STATEMENT OF OBJECTIVES from RFP

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

Moderators: tonybarry, MSimon

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

Post by TallDave »

No, Simon is right, there are cusps everywhere fields from the coils meet.

Bussard obviously knew about cusps. He put the interconnects right in front of the semi-line cusps on WB-6, and WB-6 worked extremely well. So did Nebel's team, and WB-7 is also reportedly working quite well.

I have to believe there was some reasoning that went into that decision. It may have to do with current flow or how the small magnetic from the shielded (?) interconnects interacted with the larger coil fields.
They should become more point-like with a "square" plan form coils set,
You can't generate a square field, though, so a square plan won't conform to the field shape. The result will be excess cross-field diffusion.

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

MSimon wrote:
Nope, the cusps are the center points of the "In" coils and real or virtual "Out" coils.
I take it you are unfamiliar with line cusps. Art Carlson seems to think they are a fact of life. Perhaps you need to have a go with him and explain why line cusps are a figment of his imagination. I'm sure he will be most grateful for the education.
I think you take it wrong. I believe I understand line cusps well enough and nowhere have I ever claimed that line cusps are a figment of anyone's imagination. Please do not debase this discussion with such insults.
As I understand it, line cusps are what you get when two same polarity fields meet, like what you would get if you put two round coils side by side or even face to face. When you put 6 of them in a quasi cube, you get quasi-line cusps because the virtual corner magnets of opposite polarity do "funny" things to them. I take it you aren't comfortable with DrB's "funny cusps". (Sorry, turn about and fair play and the like; please forgive me).

Funny cusps are what you get when an even number of opposite polarity fields meet like what you would get with real/real cuboctohedral magnets. The funny cusps are more "holes", singularities in the field, rather than cusps which are pinches (or perhaps more accurately rapid changes in direction) in a unidirectional field.

So what we have between the current generation of WBs are quasi-line cusps, perhaps you might think of the as "funnified" line cusps, and what we would have in a configuration more like what DrB wanted would be funny cusps. Clear?

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

TallDave wrote: No, Simon is right, there are cusps everywhere fields from the coils meet..
Anywhere that fields from SAME POLARITY coils meet. Where opposite polarities meet, they form on extended field. Where TWO PAIRS of opposites meet, they form a "funny" cusp (which isn't really a cusp in the sense that line and point cusps are.
TallDave wrote:Bussard obviously knew about cusps. He put the interconnects right in front of the semi-line cusps on WB-6, and WB-6 worked extremely well. So did Nebel's team, and WB-7 is also reportedly working quite well.
But not so well that they don't want to pay to move the nub, which seems is the prime stated purpose of this new contract.
TallDave wrote:I have to believe there was some reasoning that went into that decision. It may have to do with current flow or how the small magnetic from the shielded (?) interconnects interacted with the larger coil fields. ..
IBID re: moving / changing the nubs.
TallDave wrote:
They should become more point-like with a "square" plan form coils set,
You can't generate a square field, though, so a square plan won't conform to the field shape. The result will be excess cross-field diffusion.
Not looking for a "square" field, just a substantial restriction in the size of the funny cusp. Small corners make the "In" field and the "Out" field (aka real and virtual) meet with a more uniform field, and SHOULD result in the smallest possible "funny" cusp. This SHOULD restrict the external population of electrons which some insist is essential. (I think it desireable, but not essential)

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

Post by TallDave »

Anywhere that fields from SAME POLARITY coils meet.
Which is everywhere there is a side.
Where opposite polarities meet, they form on extended field.
They cancel at the center. You can't contain anything that way.
But not so well that they don't want to pay to move the nub, which seems is the prime stated purpose of this new contract.
Well, what it actually says is:
3.1.1.2. The review shall primarily investigate the effects of parallel electron heat loss to the coil joints with respect
to plasma stability and electron confinement time.
3.2 TESTS
3.2.1 The contractor will modify/upgrade the existing wiffleball #7 (WB-7) device by installing compact, high
temperature coil joints to investigate the electron parallel heat loss. This modified device shall hereafter be
identified as Wiffleball #7.1 (WB-7.1).
That doesn't say they're moving them. I'd say it's an open question where they are and why. Maybe Rick will enlighten us at some point.
Not looking for a "square" field, just a substantial restriction in the size of the funny cusp.
At the expense of creating a nonconformal coil shape, meaning you're going to have electron flows going near or through your coils, which is much worse than cusp losses. Again, Bussard knew a lot about cusps, and didn't design it that way. In fact, he went to considerable effort to create conformal containers, and insisted future machines must have them. I would carefully review Bussard's statements before assuming what you're doing will improve on his design. He was a clever guy and spent two decades working on the concept through many iterations.

From Valencia:
Robert Bussard wrote:...it was always known that conformal magnet coil
cans/casings were the only way to avoid B field intersect
with their surfaces, but since it was difficult and costly to
build such container shapes,
...
The need for magnetic field coil containing structures
to be conformal with the B fields they produce, to
avoid excessive electron impact losses (as above)
...
This requirement has two main consequences: (a) All
coil containers/casings must be of a shape conformal to the
B fields produced by their internal current conductors, and;
(b) The finite size of real coils forces design so that no
coils/containers can ever be allowed to touch each other, but
all corners MUST be spaced at some distance from the
adjacent coils, to avoid B field intercept.
This SHOULD restrict the external population of electrons which some insist is essential. (I think it desireable, but not essential)
Without the B field holding in the plasma, you will never get close to net power densities. Remember, power is B^4 R^3. With no wiffleball, there's no B^4 scaling. You'd need a truly gigantic reactor to get to 100MW on r^3 alone.

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

[/quote]
TallDave wrote:
Anywhere that fields from SAME POLARITY coils meet.
Which is everywhere there is a side.
Not so! The whole basis of the polywell is to have alternating polarity magnets. With WB-6, this was achieved by having 6 real magnets and 8 virtual magnets. They are side by side. The field comes in the real coil "In" point cusp, extends ALL THE WAY ACROSS to the two magnet coils (one real, the other virtual) and goes out the virtual "Out" point cusp. The only place that two SAME POLARITY fields exist side-by-side is the space between the various real magnets. And the cusps that are created are only as LONG as they are (i.e. quasi-linear) because the virtual field is too far away. This is one thing I think DrB wanted to correct with the "square" plan form magnets.
TallDave wrote:
Where opposite polarities meet, they form on extended field.
No, they cancel. You can't contain anything that way.
Language difficulty. If there is an "out" coil directly in front of the "in" coil, (read North and South if you wish) the field just continues through the second coil and you get a longer coil. There is no line cusp between them, but they sure as heck don't cancel. What cancels is when you try to have an IN cusp and an Out cusp in the same space. That is, by definition, a "funny" cusp.
TallDave wrote:
But not so well that they don't want to pay to move the nub, which seems is the prime stated purpose of this new contract.
Well, what it actually says is:
3.1.1.2. The review shall primarily investigate the effects of parallel electron heat loss to the coil joints with respect to plasma stability and electron confinement time.
3.2 TESTS
3.2.1 The contractor will modify/upgrade the existing wiffleball #7 (WB-7) device by installing compact, high temperature coil joints to investigate the electron parallel heat loss. This modified device shall hereafter be
identified as Wiffleball #7.1 (WB-7.1).
That doesn't say they're moving them. I'd say it's an open question.
I'll give you that one. They just want to see how bad it really is. Sounds like they don't think they are just fine.
TallDave wrote:
Not looking for a "square" field, just a substantial restriction in the size of the funny cusp.
At the expense of creating a nonconformal coil shape, meaning you're going to have electron flows going through your coils, which is much worse than cusp losses. Again, Bussard knew a lot about cusps, and didn't design it that way. In fact, he went to considerable effort to create conformal containers. I would carefully review Bussard's statements before assuming what you're doing will improve on his design. He was a clever guy and spent a decade and a half working on the concept.

From Valencia:
Robert Bussard wrote:...it was always known that conformal magnet coil cans/casings were the only way to avoid B field intersect with their surfaces, but since it was difficult and costly to
build such container shapes,
...
The need for magnetic field coil containing structures to be conformal with the B fields they produce, to avoid excessive electron impact losses (as above)
...
This requirement has two main consequences: (a) All coil containers/casings must be of a shape conformal to the B fields produced by their internal current conductors, and; (b) The finite size of real coils forces design so that no coils/containers can ever be allowed to touch each other, but all corners MUST be spaced at some distance from the adjacent coils, to avoid B field intercept.
In truth, I fail to see the relevance of this quotation. Conformal cans mean that the cross section of the magnet needs to be round(ish). That has NOTHING to do with the plan form. The early WBs had round plan form square cross section magnets. WB-6 went to round plan form round cross section magnets. DrB wanted WB7 to be a "square" plan form, round cross section magnet. But please note that in his quotation above (which I have highlighted) he said CORNERS of the adjacent magnets. You can't have a corner in a round plan form. But he wanted square, (or pentagonal) plan forms with corners. I think he wanted that to limit the length of the quasi-linear cusp that results from the round plan form.
Last edited by KitemanSA on Mon Mar 09, 2009 3:12 am, edited 1 time in total.

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

TallDave wrote:
This SHOULD restrict the external population of electrons which some insist is essential. (I think it desireable, but not essential)
Without the B field holding in the plasma, you will never get close to net power densities. Remember, power is B^4 R^3. With no wiffleball, there's no B^4 scaling. You'd need a truly gigantic reactor to get to 100MW on r^3 alone.
Hey, looks like you edited your post and added a bit. Either that or it somehow disappeared from the original quote when I quoted the first time.

This discussion was held before and while it is nice to have the electrons inside the coils, it helps indirectly against arcing, it is not necessary. If it were, there would never have been even a theoretical Elmore Tuck Watson fusor. YES I KNOW that ETWs can't break even, but that has nothing to do with the external population of electrons. It is because the grid intercepts too many. The electron lifetimes in an ETW are just too dang short. The only thing necessary for success is that the lifetime of the electron is very long, inside or both, and that the overall density in the inside is at a specific level. External electrons shouldn't effect the internal density.

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

Post by TallDave »

Not so! The whole basis of the polywell is to have alternating polarity magnets.
Right, with alternating sides facing each other. This is why only a polygon with an even number of vertexes at each side can work.
If there is an "out" coil directly in front of the "in" coil, (read North and South if you wish) the field just continues through the second coil and you get a longer coil. There is no line cusp between them, but they sure as heck don't cancel.
You have an S facing an N. What is the field at the center? Do you see the problem there?

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

Post by TallDave »

If it were, there would never have been even a theoretical Elmore Tuck Watson fusor
Why not? Just because it can't make net power doesn't mean it can't make interesting amounts of fusion.
YES I KNOW that ETWs can't break even, but that has nothing to do with the external population of electrons. It is because the grid intercepts too many. The electron lifetimes in an ETW are just too dang short.
Well, of course. Arcing is only a problem for ETWs if you have a magical transparent grid. It's like worrying about decorating your hyperspeed vehicle's cockpit before you've solved the problem of an engine.
Last edited by TallDave on Sat Mar 07, 2009 2:49 am, edited 1 time in total.

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

TallDave wrote:
Not so! The whole basis of the polywell is to have alternating polarity magnets.
Right, with alternating sides facing each other. This is why only a polygon with an even number of vertexes at each side can work.
Again, language difficulty. The magnets that face each other across the MaGrid are either both IN or both OUT. It is the adjoining magnets that have IN/OUT pairs. At the vertices (which the round plan form magnets don't really have so aren't true polywells) there should be an alternating IN/OUT/IN/OUT... which cancels the fields and creates the "funny" cusp.
TallDave wrote:
If there is an "out" coil directly in front of the "in" coil, (read North and South if you wish) the field just continues through the second coil and you get a longer coil. There is no line cusp between them, but they sure as heck don't cancel.
You have an S facing an N. What is the field at the center? Do you see the problem there?
If you have two simple coils, axially aligned such that the south of one faces the north of the other, the field (using the flow north to south convention) just exits the one coil, continues into and thru the next and the flows around to the south side of the first magnet. Makes it look like a bar magnet. It was the basis of one of the failed fusion machines of the past. Tried to use a pinching of the fields at each end to confine the plasma. Didn't work.

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

Post by TallDave »

Right, the plan form doesn't matter:
Robert Bussard wrote:This is the principal criterion for design and
construction of any real, finite material coil and system, no
matter the plan-form SHAPE of the coils, which is of no
major significance (i.e. round, square, polygonal or
triangular, etc).
But he wanted square, (or pentagonal) plan forms with corners.
What are you basing this on? He actually says he doesn't care, as long as they're conformal.
You can't have a corner in a round plan form.
Nonetheless Bussard makes references to "corners" of WB-6.

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

TallDave wrote:
If it were, there would never have been even a theoretical Elmore Tuck Watson fusor
Why not? Just because it can't make net power doesn't mean it can't make interesting amounts of fusion.
Your ealier statement seemed to imply that you thought it was the RELATIVE density of electrons inside and out that controls the fusion power. If that were true, ET&W would surely have known that their machine could never get a significant RELATIVE density so why try? But it is NOT the RELATIVE density that controls the power, it is absolute above neutral count that controls power. What controls NET power is power vs loss, i.e. the electron lifetime, not the relative density. Arcing, it plays a part in but with no neutrals it shouldn't matter.
Indeed, your own statement supports my point. ETWs make interesting amounts of fusion. They have an almost 1:1 ratio of internal and external electrons. But they have short electron lifetimes. The Poywell's prime benefit is that it protects the electrons against loss. True, it also increases the relative density, but that is gravy, not necessary.
TallDave wrote:
YES I KNOW that ETWs can't break even, but that has nothing to do with the external population of electrons. It is because the grid intercepts too many. The electron lifetimes in an ETW are just too dang short.
Well, of course. Arcing is only a problem for ETWs if you have a magical transparent grid. It's like worrying about decorating your hyperspeed vehicle's cockpit before you've solved the problem of an engine.
Not sure what you are getting at here, but it sure sounds nasty.

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

Post by TallDave »

Your ealier statement seemed to imply that you thought it was the RELATIVE density of electrons inside and out that controls the fusion power.
No, of course not. That would make absolutely no sense.
If that were true, ET&W would surely have known that their machine could never get a significant RELATIVE density so why try?
Heh. Why build them at all? They must have known the grid problem was unsolvable. But it was interesting. And look what it led to.
Robert Bussard wrote:Wiffle Ball behavior is of value (and is essential)
ONLY to establish the density ratio from the machine
interior to its exterior, and this is important ONLY to assure
suppression of Paschen arc breakdown outside, which
destroys the electron injection drive and well potential.

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

[/quote]
TallDave wrote:
it is absolute above neutral count that controls power.
I'm not sure what that means, but I hope it translates to "ion density and volume."
Yup, via well depth.
TallDave wrote:
If that were true, ET&W would surely have known that their machine could never get a significant RELATIVE density so why try?
Heh. Why build them at all? They must have known the grid problem was unsolvable. But it was interesting. And look what it led to.
Robert Bussard wrote:Wiffle Ball behavior is of value (and is essential)
ONLY to establish the density ratio from the machine
interior to its exterior, and this is important ONLY to assure
suppression of Paschen arc breakdown outside, which
destroys the electron injection drive and well potential.
Seems you found the quote to support my position. Thanks! The other way to suppress arcing is thru the neutrals, which are the primary cause. So, no neutrals and the system would still work even with a lot of external electrons. Wiffleball improves things a LOT (hallelujah) but it is the MaGrid protecting the electrons which is essential.
I wish I could remember what started this argument in the first place.

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

Post by TallDave »

Seems you found the quote to support my position.
He said it was essential to prevent arcing. Was that your position? Good, we agree then.
The other way to suppress arcing is thru the neutrals, which are the primary cause.
Presumably if that were easy, we wouldn't bother with a wiffleball.
improves things a LOT (hallelujah) but it is the MaGrid protecting the electrons which is essential.
Did you read the quote? He said the density ratio was also "essential."
Not sure what you are getting at here, but it sure sounds nasty.
You were asking why the ETW guys didn't worry about arcing. My guess is they knew the grid losses were so high it wasn't worth worrying about what would happen at energies an ETW could never reach.
Last edited by TallDave on Sat Mar 07, 2009 3:45 am, edited 1 time in total.

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

TallDave wrote: Did you read the quote? He said the density ratio was also "essential."
To prevent arcing (which by the way DrN seems to believe is not the big deal DrB did, but I may have mis-read his post). It is the neutrals that are causitive, the electrons are assistive. Too many neutrals, and you will have arcing no matter the relative density. No neutrals and the relative density won't matter. The relative density is only "essential" in that middle case. JMO.

Post Reply