magrid configuration brainstorming

[rjaypeters]

Re: Micro coil technique application for macro coil?

interestingly, you'd only have to put current in a few of the wires (e.g. just the base polyhedron), and the mag fields should induce the correct current flow in the entire grid. even if the wires touch. interesting to see what such a superconducting wire sphere would do by itself - that is, with no energy input. or maybe just in the presence of a few external electromagnetic fields to charge it up. (in this way it would act sort of like a transformer.)

Can we play this game with the macro (rather than micro) grids? For example, can we take a finer mesh example (e.g. the "false" geodesic) turn it into a dual-coil magrid (I think the geometry will work) and let the active coil induce the current in the passive coil?

Re: Micro coil dynamic stability

however, since the grid is all similarly charged, it would repel itself, resulting in a "puffing out". this can help counter-act the stellating instability produced by the mag fields, as well as the attraction to the net-negative core.

I thought further about the balance required to keep it all together. It would have to be a dynamic balance because, I think, the rate of current increase and injected ions would now be tied together to keep the micro-magrid in the proper geometry before it rips apart. I think that might be a fun control problem for people who like such things.

that seems like the trade-off when u go to higher order polyhedra: the ratio of the wiffleball radius to magrid raidus is bigger, so you dont have as large of an electrostatic gradient (potential well) to focus and accellerate the ions. i suppose once you reach optimal kes for fusion cross-section, confinement becomes more important than kinetic energy. so maybe in larger machines, it'd be just as well.

I think this is the first I have read of these general results. Are they intuitive or based on what you have seen in the sims?

It is unfortunate, but entirely understandable, you have lost enthusiasm for the sims. We are all volunteers but the work you did was some of the neatest I've ever seen (not that I've seen much of this kind of work). I was just pleased to be able to provide some tiny bit of help.

I'd really hoped you'd gotten the gig with EMC2 and gotten your hands on some really cool hardware for higher-fidelity sims. And you were having a ball! Oh, well.

On simming the micro-grid, if you don't want to sim it, I definitely don't want to model, pointilize and create the segment list for it!

Although I count myself an optimist for the overall success of the polywell concept, I am not optimistic we will see directly published results from EMC2's work for a long time. So April 2011 doesn't mean much to me except it contains my 51st birthday.

So far, I haven't heard about any industrial accidents from a certain industrial park in San Diego which I count to the good. Still, I think the best information we will get for the next year is ladajo's tracking of the money. Indirect results at best.

[happyjack27]

one need not go to a super fine grid, so this'll work for any level of polyhedra u want.

re: induction. an active grid would have a much finer magnetic field structure, and i don't think you'd get anything like that induced from a coarser active grid of a compatable polyhedra. though i did kinda suggest exactly this idea at first - i guess i now have doubts about the ability of a wire mesh to iterate the mag field topology.

it is both from intuiton and playing with sims. but firstly, from the sims. i see it, and then i realize it makes sense. say the fields are balls. the more balls u have the less each ball sticks out into the center. 'tis a rough analogy, but at equal total ball volume (total amp-turns), the basic principle works out the same. they're just sort of squishy stochastic toroids instead of balls.

re what we will hear. we will at least know whether or not they give the go-ahead to build wb 8.1 - p-b11 mods. and at 8x field strength from wb 7, such results will be a lot more conclusive. (testing of scaling laws, sustained power mode, etc., which orders of magnitude higher expected fusion rates)

[rjaypeters]

Snub dodecahedron with two coils: one active, the other passive.

[happyjack27]

also i think maybe passive coils might have to be superconductive for the kind of induction we're talking about to work.

then by the act of rejecting magnetic field lines, well the field lines will just kind of wrap around it like a skin, and follow along it.

so you'd get sort of the topological equivalent of an electromagnetic, except the field strength would not vary at r^2 from distance because it's not actually "coming from" the wire, but rather being blocked by it.

they might sort of act as airfoils in a magnetic "wind" or rocks in a magnetic river. as such, there might be magnetic turbulence between them.

though i think it would be pretty safe (non-turbulent), at least, on a superconducting triangular mesh, to make every other wire (in each of the 3 directions) passive.

in any case, it would be an interesting experiment.

[rjaypeters]

From my primitive understanding of Faraday's law (induced current comes only from change of magnetic field), the passive coil must be superconducting, else the induced current will decay through resistance. Am I right in thinking the current-induction in the passive coil will come from the increasing magnetic field of the active coil.

Similary, the passive coil will loose current when the active coil decreases.

I didn't look carefully, but is the induced current in the direction we need? Also, what is the magnitude of the induced current? The same as the active current? That would seem to be too good to be true.

Please go easy on me, basic electrical engineering was long ago and far away...

[KitemanSA]

I don't get the desire for a "passive" coil. Pump it to the max, dude!

[happyjack27]

ya, i'm thinking it must be superconducting, too. regarding whether it's in the right direction: as in electrical circuits, it's a matter of looking at all the different combinations that don't violate the laws of electromagnetics. (current is proportional to voltgage differential over resistance, etc.) if there is ony 1 such solution, well then, that must be the way it goes.

yes, it's not actually going to "generate" a field. strictly speaking it will _reject_ a field. it's sort of like a wave-guide for magnetic fields.

why? why not crank it to the max? well, because this is FREE. (minus cooling.) also, because it's different. it may -- hell it probably does -- result in power power ratio efficiency. but it's be interesting to try and see what it does, on a small scale.

unfortunately i dont know how to go about simulating a superconductor in my lagrangian n-body sim, otherwise i'd just sim it. hmmm... maybe just finding the mag field strength at each point in the superconducting wire, then have that point generate a mag field of precisely opposite strength (same strength, opposite direction), thus canceling out any field lines that pass through it?

i suppose at that rate, then such a wire would be a huge electron loss channel. or would it's cancelling effect on an approaching electron's mag field deflect the electron? ... further proof that i'm not a physicist.

but if that way of simulating a superconducting wire (w/no power) is correct, then, say, a superconducting ring inside a powered ring (electromagnet) - concentrically - would create inside the central superconducting ring a magnetic field _of opposite orientation_ from the one between the two rings _even though there is no power being supplied to the inner ring_.

and in this general manner, one could iterate a magnetic field. - one could create a finer topology from a coarser one, using only _unpowered_ superconductors.

in fact, i would think such a ring would gravitate towards the exact center of the powered ring in order to minimize the total amount of field line displacement. thus you could have a virtual inverted _levitated_ magrid inside another, physical magrid. and you wouldn't even have to supply any power to the inner magrid (though u would have to keep it at superconducting temperatures. you could also reverse this, having the induced magrid on the outside. this would make it a lot easier to cool, and would probably improve recirculation, as it would bend escaping field lines back around. and again - it wouldn't require any power. (except for cooling.)

maybe just have the faraday cage a superconducting triangular mesh?

[rjaypeters]

quote: "...thus you could have a virtual inverted _levitated_ magrid inside another, physical magrid. and you wouldn't even have to supply any power to the inner magrid (though u would have to keep it at superconducting temperatures."

These conductors are too big for the cryo-cooled wall to keep up by passive cooling (unlike the micro-grid, maybe). How do we cool levitated conductors? Flexible tubes?

quote: "...maybe just have the faraday cage a superconducting triangular mesh?"

All conductors connected to each other? I believe you suggested this before. How does the current flow to create an organized, patterned magnetic field useful for wiffleball formation?

[happyjack27]

quote: "...thus you could have a virtual inverted _levitated_ magrid inside another, physical magrid. and you wouldn't even have to supply any power to the inner magrid (though u would have to keep it at superconducting temperatures."

These conductors are too big for the cryo-cooled wall to keep up by passive cooling (unlike the micro-grid, maybe). How do we cool levitated conductors? Flexible tubes?

yes, cooling an internal levitated magrid would be a major engineering problem. i have no answers for that, except that it's probably not feasible on that account.

quote: "...maybe just have the faraday cage a superconducting triangular mesh?"

All conductors connected to each other? I believe you suggested this before. How does the current flow to create an organized, patterned magnetic field useful for wiffleball formation?

u have 2 options: connect the wires or don't. either case the desired current flow is the same: that shown in the link on my original (or second?) post on trainguler mesh. to recapitulate, there are three orientations:

\_
/

if you turned each into an arrow pointing outward, that would represent the current direction for each wire orientation in a triangular grid. that is the _unique_ orientation that results in a) no divergence in the electrical field (current in = current out) and b) a closed directed current loop around every triangle. (notably the current loops also alternate direction (in-out) as required.)

regarding passive, induced, grids, well result a) is automatic - it is required by the laws of physics. result b) represents a solution of least action. so though it may not be required by physics, it is a configuration that the system will be attracted to. now the other part of this is the iteration of magnetic fields via superconductors as mentioned above (re:concentric rings) by superconductors rejecting field lines. and as mentioned in a prior post - those field lines "wrapping" around the superconductors and snaking along them. this strongly favors a magnetic field structure that is globally consistent with c) the superconductors rejecting all field lines, via d) minimal disruption of the mag field, and by entropic effects of the laws of electromagnetics e) minimal overall curvature of the field lines (and simplest possible topology (e.g. fewest "holes")). given the restriction that c) is non-negoitiable (as is a), btw.), c, d, and e all have the same unique minimal solution: the one desired.

[happyjack27]

on another note, i think in the concentric rings example, the inner ring might want to turn so its perpendicular to the outer ring, so as to minimize the magnetic field disruption. that, and it would want to leave the plane of the outer ring for the same reason. so the configuration probably would not be kinetically stable.

[rjaypeters]

u have 2 options: connect the wires or don't. either case the desired current flow is the same: that shown in the link on my original (or second?) post on trainguler mesh. to recapitulate, there are three orientations:

\_
/

if you turned each into an arrow pointing outward, that would represent the current direction for each wire orientation in a triangular grid.

Except this concept does not work wrapped around a sphere (see my posts after your hand-drawn picture). A cylindrical/helical concept might work with end plugs (I'm pretty sure closing the ends with a couple rings would work) except the sphericity would not be very good.

[happyjack27]

u have 2 options: connect the wires or don't. either case the desired current flow is the same: that shown in the link on my original (or second?) post on trainguler mesh. to recapitulate, there are three orientations:

\_
/

if you turned each into an arrow pointing outward, that would represent the current direction for each wire orientation in a triangular grid.

Except this concept does not work wrapped around a sphere (see my posts after your hand-drawn picture). A cylindrical/helical concept might work with end plugs (I'm pretty sure closing the ends with a couple rings would work) except the sphericity would not be very good.

that's why i suggested certain base polyhedra. those polyhedra have only 6-segment vertices and 4-segment vertices. the 6 segment ones are trivial. the 4 segment vertices are not that difficult, either. the 4-way intersections are crosses, with the one orientation the current going inward, the other one, the current going outward:

Code: Select all

      |
     \|/
/___  . ____\
\           /
     /|\
      |

now in this case, for the triangles in the 4 corners, there is only 1 direction the 3rd side can go which results in a loop. and from there, all intersections are 6-way and uniquely determined, inductively.

[rjaypeters]

Let me see if I get this: a superconducting mesh will self-organize the current flows? Cool!

[happyjack27]

Let me see if I get this: a superconducting mesh will self-organize the current flows? Cool!

yep. that seems to follow inevitably from the logic i laid out. however, the _strength_ and _stability_ of the resulting fields is another question altogether. we're not necessarily getting anything for free here, besides a different (finer) magnetic field topology. it's quite possible that the net effect of putting a passive superconducting grid in the system would _worsen_ confinement.

still pretty cool, though!

[rjaypeters]

So, put all your ideas together. Suspend a micro-grid mesh inside a cryo-vacuum chamber and induce current in the mesh. Need to maintain the orientation of the grid so the alphas (we hope it's b-B11) come out aimed at the venetion blinds.

Could get fancy and levitate the micro-grid above magnets in the chamber walls. Orientation controlled by extensions from the grid to the chamber walls, etc.