. raw momentum.

And the reason for this is simply that inertia doesn't rise to compensate velocity in any way - it's temporally-invariant, so insensitive to velocity.

Due to the momentum asymmetry we can drag our reaction mass along with us, by sharing our momentum gain with it following each not-so-mutual acceleration. Hence both masses always begin each cycle at the same relative velocity, and hence the cost of each subsequent kg-m/s of momentum remains the same fixed price as the very first...

...in principle, all the way up to lightspeed, and millions of times OU.

Although so far, it looks like a vertical rotor powered by this exploit will top out at around 50 - 60 RPM, beyond which synchronisation is lost (the weights have to be mutually driven apart while falling, and then re-collided while traveling horizontally, ie. uninfluenced by gravity, to redistribute the gain symmetrically and thus accelerate the net system of both masses).

So all thermodynamic doubts have been eliminated. We have a constant, speed-invariant input vs a squaring output. Job done, in terms of identifying the OU energy gradient, and thus what we need to design and build.

Again, it is essential to recognise that this is not the only implementation possible - the raw ingredients of this 'prime mover' are simply two inertias, one or both of which are subject to gravitation.

They could be angular, linear, or a combination of each; i've demonstrated examples of each over the preceding pages.

The only reason i've gone with the linear version here is to make it visually and mathematically simpler to follow (since a combination of angular and linear momentum terms would seem unnecessarily complicated for the purposes of simply communicating the concept).

However the angular examples, or angular/linear, do seem conceptually simpler to follow - the lower mass can be substituted by the angular inertia of the net system, instead... really, using a descending mass like this seems kinda redundant. But it is easy to follow, and it works.

So the potential mechanical schemes that might be employed to attack this front are entirely up to your imagination. So far, the only real mechanical considerations i've made are towards physical viability, and on that count, i see nothing inherently contradictory or impractical about it. It's do-able. A machine can be designed to draw upon this I/O disunity.

Re-check the demos, cos if you're not seeing magic, you've missed something..

[MrVibrating]

Hey Daan - I will get back to you but for now, this is a very related development!

http://www.sciencealert.com/leaked-nasa ... -does-work

Yes, the EM drive, it it works, will also be OU in exactly the same manner as this system. Shawyer would perhaps argue that it's payed for all its momentum... but then so have we. We're just buying at a fixed, rather than speed-dependent price, and then cashing in on its speed-dependent value..

[MrVibrating]

Hi guy's

Cc ,

Your gif ,
If you have your mechanism in a a horizontal position , in a balanced state ( all pair's closed ) and you open the lh side by a force , was it not for gravity and therefore ob , the wheel would not turn ! .

That to me is the indication that this is the mechanism Mr V wants for his non backtorqe reaction less drive .

If I had a pivoted bar , with a pulley on either end , ( also in a horizontal orientation ) and 2 equal wights tethered around the pulley's on each end , I can move the weights up and down either end with out any back torque to the system .On the lh side of the bar , it would not matter if I had the lh side weight pushed down or the rh side weight pushed up , the weights in this case is independent of the system it is connected to ,

I do not think you will have any joy with a asymmetric break , for N 3rd you can not have a action without a reaction , that is a fundamental property of putting a mass into motion , you cannot accelerated a mass without a action , not inside or outside of a gravitational field .
If you use gravity or a magnetic field to increase the velocity of one of a pair of masses , that field will be the stator in your wheel for one of the masses , and if there is no
communication between the 2 masses , there cannot be a reaction between them .

KE =1\2mv² , but I think that is where the 2 masses is in the same frame of reference .
If I have 2 masses on a bus , and send one mass off at the expence of the other , so that one mass is brought to a standstill , that equation is true , but relative to the bus , that is not . Bus and 2 weights V = 10 m\sec , after separation of the 2 masses , one mass is 0 m\sec , the other is 20m\sec , but relativ to the bus , the mass is doing only 10m\sec . The faster the bus goes , does not make the mass go faster after separation .

I think the only way this idea could work is if the accelerated mass is part of the rotating mass ( the wheel ) in the first place , so that the wheel have to looses mass , and later collect t it again .

Hi Daan,

Not sure you've quite got the concept - we're using gravity's innate acceleration to cancel our counter-momentum, by vectoring an inertial interaction against it.

The lower mass could be a linear displacement, as per the last example, or else angular - just the wheel body would suffice.

The upper mass could also be an angular displacement. Angular vs angular, linear vs linear, and angular vs linear examples have been demonstrated already.

The N3 break works. The above demonstrations are incontrovertible. We can input momentum of one sign only, and accumulate it over successive cycles, at a fixed energy cost per unit of momentum added.

But the mounting KE value of all that accumulated momentum is a standard function of 1/2mV^2 - ie. it's worth more than we've paid for it, and moreso the more we buy.

All this is proven already, and follows from simple first principles.

As explained, it is fully consistent with all three of Newton's laws:

A mass or system of mass's momentum remains constant unless acted upon by an external force; check.

F=mA; check: everything that should accelerate, does.

Counter-momentum is imparted, to Earth, via its attraction upwards towards the inertially-suspended weight. It isn't applied to the inertially-suspended mass, due to its equal opposite counter-acceleration under gravity.

So we're fully consistent with N1, 2 & 3, so consistent in fact that we arguably end up breaking all three, too, if you wanna get super-pedantic about it - but then that just underlines how superlatively consistent we really are; it's all in the T&C's, check the small print..

[MrVibrating]

Ok here's what an ideal implementation might look like..


Suppose we have a significant OB weight. Lotta mass, good radius. plenty of drop.

By falling through an angular trajectory, as when fixed to a wheel or balance beam, CF can be generated.

This CF load can be independent of the GPE load; we could have a pair of masses, even heavier than the GPE load, and at even greater radius, but balanced, 180° opposite one another, on either side of the axle.

Both these CF masses can slide outwards, increasing radius in step together, while remaining counter-balanced.

This radial displacement of the CF masses can be quite small - the forces will be so high, little displacement will be required to perform useful work.

The CF masses are used to drive the interacting inertias in free-fall; details to be worked out, but no magic required.

The optimum config would seem to involve using that GPE output to efficiently convert GPE to centrifugal PE... and then using that CFPE to apply as many discrete asymmetric inertial interactions as possible.

So the mental image i'm trying to evoke is something like a jack-hammer, that fires rapid bursts of these asymmetric interactions in a chained sequence, whilst being powered by one long smooth continuous GPE output, via its conversion to CF/CP.

The reason is, that the number of complete successive inertial interactions is the single most important factor determining how soon into a full rotation of the system - after how may degrees of rotation - we can break unity and so repay that GPE..

At the other (problematic) extreme, if we only had one or two reactionless accelerations per GPE input, we wouldn't even necessarily be capable of breaking unity, even with a push-start..

Maybe MT 138 hints at a simple way of "chaining" these interactions, perhaps literally.. what you really want is to be able to crank 'em round like a Gatling gun, hammering the net system into a net momentum asymmetry before re-raising the weight even becomes an issue..

[MrVibrating]

...so you've got a spinning wheel, with CF workloads rippling in and out a little way, over a short radial displacement, as the whole thing rotates..

These are somehow in centripetal tension with a series of spring-loaded sissorjacks, and rapidly-reloading them after they've fired. Doubtless our US contingent will be the go-to team on this part of the mechanism.. we need to make a simple reciprocal interaction cycle in full-auto..

These interactions hammer their momentum gains into the net system by colliding with their stops, bouncing back with a little KE and momentum left (so subsidising part of the cost of successive interactions, making the momentum gains even cheaper..!)

The only calculation required is how many of our asymmetric interactions we'll need to break unity, before we can start re-lifting the GPE that caused the initial rotation.. it doesn't matter if it's four, five or 15.. whatever it takes, the break-even - and thus OU - threshold is the very definition of low-hanging fruit.. if it's mathematically viable, it's buildable.

So to recap, we wanna rapid-fire these momentum gains via a powerful CF workload, itself powered by one long continuous GPE... using a very high MoI of the CF masses to maximise the GPE-to-inertial-exchange conversion efficiency.

Which means my initial mock-up rig on the previous page probably isn't up to the job, beyond basic data acquisition..

Need to suss some kind of rapid-fire, fully-automatic, burst-fire mechanism..

[daanopperman]

Mr V ,

Sorry for it to take so long to sink in .working on a reaction less drive .

[MrVibrating]

No worries, it's still sinking in for me too.

I only realised after i went to bed last night, that yesterday i demonstrated that the 96.23 J per 9.81 kg-m/s of momentum is a constant, speed-invariant price...

...therefore a second, identical helping, of another 9.81 kg-m/seconds, will take our net input energy up to 2 * 96.23 = 192.46 J, for which we have bought 2 * 9.81 = 19.62 kg-m/seconds of momentum, shared between our two 1 kg masses, so with 9.81 kg-m/s on each.

So if we make a third and then fourth identical purchases, we'll have spent 4 * 96.23 J = 384.92 J, for 4 * 9.81 = 39.24 kg-m/s, divided between our two 1 kg masses, means they have 19.62 kg-m/s, and 192.47 J each.

A fifth identical purchase takes us up to 5 * 96.23 = 481.15 J total input energy so far, for which we've bought 5 * 9.81 = 49.05 kg-m/s of momentum, divided between two 1 kg masses, so there's 24.52 kg-m/s on each 1 kg mass..

Right?

And yet, at a speed of 24.52 kg-m/s, a 1 kg mass has precisely 300.61 Joules, no more or less.

Work it out for yourself - KE = 1/2mV^2...!

We have two 1 kg masses, each carrying 24.52 kg-m/s of momentum...

...for which we've paid a net total of 481.15 Joules...

...and yet we have 2 * 300.61 = 601.22 Joules of KE...!

So after precisely five identical interactions, we've broken energy unity, at an efficiency of 601.22 / 481.15 J = 1.24 CoP.

Five reactionless accelerations = OU.

Gentlemen, i believe we may have cracked the Toys Page...!



PS - Daan - it's entirely your call, if you understand what you're getting into, but i just want to be clear - i have far higher confidence in OU at this stage, than reactionless propulsion.

The latter would be an absolute bonus of course.

It may not pan out though, so moderate your expectations!

Even if it does, it may turn out that it is a requirement to convert cancelled counter-momentum into KE like this, in other words, reactionless propulsion would just be a side-effect of running a Bessler wheel, and no special redesign would strictly be necessary.

So right now my confidence in propulsion applications is around 50%.

My confidence in KE generation is 100%.

Working on the latter, rather than the former, is still going to inform upon the former.. Your call though, and glad you're as excited as i am!

[MrVibrating]

Isn't it funny how 'five' it is, tho?

I mean, you can appreciate why he was so wrapped up in the number..

..because at just four cycles, everything is at unity - the masses have precisely the right amount of KE, equal between input and output..

But buy just one more reactionless rise in momentum, and we suddenly have 24% more KE than we've input.

Buy much more after that, and that OU percentage escalates to ludicrous proportions.

But the magic kicks in at five.


Hmmm... in that case, do we really need the preceding four reactionless accelerations? I mean if they start out under-unity anyway, and then reach unity at precisely four successive interactions, then why not just skip them entirely and unleash some kind of PE load that gets us straight up to that threshold unity energy level, and then buy that 'fifth' portion as the first reactionless purchase in our sequence? Unless that causes some key change in conditions, we should still get that 124% efficiency, so we can immediately repay that PE load..

Some kind of GPE 'kick-start' principle, to jump straight into the OU action without the redundant foreplay.. type stuff..

[cloud camper]

OK folks, been just creeping along, taking days to just get this far.

Since we have now interconnected the timing mechanism left to right we are getting much better control over the weight motions just like MT138 depicts.

The timing is still off but the motions are much better controlled with much less bouncing I believe. Once we get the timing close to right then we can start working on impacts to try and capture the excess KE.

We can observe also that the timing angle is a compromise, ideally it would advance at higher rpms, just like the ignition timing advances on an IC engine.

The timing is currently set to 45 degrees advance but looks like it is too much so will try retarding a bit more. Should help demonstrate why sliding weights don't work like you think as there is a horrendous lag time that must be taken into account!

We can also note the timing weights show a very nice overbalancing motion yet do not power the wheel.

[MrVibrating]

CC great work, but i'm still concerned this can't work since it's not strictly following the principle; it is incidental that the weights move up and down in pairs!!! That is not the exploit! Specifically, they must be accelerating against one another's inertia, not that of the main system / rotor body!

The examples i've given demonstrate different potential approaches to exploiting the principle. Angular-angular interactions look much easier to implement into a working mechanism.

I've continued using linear masses for the sims because their momentum and KE terms are so easy to understand. But this would be very difficult to implement into a working design.

To strictly follow the maths you'd want something like a lightweight linear actuator that can push the masses apart (again, exclusively against one another's inertia, not that of a third body or the net system) - a carbon-fibre scissorjack or god-knows what lol.. and a spring or something with which to fire it, as it falls. You can then harvest the momentum gain by letting the lower, accelerated, mass, strike or be caught by the wheel / main system. The RKE you will have after 5 such interactions should be 124% more than you've used to fire all five mechanisms. But then how to transfer that RKE back in to reset the springs or whatever? Some kind of CF workload would seem the way to go.. it just gets really complex doesn't it? Nothing mechanically impossible, but obviously there's gotta be a simpler implementation.

In the thought-experiments and sims i've shown, the displacements were highly exaggerated, just to emphasize the principle, but it's fully-scalable - we could have lots of little asymmetric interactions per cycle, from a simple repeating mechanism, rather than a couple of great lumbering ones...

The logical way to proceed would be to identify the simplest possible mechanisms that can perform the key ingredients of the interaction across a narrow angle, of say around 60° per segment.

We don't have to generate large amounts of momentum per interaction - if it meant it was more robust, especially across a range of RPM, then it would be more efficient to have lots of smaller ones, than one big one. Smaller gains per interaction, but at a higher frequency of interactions per cycle.

Sorry CC i don't mean to denigrate your sterling efforts, but hey i'm going off half-cocked here, it's hard not to once you've understood that this is an open flank.. It's a fait accompli alright, but it's all about following the momentum, at each critical step.

- step 1: Generate a momentum asymmetry, by forcing two inertias apart or together, the effective inertia of one of which is temporarily augmented by an applied static force field.

These two inertias can be angular, linear, or any combination. Maybe even radial, too.

The static force field can be gravity, permanent magnets, or possibly even CF/CP.


Having chosen which two inertias, and which static force to use, proceed on to step 2.. Obviously, though, it would seem to make things easier if one of these inertias is actually just the angular inertia of the net system..


- step 2: apply a force between your two inertias, while one or both of them are also subject to the static force, such that only one of them is caused to accelerate.

So the motion of the other inertia remains constant, or as close to constant as possible.

If you've succeeded there, you've just input momentum between the two inertias, of one sign only.. so proceed to step 3.


- step 3: Consolidate your gains. The two inertias need to interact a second time, but without interference by the static force, and so just a regular elastic collision would do, or maybe some kind of catch.. doesn't matter too much, the point is simply that after this second interaction, both inertias should now be at the same speed relative to one another, in basically the same relative condition they were prior to step 2.. except now both of them are at a higher ambient velocity, caused by this influx of unidirectional momentum.

So if you've met this condition, then simply repeat steps 2 & 3 four more times, and then go to step 4.

- step 4: after five cycles through steps 2 & 3, your system is now rotating with 124% more kinetic energy than you've spent actually performing those interactions.

The actual figures you've spent on them doesn't matter, provided losses are kept low, you'll break unity after five of them. Obviously, Bessler even didn't have the benefit of roller bearings, so this thing tolerates medieval friction..

So now you need to channel that RKE back in to power the interactions.

Without recourse to a stator.

Which basically leaves CF / CP workloads as the only way to harvest the RKE from within the rotating system..

[MrVibrating]

I think a little more elucidation on the core principle, and different potential means of applying it, might help..


..come back to this doodle i was playing with the other day:




...it's just trying out ideas, but can you see how i was trying to follow the maths?

It's all over a bit fast, but the green mass is a GPE input, dropping against gravity, and so applying a mutual torque between the red inertia and that of the wheel body.

Because the red inertia is subject to gravitation, but the wheel body is not, the wheel receives a greater share of momentum from the GPE output, ie. it's caused to accelerate, but the red inertia isn't.


The trick, then, would be to subsequently re-collide the red mass with the wheel, such that it shares its momentum back with it, summing them to a new net equilibrium velocity, and ready to repeat the interaction.

The above mechanism is only halfway there, so far - this is just step 1.


But to get a better idea of what inspired it, take another look at this exaggerated reduction of the case of an angular-angular interaction:



..so that's just two angular inertias, one of which is also subject to gravity.

Obviously, on the face of it, this looks completely different to the linear-linear version i've spent more time on.. however what i'm actually doing with them is exactly the same thing; they're just a pair of inertias, being propelled back and forth against one another, whilst a static uniform force field interferes in the resulting distribution of momentum. We generate the momentum asymmetry by vectoring one half of the inertial interaction against the static force, and then consolidate it by not vectoring the second half of the interaction against it, ie. when the inertial displacement is anti-parallel to the static force vector.

But as you can see, that above example is not practical to develop, since it's bitten off more momentum than it can chew on the first bite.

Obviously we need to nibble at it in more manageable chunks.. make it, consolidate it, repeat.

So exactly what that anim above is doing, but with much smaller angles of relative displacement..

For instance, imagine if there were multiple arms rotating with the system, but as they fall on the descending side, there is some mechanism that torques them against the wheel - so torquing the weighted armature back upwards, while torquing the wheel in its current direction; the intention is not to reverse the armature's direction or even to lift it, but merely to cancel its acceleration under gravity, so that its speed remains constant during the impulse..

It could be a very brief impulse.. just a lil' hiccup on the way down.. But after five of them, the system's OU.

Now, take a look at MT 24, MT 25 and MT 60, and you'll perhaps see them in a new light:







...as you can probably now see, these kind of fit the profile we're after; these systems can apply a mutual torque between the wheel's angular inertia and some other angular inertia - that of a weight (so subject to gravity) - constrained to an angular trajectory.. and additionally, there is a radial motion between them which can apply or receive this torque, potentially as inputs, outputs or either.

With multiple mechanisms operating across limited arcs, each mechanical section must therefore accomplish an independent asymmetric inertial interaction, once per cycle. They may also interact with one another during their operation, so one might potentially reset another..

So a good goal would seem to be to find some mechanism that can accomplish the key steps across a much narrower angle of rotation.

This is what i'm gonna try focus on next. Simplify, reduce, compartmentalise etc.

[MrVibrating]

The basic concept is incredibly simple: momentum generated via asymmetric inertial accelerations has a constant, speed-invariant energy cost.

Whereas, under all normal (N3-compliant) circumstances, the energy value of momentum squares with velocity.


Therefore the exploit is simply cut-price momentum.

The more you buy, the more momentum and thus velocity you have and thus the greater the disparity between its constant cost, and speed-dependent value.

This is a viable OU gradient. It is 'the magic'. The 'impossible' part, done. No further magic is required for a mechanism to exploit it, it's just an engineering challenge to deduce the optimal configs for whatever given application..

Don't worry, we'll have a runner within weeks, i'm sure..

[MrVibrating]

..just noticed a good consistency with the one-way wheel - if we take that above scheme, of blipping the weighted armature with counter-torque during its descent, some kind of spring could be used (in place of the motor in the sim).

The pre-load on that spring could be adjusted via a screw on the axle, which would vary the torque/angle/time of the resulting impulses and thus the momentum distribution, and hence coasting RPM...

But whatever, consistency with any particular clues is by the by; i've shown that the momentum asymmetry can be generated, at equal magnitude and equal energy cost across a viable range of velocity, its accumulation depends upon nothing more than N3 doing its normal job, and its OU value depends upon nothing more than KE=1/2mV^2.