magrid configuration brainstorming

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

Moderators: tonybarry, MSimon

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

Nebel has a nondisclosure agreement. It's not up to him.

Ironically, the better WB-8 works the less data we may see.

I think you're right about University funding, though. What does UW spend on IEC fusion already?

These really aren't that expensive. If WB-8 news gets people interested, we could Polywell progarms sprouting up all over.

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

KitemanSA wrote:
rcain wrote: but i believe we are after 'order of magnitude' improvements X 3 - if we are to entertain Art's worst-best-case numerical models predictions/approximations.
Dr.B mentioned a 3-5x improvement in one document that SEEMED to be due to improved sphericity
I know Rick seemed to indicate confinement is the focus. I think they really want to know more about how losses scale with B, and what else might be done to improve losses.

A factor of 8 should tell us a lot about B scaling.

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

Build WB7 (square planform cuboctahedron)
Build WB8 (higher order polygon)
Build big.

They called his rebuilt WB6, WB7. Oh well.
But rather than going on to his WB7, they are building a bigger (I think) WB6. Oh well again! :sigh:
My guess is the funders/reviewers are more worried about B scaling than about dodec improvements.

Have we actually been told WB-8 is a truncube? I can't recall how we arrived at that. It may have been inferred from the fact no one said dodec, and inductive reasoning is always suspect.

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

Post by tombo »

A followup to an objection to the design from a long time ago:
Here are 2 solutions to the planar leak paths leading between the feed lines.
I finally got back to documenting what I was visualizing.
There are still the line cusps at the close approach points. But they should be no worse than for a truncube.
(Sorry about the humongous images but photobucket won't give me a medium size today. It's either this or tiny.)

This one has diverging feedlines for better load bearing.
Image
other views:
http://i299.photobucket.com/albums/mm31 ... tView2.png
http://i299.photobucket.com/albums/mm31 ... tView3.png
http://i299.photobucket.com/albums/mm31 ... tView4.png

This one has parallel feedlines for more constrained fields outside the magrid.
Image
other views:
http://i299.photobucket.com/albums/mm31 ... iew2-1.png
http://i299.photobucket.com/albums/mm31 ... iew3-1.png
http://i299.photobucket.com/albums/mm31 ... BView4.png
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

krenshala
Posts: 914
Joined: Wed Jul 16, 2008 4:20 pm
Location: Austin, TX, NorAm, Sol III

Post by krenshala »

As I understand the discussions previously (and, admittedly, I would not be surprised if I have misunderstood something) wouldn't both of those designs give you four line cusps that go "across" the device from one cluster of feedlines to the other?

Also, why the twist just after leaving the "sphere" of the magrid on the second model?

TallDave
Posts: 3140
Joined: Wed Jul 25, 2007 7:12 pm
Contact:

Post by TallDave »

Ack! Triangles! [runs away]

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

Post by tombo »

The line cusps going across would be the same as for a truncube but only 4 of them not 8 of them.
But, yes they exist and I am not entirely happy with that.
The twist closes off a planar cusp (4 of them) similar to the original "cusp device".
Notice, both pictures have the twist.
Triangles exist in the corners of the truncube too between the round or the square coils. Embrace the shadow of what the artists call negative space.
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Post by D Tibbets »

With the opposing triangles being rotated like this - one base faces a point of the opposing triangle, wont the magnetic surfaces inside be distorted with waffling. I think that would introduce local concave surfaces in the magnetic field that are bad.


Dan Tibbets
To error is human... and I'm very human.

Randy
Posts: 82
Joined: Mon Sep 20, 2010 5:40 am
Location: Texas

Post by Randy »

Optimum Magrid Configuration

I know this is a very old thread but hopefully some group members will respond to my questions.

Hi everyone,
My name is Randy. This is my first post to the group. Back in the ‘70’s I was trained in the US Navy’s nuclear power program. I learned mainly about nuclear fission but also a little about fusion. I was fascinated by fusion and read everything I could find about it. So far I’ve been very disappointed by the mainstream tokomak progress. I’ve been aware of the Farnsworth fusor concept and its grid problems for many years.
Recently I stumbled onto Dr. Bussard’s polywell concept and was quite impressed by his ingenious virtual cathode idea. However, I’m no plasma physicist so please bear with me.

I have a serious question as to why, what appears to me to be the optimum magrid configuration (octahedron) is not being used in any of the pictures of actual magrid prototypes I’ve seen built so far.

First, let me reproduce some definitions from the Theory Section F.A.Q. pertaining to cusps in magnetic containment fields:

1) Point cusps occur at the center of magnetic coils.
2) Line cusps occur when conductors with currents in opposite directions are placed beside one-another.
3) Funny Cusps occur when multiple pairs of fields with opposing directions meet at a common vertex.

And second, here are some quotes from Dr. Bussard’s Google techtalk of November 9, 2006:
“There is only ONE configuration that works, and that is the one that we patented. It is a configuration that is a polyhedron where the coils are all on the edges of the polyhedron (by this I believe Dr. Bussard means that the magnet coil wires are wound along the edges of each polygon face of the polyhedron such as to produce either a north or south field coming out of that polygon face), and the polyhedron has the property that there are an EVEN NUMBER OF FACES around every vertex so that ALTERNATE FACES are north, south, north, south, north, south.

If you look at the CUBE which constitutes the normal biconic cusp, it only has THREE FACES around every vertex, so you have the LINE CUSP PROBLEM. The only thing we could find to solve it is to make a system that is quasi spherical with no magnetic monopole (by this I believe Dr. Bussard means that the magrid configuration doesn’t have all north or south poles facing toward the core), so you have to do it from the surface, so you ONLY HAVE POINT CUSP LOSSES.

...The first problem is the electron losses necessary to drive the well; how many electrons you can lose and still make fusion. The MAIN LOSSES OCCUR THROUGH MAGNETIC CUSPS to the walls.”
End of Dr. Bussard’s quotes.

So, in order to achieve the deepest possible potential well in a magrid, you need to keep as many core electrons from getting lost through the magnetic cusps as possible. My first guess would be that since a line is composed of an infinite number of points that eliminating all line cusps would be a definite top priority in any magrid design. i.e., apparently (from Dr. Bussard’s comments) line cusps easily allow the loss of many more electrons than either point or funny cusps do.

It is very clear to me why Dr. Bussard basically implied that the CUBE magrid configuration has the line cusp problem. Just draw, on a flat sheet of paper, three line segments 120 degrees apart that intersect at a single vertex. Then draw three magnetic poles on the faces around the vertex, either N,S,S or N,N,S. There is simply no way to alternate the pole placement without having two similar poles touching each other along one line segment of the vertex. By definition (see F.A.Q. above) this introduces a line cusp along the line segment separating the two similar pole faces (because the coil currents flow in opposite directions along the line segment separating the two similar poles). This same line cusp problem will necessarily exist along at least one line segment of each of the eight vertices of any cubic magrid coil configuration. If you analyze the problem you will quickly see that for ANY arrangement of the cubic (six coil) magrid configuration that you can only achieve a MINIMUM of four line cusps, either two north::north line cusps and two south::south line cusps, or four north::north line cusps or four south::south line cusps… Just draw a square on a flat sheet of paper and alternately label each side N,S,N,S (imagine the sides of the square as being the pole faces of magnets lying on the sheet of paper). Then imagine placing the north pole face of a magnet on the front side of the paper and then placing the south pole face of a magnet on the back side of the paper (with each front and back magnet face covering the drawing). You get two north::north line cusps from the page front magnet and two south::south line cusps from the page back magnet.

From the patent definition it’s clear that the optimum magrid configuration is an octahedron not a cube. Another member in this same thread (I believe it was ‘tombo’) recognized this back in 2008. But, in my interpretation, tombo had a problem with understanding the significance of the different kind of magnetic cusps involved with each design.

Consider the octahedron (two pyramids connected together at their bases), each vertex of the octahedron has four faces (an even number) surrounding it as specified in the patent. You can easily alternate the faces of the octahedron N,S,N,S,N,S,N,S without ever having two similar poles touching each other along any line segment separating any two pole faces. This means that the octahedron magrid configuration eliminates all line cusps. The octahedron magrid configuration features only eight point cusps (one at the center of each coil) and six funny cusps (one for each vertex of the octahedron) and no line cusps. It is now very clear to me why Dr. Bussard patented this magrid concept.

NNNNNN
\\\\\\\\\\
---------- <- Line Cusp
//////////
NNNNNN

N N
\ | /
--.-- <- Point cusp
/ | \
N N

N __ S
\ | /
|--.--| <- Funny cusp
/ | \
S __ N


As I said in my introduction, I’m no plasma physicist, but I can’t imagine that a funny cusp could introduce more core electron losses into a magrid design than a line cusp would.

I’m just guessing here, but I think, as far as electron loss from the core is concerned, that line cusps are the worst, funny cusps are next and point cusps are the least significant.

The whole point in building a magrid is to develop the deepest potential well (virtual cathode) as possible. So if you’re going to build a prototype magrid, why use a cubic configuration (using six coils – and necessarily introducing at least four line cusps into your design) when you can just fabricate two more coils to build an octahedral magrid that only introduces eight point cusps and six funny cusps into your design. You could still use circular coils on each face of the octahedron for the prototype.

As I stated earlier,
I have a serious question as to why, what appears to me to be the optimum magrid configuration (octahedron) is not being used in any of the pictures of actual magrid prototypes I’ve seen built so far.

The only reason I can guess as to why all the pictures of Dr. Bussard’s magrid prototypes depict cubic designs is to protect their main (most likely octahedral) magrid concept from any serious competition.

~Randy

WizWom
Posts: 371
Joined: Fri May 07, 2010 1:00 pm
Location: St Joseph, MO
Contact:

Post by WizWom »

tombo wrote:A followup to an objection to the design from a long time ago:
Here are 2 solutions to the planar leak paths leading between the feed lines.
I finally got back to documenting what I was visualizing.
There are still the line cusps at the close approach points. But they should be no worse than for a truncube.
(Sorry about the humongous images but photobucket won't give me a medium size today. It's either this or tiny.)
well, you DON'T want the outgoing magnetic fireld to be the same size as the inward pointing magnetic field. You lose the whole whiffleball containment then.
Wandering Kernel of Happiness

Randy
Posts: 82
Joined: Mon Sep 20, 2010 5:40 am
Location: Texas

Post by Randy »

Sorry,
I should have drawn the point cusp representation as:

N N
\ | /
N--.--N <- Point cusp
/ | \
N N

~Randy

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

Welcome to the nut-house, Randy..... here you will have endless opportunity to debate as many speculations as you like over a machine from which the answers will arise from experimentation rather than theorising and jaw-wagging, as its operation does not appear to quite lend itself to normal analysis with normal physics. Unfortunately, the only folks who are doing such experiments don't want you to know the details...

Sorry, I don't see it. Can you draw arrows showing the current in each edge of each face of your octahedron? Would you not end up with a couple that do not have currents running around in the same direction?

Randy
Posts: 82
Joined: Mon Sep 20, 2010 5:40 am
Location: Texas

Post by Randy »

chrismb wrote:Welcome to the nut-house, Randy..... here you will have endless opportunity to debate as many speculations as you like over a machine from which the answers will arise from experimentation rather than theorising and jaw-wagging, as its operation does not appear to quite lend itself to normal analysis with normal physics. Unfortunately, the only folks who are doing such experiments don't want you to know the details...

Sorry, I don't see it. Can you draw arrows showing the current in each edge of each face of your octahedron? Would you not end up with a couple that do not have currents running around in the same direction?
Looking into one vertex of the octahedron you would see something like this:
Image
From the F.A.Q. definition that, “Line cusps occur when conductors with currents in opposite directions are placed beside one-another.”
You can easily see that the coil currents flow in PARALLEL directions along the line segments separating any two opposite poles. This fact shows that no line cusps will be formed along these line segments.
A similar symmetric picture will be viewed looking into any vertex of the octahedron.
Hope this diagram clears it up for you.
~Randy

Randy
Posts: 82
Joined: Mon Sep 20, 2010 5:40 am
Location: Texas

Post by Randy »

Randy wrote:
chrismb wrote:Welcome to the nut-house, Randy..... here you will have endless opportunity to debate as many speculations as you like over a machine from which the answers will arise from experimentation rather than theorising and jaw-wagging, as its operation does not appear to quite lend itself to normal analysis with normal physics. Unfortunately, the only folks who are doing such experiments don't want you to know the details...

Sorry, I don't see it. Can you draw arrows showing the current in each edge of each face of your octahedron? Would you not end up with a couple that do not have currents running around in the same direction?
Looking into one vertex of the octahedron you would see something like this:
Image
From the F.A.Q. definition that, “Line cusps occur when conductors with currents in opposite directions are placed beside one-another.”
You can easily see that the coil currents flow in PARALLEL directions along the line segments separating any two opposite poles. This fact shows that no line cusps will be formed along these line segments.
A similar symmetric picture will be viewed looking into any vertex of the octahedron.
Hope this diagram clears it up for you.
~Randy
Sorry, instead of saying PARALLEL directions, I should have said: "in the SAME directions."
~Randy

chrismb
Posts: 3161
Joined: Sat Dec 13, 2008 6:00 pm

Post by chrismb »

But they aren't going in the same direction, wrt a radial sight looking into the coils. Isn't it necessary to have all currents going around clockwise, or ac, when looking towards the centre? Else the mag fields will simply curl around the edges and there would be no 'cusp-containment'? What you'd be generating then is a set of inductively coupled plasmas around each edge?

Maybe I got it wrong, but there it is as I thought it was meant to be?

Post Reply