magrid configuration brainstorming

Discuss how polywell fusion works; share theoretical questions and answers.

Moderators: tonybarry, MSimon

Betruger
Posts: 2310
Joined: Tue May 06, 2008 11:54 am

Post by Betruger »

MSimon wrote:
olivier wrote:I found some nice Java applets on the web, designed by Paul Falstad.
I have played with the 3D Magnetostatic Fields Applet. Select the "loop paired stack, opp." and "Display: Particles (Vel.)" options and within seconds, you will have built one third of a minimalist Polywell and will observe some confinement in the middle. Many losses from the sides, are you surprised ?
You can rotate the picture, adjust the B-field or the coil size and see the movement of the particles change in real-time.
The Java code is available and, from what I have seen, it would not be too difficult to add four more coils. It is really nice for educational purpose, but might be more difficult to turn into the visualization tool drmike is longing for.
I wrote the author an e-mail and asked if he would add the Bussard Configuration to the applet. He said it was a good idea.
Any feedback since?

rjaypeters
Posts: 869
Joined: Fri Aug 20, 2010 2:04 pm
Location: Summerville SC, USA

Post by rjaypeters »

The "Tombo'd" dodecahedron:
Image
Image Image

The longer radius fillets are to keep the close approaches approximately the same size. It makes the pentagons look funny, but I didn't feel like realigning any of the straight segments tonight.

Just one long coil which might help field asymmetry problems on start up, if that's even a problem.

I looked at a Tombo'd icosahedron, but the resulting triangles seem small compared to the volume, I might do it later...the dodedahedron was kind of fun (more than creating a segment list, LMTY).
"Aqaba! By Land!" T. E. Lawrence

R. Peters

rjaypeters
Posts: 869
Joined: Fri Aug 20, 2010 2:04 pm
Location: Summerville SC, USA

Post by rjaypeters »

Single-coil dodecahedron on a 1m radius sphere.

Image Image

Notes: Small radius is 0.2m vice 0.25m of previous practice (0.25m tight turns weren't enough fun to model many times for a non-paying gig).

Icosahedra and dodecahedra share a type of spherical symmetry. For this exercise, they are the same.

Playing the single-coil game with the two higher-order polyhedra (unnamed here because I can't remember the names) I've already modeled would probably look like a human brain hemisphere or some of the cooler-looking corals). I'm not tempted, I have things to do...and some new fiction to read...
"Aqaba! By Land!" T. E. Lawrence

R. Peters

tombo
Posts: 334
Joined: Fri May 09, 2008 1:10 am
Location: Washington USA

Post by tombo »

Sweet!

Thank You RJ for taking up the slack.
I had to stop when my new job started taking up all my time

My original inspiration for the one-pass coils was the bent copper tubing coils with water flowing through them that I used to make for high powered induction furnaces. They handle huge currents. One of our biggest problems was condensation because the cooling was too good. Msimon pointed out that Cu has very low resistance at LN2 temperatures. My thought was a proof of concept machine for polywell but it is not low enough resistance to make break-even. And of course they have no finicky T's with the many layers needed for superconductors. And, yes, I admit they would be difficult to fabricate with too many layers or with too many bends.

What program do you use to create your models?
I use AutoCAD and Inventor.
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

rjaypeters
Posts: 869
Joined: Fri Aug 20, 2010 2:04 pm
Location: Summerville SC, USA

Post by rjaypeters »

High praise, indeed. Have a care or you will turn my head! :D :oops:

happyjack27's simulations show the dodecahedral concepts have promise and I think the single-coil concept has some legs because of operational simplicity - just one coil to think about. At the price of difficult fabrication.

I'm wondering if the best fab location is micro-gravity, probably not until we have other details sorted.

I'm curious if anyone here knows enough about SC manufacture and handling (i.e. is there such a thing as an SC splice?) to offer an opinion on the single-coil concept as shown in the octahedrals you designeds and the dodecahedrons above.

I use ViaCAD 3D V7. It's relatively cheap and does a good job for this work. My favorite tool is Catia V5, but the license is prohibitively steep. Using ViaCAD, I actually have to think about certain concepts I could hopscotch through in Catia V5.

My respect for your work may increase further if you use AutoCad for 3D work. The "real" modelers I worked with said the AutoCad 3D user interface was torturous. Inventor was supposed to be better, but my information is three years old.

That's the thing about real work, it gets in the way of what we want to do! "The need to make ends meet keeps a man upon his feet."
"Aqaba! By Land!" T. E. Lawrence

R. Peters

tombo
Posts: 334
Joined: Fri May 09, 2008 1:10 am
Location: Washington USA

Post by tombo »

Yes, Inventor has a much better 3D interface. But, it still sucks when dealing with non-rectilinear 3D stuff. The part I really like is that I can drive the parameters from an excel file, so after the model is done I can change the sizes of the features, like thickness & bend radius.
I use the AutoCAD to display the Excel models of B fields I made.
I can use Excel to generate an AutoCAD script to draw the coil shape and then to enter the field vector arrows. It works quite well. Inventor does not do that so well. It's too complicated. I bought my ancient version of AutoCAD from a company that closed down. Otherwise it would be prohibitively expensive for personal use. I'm comfortable with it because I've been using it at work since 1982.

I've seen people bend amazingly intricate and accurate forms from stainless steel and copper tubing.
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

rjaypeters
Posts: 869
Joined: Fri Aug 20, 2010 2:04 pm
Location: Summerville SC, USA

Post by rjaypeters »

rjaypeters wrote:Playing the single-coil game with the two higher-order polyhedra (unnamed here because I can't remember the names) I've already modeled would probably look like a human brain hemisphere or some of the cooler-looking corals). I'm not tempted, I have things to do...and some new fiction to read...
So, I was wrong, I was tempted... Have the pipe-fitters "say hello to my little fren'!", the single-coil rhombicosidodecahedron:

Image Image
"Aqaba! By Land!" T. E. Lawrence

R. Peters

rjaypeters
Posts: 869
Joined: Fri Aug 20, 2010 2:04 pm
Location: Summerville SC, USA

Post by rjaypeters »

In the quest for ever-greater masochism, this is half of a snub dodecahedron:
Image

By itself, it's not a very satisfactory single coil candidate, in fact, it's impossible because some of the vertices have an odd number of segments. But before we get to the real fun, I filled in the large pentagons:
Image

Then, I create what I call the "conjugate," I draw in segments that begin and end at the midpoints of the original polyhedron until I get all the vertices with an even number of segments:

Image

In this case, none of the original or pentagon-filling segments remain.

Then, I create the single-coil path, "pipe" the segments and arcs and we have the results:


Image Image
"Aqaba! By Land!" T. E. Lawrence

R. Peters

icarus
Posts: 819
Joined: Mon Jul 07, 2008 12:48 am

Post by icarus »

Wow rjaypeters you're getting pretty close to getting into tessallation theory here, and on a spherical surface no less. Maybe a drop by some of Penrose's works is in your future. :wink:

imaginatium
Posts: 55
Joined: Mon Jan 05, 2009 10:46 pm

Post by imaginatium »

rjaypeters wrote:In the quest for ever-greater masochism, this is half of a snub dodecahedron:
Image

By itself, it's not a very satisfactory single coil candidate, in fact, it's impossible because some of the vertices have an odd number of segments. But before we get to the real fun, I filled in the large pentagons:
Image

Then, I create what I call the "conjugate," I draw in segments that begin and end at the midpoints of the original polyhedron until I get all the vertices with an even number of segments:

Image

In this case, none of the original or pentagon-filling segments remain.

Then, I create the single-coil path, "pipe" the segments and arcs and we have the results:


Image Image
How about taking this a few steps forward, by including coolant inflow and outflow pipes, and support structures. If you really feel up to the challenge you can create the whole vacuum chamber assembly.

rjaypeters
Posts: 869
Joined: Fri Aug 20, 2010 2:04 pm
Location: Summerville SC, USA

Post by rjaypeters »

icarus wrote:... Maybe a drop by some of Penrose's works is in your future. :wink:
Aww, that dude is way too smart for me. Your mention of Penrose reminded me of Godel, Escher and Bach which I have read, but from which I don''t remember much. Why Penrose reminds me of Godel, Escher and Bach, I don't know.
"Aqaba! By Land!" T. E. Lawrence

R. Peters

rjaypeters
Posts: 869
Joined: Fri Aug 20, 2010 2:04 pm
Location: Summerville SC, USA

Post by rjaypeters »

imaginatium wrote:How about taking this a few steps forward, by including coolant inflow and outflow pipes, and support structures. If you really feel up to the challenge you can create the whole vacuum chamber assembly.
I have designed support structures for a traditional six-coil magrid, but gone no further with it because there is no engineering feedback on what is a good design. Quite rightly, the simulator(s) of whom I am aware have concentrated on the more fundamental aspects of the polywells, e.g. confinement, recirculation, etc. and run into computer hardware limitations. Truly, it wasn't the polywell design that caused the computer hardware limitations to become important.

I have not supplied segment lists for the higher order work I've been doing because the simulators have not been active, to my knowledge. So, why am I creating still? I am hoping the simulators will become active again if they see something interesting.

What I hope for out of the simulators is statistical work for the increasing higher order polyhedra that gives us a hint of the relationship between more coils and confinement (for example) behavior. Serious students and professionals can estimate the costs and difficulties of building ever more complex polywells.

If it turns out confinement for the higher polyhedra is not worth the effort then the simulators will probably have saved humanity unneeded work. OTOH, if the higher polyhedra are worth the effort, that would be really good to know.

Back to your question, I'll design support structures when the simulators ask for them. The only design so far that, sort of, includes support structures and has been simulated is hanelyps "winged" octahedron. See here:

"Quick and dirty"
viewtopic.php?t=2720&postdays=0&postorder=asc&start=255

"Complete" with the simulation linked later on the same page.
viewtopic.php?t=2720&postdays=0&postorder=asc&start=270

[EDIT]And on the snub dodecahedron, it's hard enough to look at just with .jpgs! Who wants to see support structures on top of that ball of snakes? :)
"Aqaba! By Land!" T. E. Lawrence

R. Peters

MSimon
Posts: 14332
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

A continuous loop is going to be hard to cool. And then there is the problem of stiffness without "nubs". Geometry is only part of the problem.

And then there is the magnetic field problem. What does the field look like? i.e. you need the major field generators pointing "in" and the cusps pointing "out". And the field needs to null in the center.
Engineering is the art of making what you want from what you can get at a profit.

rjaypeters
Posts: 869
Joined: Fri Aug 20, 2010 2:04 pm
Location: Summerville SC, USA

Post by rjaypeters »

The pictures I have shown of the "single-coil" configurations include only the basic conductor paths. Cooling and structural supports are omitted for clarity and because I'm pretty clueless about how those things should go.

WRT magnetic field configuration, it is my contention the single-coil looks little different from lots of little torii. viz:

Image

[edit: added better photo]
Image

This single-coil "starts" in the +X (red) coil, trace your eye counter-clockwise, as seen from the outside of the assembly, to where it meets the +Y coil (green), again counter-clockwise to the +Z (dark blue) coil for a short trip around to the -X (gray) coil, around to the -Y (cyan) coil for another short trip around to the -Z (yellow) coil where the trip reverses itself and the current flows back to the +X coil via the four other coils.

This really is nothing other than the trick Tombo, I think, invented some time ago. I've just extended it to polyhedra with more faces. Although, the example above is six faces and Tombo started with [edit] four. [edit] It's simpler to show with a tetrahedron, but easier to model with a "classic" WB.
"Aqaba! By Land!" T. E. Lawrence

R. Peters

imaginatium
Posts: 55
Joined: Mon Jan 05, 2009 10:46 pm

Post by imaginatium »

MSimon wrote:A continuous loop is going to be hard to cool. And then there is the problem of stiffness without "nubs". Geometry is only part of the problem.

And then there is the magnetic field problem. What does the field look like? i.e. you need the major field generators pointing "in" and the cusps pointing "out". And the field needs to null in the center.
That was precisely the point I was trying to make when I presented the challenge. Theoretical models don't always translate into functional structures.
rjaypeters wrote:
imaginatium wrote:How about taking this a few steps forward, by including coolant inflow and outflow pipes, and support structures. If you really feel up to the challenge you can create the whole vacuum chamber assembly.
I have designed support structures for a traditional six-coil magrid, but gone no further with it because there is no engineering feedback on what is a good design. Quite rightly, the simulator(s) of whom I am aware have concentrated on the more fundamental aspects of the polywells, e.g. confinement, recirculation, etc. and run into computer hardware limitations. Truly, it wasn't the polywell design that caused the computer hardware limitations to become important.

I have not supplied segment lists for the higher order work I've been doing because the simulators have not been active, to my knowledge. So, why am I creating still? I am hoping the simulators will become active again if they see something interesting.

What I hope for out of the simulators is statistical work for the increasing higher order polyhedra that gives us a hint of the relationship between more coils and confinement (for example) behavior. Serious students and professionals can estimate the costs and difficulties of building ever more complex polywells.

If it turns out confinement for the higher polyhedra is not worth the effort then the simulators will probably have saved humanity unneeded work. OTOH, if the higher polyhedra are worth the effort, that would be really good to know.

Back to your question, I'll design support structures when the simulators ask for them. The only design so far that, sort of, includes support structures and has been simulated is hanelyps "winged" octahedron. See here:

"Quick and dirty"
viewtopic.php?t=2720&postdays=0&postorder=asc&start=255

"Complete" with the simulation linked later on the same page.
viewtopic.php?t=2720&postdays=0&postorder=asc&start=270

[EDIT]And on the snub dodecahedron, it's hard enough to look at just with .jpgs! Who wants to see support structures on top of that ball of snakes? :)
Up until now i haven't read that thread at all, sitting in the middle of it is the most practical design I've seen to date
Image

This design is structurally well supported, with short segments capable of cooling that sufficiently removes heat. The only line cusps are between the feed tubes, but they are perpendicular to the wiffleball, and so are for all practical purposes point cusps.

I love you enthusiasm for modeling, but if you put out every conceivable configuration under the sun, the truly useful ones get buried in the pack. If you used the same short segment design found here, in higher order polyhedra, you might actually come up with the ultimate polywell design.

Post Reply