from confinement to ignition to (continuous?) burn

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

Moderators: tonybarry, MSimon

Post Reply
rcain
Posts: 992
Joined: Mon Apr 14, 2008 2:43 pm
Contact:

from confinement to ignition to (continuous?) burn

Post by rcain »

on transitioning from confinement to ignition to (continuous?) burn:

can anyone explain to me, what the current plan is to achieve 'ignition' and 'sustained' fusion in the WB, and what happens to our containment model once these states obtain?

Just as some background, Ive been reading through 'Aneutronic fusion in a degenerate plasma' - S. Son ., N.J. Fisch, 2004. ( http://w3.pppl.gov/~fisch/fischpapers/2 ... PLA_04.pdf ). This paper seems 'kinder' to the idea that IEC is capable of achieving ignition than Rider (who we largely seem to have discredited), and likewise favours p-B fuel regime; nevertheless, it puts some pretty 'mean' constraints on what must be physically established (in terms of density, relative velocity of ions and electrons, and thermalisation), in order to get where we want to be and sustain it. Overall, its conclusions are still pessimistic.

On a more simplistic level, i am really wondering what stops the core (or the well) simply 'blowing out' as soon as any appreciable fusion starts happening. Also, how our confinement recipe copes with intermediate and final products.

can anyone provide a clear picture of how the WB should behave through these phases?

thanks for any insight...

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

Post by TallDave »

I don't think polywells achieve "ignition" in the sense of keeping themselves thermally excited enough to keep fusing like a tokamak. It's more a question of how much fusion you get at a given drive well depth and ion focus versus what it costs you to maintain that depth/focus.

In a D-D BFR the neutrons generated have no effect on the electrostatic balance (they's neutral). p-B11 BFRs may have issues with alpha sputtering.

I think it's probably impossible to say exactly what will happen at high energies without experimental data.

jmc
Posts: 427
Joined: Fri Aug 31, 2007 9:16 am
Location: Ireland

Post by jmc »

This paper has nothing to do with Todd-Rider's assertions which contained no quantum phenomena and nothing to do with Polywells which are insufficiently dense to achieve fermi degenaracy.

It does however open up the possiblity of igniting and producing aneutronic fusion in a laser inertial confinement reactor at super high densities.

Art Carlson
Posts: 794
Joined: Tue Jun 24, 2008 7:56 am
Location: Munich, Germany

Re: from confinement to ignition to (continuous?) burn

Post by Art Carlson »

rcain wrote:Rider ... who we largely seem to have discredited
What do you mean by this?

Munchausen
Posts: 226
Joined: Mon Sep 17, 2007 5:36 pm
Location: Nikaloukta

Post by Munchausen »

What do you mean by this?
rnebel



Joined: 24 Dec 2007
Posts: 41

Posted: Wed Apr 30, 2008 6:23 pm Post subject:

--------------------------------------------------------------------------------

to JMC:

Since you are worried about Rider, let me suggest the following exercise. Let's assume that a Polywell reactor is in the wiffleball mode, namely that:

n*kBolt*Te = B**2/(2*mu0)

to make it simple, let's use mks units and assume B = 10 Tesla, mu0 =4.0e-7*pi, Te = 1.0e4 eV and kBolt = 1.6022e-19 Joules per eV.

Calculate what n is and compare it to the ITER value at

http://www.iter.org/a/index_nav_4.htm

Tell me what you get.

rnebel



Joined: 24 Dec 2007
Posts: 41

Posted: Thu May 01, 2008 4:07 pm Post subject:

--------------------------------------------------------------------------------

JMC and MSimon:
Actually, you need to click on “read more” under the design section, then “main parameters” then on the “more” button. What you will find is that the average density of ITER is ~ 1.0e20/m**3. If you use the formula I sent you for the Polywell, you will get a density ~ 2.5e22/m**3. The upshot of this is that the Polywell has a power density that is ~ 62500 times bigger than ITER EVEN IF THERE IS NO ION CONVERGENCE! Thus, a Polywell should far outperform a Tokamak even with a constant density Maxwellian plasma. Even if Rider and Nevins were correct (which Chacon has pretty clearly shown they aren’t) this isn’t a show stopper. It has a lot more significance for Hirsch/Farnsworth machines that have low average densities than it does for the Polywell.
The best analogy that I can think of is that the wiffleball mode is the jet engine and the ion convergence is the afterburner. The 2.5e22/m**3 density is what the Polywell should have on the edge, and then it hopefully goes up a few orders of magnitude as it goes into the interior. I don’t mean to imply that ion convergence isn’t important. This power density boost is what enables the Polywell to be built in small attractive unit sizes and to easily use advanced fuels.
However, the wiffleball mode is essential and the ion convergence simply makes things better. If we can’t get the wiffleball, then we can kiss our behinds goodbye. That’s why we are focused on achieving the wiffleball and we aren’t paying any attention to Rider and Nevins. They’re just a distraction. Does this kind of make sense?

Art Carlson
Posts: 794
Joined: Tue Jun 24, 2008 7:56 am
Location: Munich, Germany

Post by Art Carlson »

Munchausen wrote:
What do you mean by this?
rnebel



Joined: 24 Dec 2007
Posts: 41

Posted: Wed Apr 30, 2008 6:23 pm Post subject:

--------------------------------------------------------------------------------

to JMC:

Since you are worried about Rider, let me suggest the following exercise. Let's assume that a Polywell reactor is in the wiffleball mode, namely that:

n*kBolt*Te = B**2/(2*mu0)

to make it simple, let's use mks units and assume B = 10 Tesla, mu0 =4.0e-7*pi, Te = 1.0e4 eV and kBolt = 1.6022e-19 Joules per eV.

Calculate what n is and compare it to the ITER value at

http://www.iter.org/a/index_nav_4.htm

Tell me what you get.

rnebel



Joined: 24 Dec 2007
Posts: 41

Posted: Thu May 01, 2008 4:07 pm Post subject:

--------------------------------------------------------------------------------

JMC and MSimon:
Actually, you need to click on “read more” under the design section, then “main parameters” then on the “more” button. What you will find is that the average density of ITER is ~ 1.0e20/m**3. If you use the formula I sent you for the Polywell, you will get a density ~ 2.5e22/m**3. The upshot of this is that the Polywell has a power density that is ~ 62500 times bigger than ITER EVEN IF THERE IS NO ION CONVERGENCE! Thus, a Polywell should far outperform a Tokamak even with a constant density Maxwellian plasma. Even if Rider and Nevins were correct (which Chacon has pretty clearly shown they aren’t) this isn’t a show stopper. It has a lot more significance for Hirsch/Farnsworth machines that have low average densities than it does for the Polywell.
The best analogy that I can think of is that the wiffleball mode is the jet engine and the ion convergence is the afterburner. The 2.5e22/m**3 density is what the Polywell should have on the edge, and then it hopefully goes up a few orders of magnitude as it goes into the interior. I don’t mean to imply that ion convergence isn’t important. This power density boost is what enables the Polywell to be built in small attractive unit sizes and to easily use advanced fuels.
However, the wiffleball mode is essential and the ion convergence simply makes things better. If we can’t get the wiffleball, then we can kiss our behinds goodbye. That’s why we are focused on achieving the wiffleball and we aren’t paying any attention to Rider and Nevins. They’re just a distraction. Does this kind of make sense?
That's not what I would call "discredited". Nebel says parenthetically that he thinks that Chacon has shown the Rider is wrong, but it is not clear in what point. The rest of the post is about power density, but Rider talks mostly about power balance.

When it comes to power density, Nebel is playing a bit fast and loose with the numbers. When he suggests using 10 T for the polywell field but then compares it to ITER with 5 T on axis, he is giving the polywell an illegitimate factor of 16 advantage. I'd start the calculation another way and say that ITER has a 2.5% beta, and the polywell is supposed to have 100% beta, so there is a factor of 40 in the pressure. Since the polywell runs about ten times hotter, that is about a factor of 4 in the density, leading to a factor of 16 - not 62500 - advantage in the power density. My calculation is an underestimate because I did not take the increase in cross section with temperature and some other fine points into account, but that is not really the issue. No one disagrees that, all other things being equal, a high beta machine will have a much higher power density than a tokamak. The point is whether the power you have to put in to make it run is even higher than the fusion power you get out.

rcain
Posts: 992
Joined: Mon Apr 14, 2008 2:43 pm
Contact:

Post by rcain »

Hi

To TallDave:: a different definition of 'ignition' - thats a useful idea to work with - thanks. i still do wonder though, how we maintain the well (and the core) whilst at the same time producing and extracting a 'useful' amount of energy out of it. If some degrees of asymmetry were to be incorporated into the design I could probably get my head around possible ways to control it, but since its (so far) symmetrical, i cant help thinking that scattering will be an 'uncontrolled' problem, also as you suggest alpha sputtering. I also wonder about the 0.1-0.2% neutrons likely (from side reactions) will have a similar disruptive effect at the core.

To jmc:: yes, I realise that Fermi degeneracy wouldn't appear to have any dominant role in current Polywell design, but I wonder whether it doesn't play any role at all. Also, the article refers generally to Laser OR Ion beam collision (albeit at extreme densities). Son/Fisch does make reference to Rider: vis::

'Rider [T.H. Rider, Phys. Plasmas 4 (1997) 1039.]
pointed that the fusion power of an aneutronic
plasma is substantially smaller than the minimum
recirculating power to maintain the non-equilibrium
condition (high Ti and low Te), which diminishes the
prospect for utilizing aneutronic fuel. However, his
derivation is under the assumption that the two-body
effects are proportional to d3x |n(x)|2. In an ultra
dense plasma, we showed the stopping power is not
proportional to the electron density, breaking that assumption,
and thus avoiding the negative conclusion
by Rider.'

Another of Riders 'assumptions' deemed inappropriate (see also my point below). True, Rider doesn't treat things quantum, to the best of my knowledge, but I think the point here was more general.

One other interesting set of calculations Son/Fisch performs, on system gain (albeit wrt degenerate plasma):

"The feasibility as a reactor for either of these fuels (p-B11, D-He3)
is low because the gain is smaller than 20, and
more than 200 is usually required [36]. We note that
the gain can be as large as 1000 in D–T fuel [32].
The creation of a hot spot for the fast ignition [33]
might be a way of improving the gain substantially."

- similarities with the POPS approach?


To Art Carlson:: I obviously didnt mean to imply Rider himself has been discredited, simply that his early critique of Polywell-WB IEC based fusion seems to have been at least equally-well refuted by many others. In particular his paper: 'A general Critique of Inertial Electrostatic Confinement - 1992' ( http://dspace.mit.edu/bitstream/1721.1/ ... 763419.pdf ) has been treated by for example Rostoker, et al in 'Colliding Beam Fusion Reactors - 2003' ( http://forum.nasaspaceflight.com/forums ... ntid=25188 ) - vis::

"In section 2 we show that the very large circulating
power obtained by Rider [1] is a consequence
of the assumption of particle distribution functions
that simplify calculations but have no physical basis."

In particular Rostoker proposes that a 'drifted Maxwell distribution' is more appropriate for calculation of collision distribution functions than Riders 'isotropic' formula which Rostoker asserts does not correctly model the cancellation of factors for 'like-particle' collisions.

There are other refutations of Rider, authors from this forum as well as Nebel himself I believe. Notable others include Chacon - Nov. 2000 Physics of Plasmas paper by Chacon, Barnes, Miley, and Knoll - (who's paper I cant get to - can anyone supply a 'free' link to this?) .

So, whilst to my simple brain, Rider seems to make some very good, though critical, general observations, far better brains have called his maths/assumptions into question - which I find even more difficult to argue with.

I was also reading your own very interesting discourse with Rostoker to related issues in Colliding Beam (Field Reversed Configuration) - ( http://www.sciencemag.org/cgi/content/f ... /5375/307a ). It seems to me that these general issues are the same in Polywell IEC, but the maths is even more complicated and come to a still different conclusion.

To: Munchausen:: yes, yet another factor missing in Rider's analysis - the 'Wiffleball' effect itself.

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

Post by TallDave »

Is Rider's paper discredited? Probably not. Disputed, certainly.

We probably need experimental data at MW energies to really know for sure.

rcain
Posts: 992
Joined: Mon Apr 14, 2008 2:43 pm
Contact:

Post by rcain »

TallDave wrote:Is Rider's paper discredited? Probably not. Disputed, certainly.

We probably need experimental data at MW energies to really know for sure.
I think that was really one of the points of my thread here - are we really likely ever to achieve data at those levels - at least in any 'continuous' burn regime?

or to put it another way, not 'if' but 'how' - to put our fusion engineering hats on.

MSimon
Posts: 14334
Joined: Mon Jul 16, 2007 7:37 pm
Location: Rockford, Illinois
Contact:

Post by MSimon »

i cant help thinking that scattering will be an 'uncontrolled' problem, also as you suggest alpha sputtering.
I am of a different opinion. I believe the real annealing mechanism is beam bunching (klystron like). If that is predominant [POPS results and several simulations including the MIT papers (McGuire/Dietrich) argue that it is or may be enhanced to that point] then scattering will not be an uncontrolled problem.

As to alpha impingement. I think there are limits to what can be worked out in advance. Experiments will need to be done.

ITER is having a lot of trouble with that at much lower energy levels.

Boron 11 coating everything might work in a pB11 machine. Possibly pulse operated if the B11 concentration gets to high. Not good. Not the worst of all possible worlds either.

Neutron flux? I think it is solvable at D-D levels (10 to 100 year lifetimes). pB11 at 1/1,000th of that or less is not going to be a problem.
Engineering is the art of making what you want from what you can get at a profit.

Roger
Posts: 788
Joined: Fri Jul 06, 2007 2:03 am
Location: Metro NY

Post by Roger »

Do I have this right ?

Bremsstrahlung radiation relies on a square topped potential well, a broad wide well, with less than optimum ion focus.


Except that shape well doesnt occur in a Polywell, its sort of parabolic. And Bremsstrahlung radiation occurs when you have hi energy electrons that are also dense. In a Polywell electrons are dense in the center... they make up the "Potential well". And where does the potential well gets its energy from ? The inetic energy of the electrons. So... you have low kinetic energy electrons in the center, so there is nearly no Bremsstrahlung radiation.

Or do I need to go back to school ?
I like the p-B11 resonance peak at 50 KV acceleration. In2 years we'll know.

rcain
Posts: 992
Joined: Mon Apr 14, 2008 2:43 pm
Contact:

Post by rcain »

MSimon wrote:
i cant help thinking that scattering will be an 'uncontrolled' problem, also as you suggest alpha sputtering.
I am of a different opinion. I believe the real annealing mechanism is beam bunching (klystron like). If that is predominant [POPS results and several simulations including the MIT papers (McGuire/Dietrich) argue that it is or may be enhanced to that point] then scattering will not be an uncontrolled problem.

As to alpha impingement. I think there are limits to what can be worked out in advance. Experiments will need to be done.

ITER is having a lot of trouble with that at much lower energy levels.

Boron 11 coating everything might work in a pB11 machine. Possibly pulse operated if the B11 concentration gets to high. Not good. Not the worst of all possible worlds either.

Neutron flux? I think it is solvable at D-D levels (10 to 100 year lifetimes). pB11 at 1/1,000th of that or less is not going to be a problem.
Hi MSimon, interesting answers - thanks.

Bunching, annealing and POPS - sounds a promising approach. The article you cite - is it this one http://ssl.mit.edu/publications/theses/ ... chCarl.pdf ?

(If not, then it appears to be a very interesting and germane paper in any case, though I have yet to read it detail all the way through yet - I particularly like the concepts of 'Multigrid' and 'Asymmetry', so far).

Re alpha sputtering - Member 93143 had in an interesting idea in thread viewtopic.php?t=629. Could it be employed to mitigate sputter do you think?

Re Neutrons: they may be few, but still energetic - enough to disrupt the core or not? I suppose there's little one can do about them in any case, apart from minimise their occurrence/energy in the first place.

To Roger - I'm not an expert, but my understanding so far is that there will be Bremsstrahlung from various regions in the device, wherever collisions, between charged particles (or rapid changes in velocity) occur (bremsstrahlung from P-B11 fusion itself is high I believe, but 'might be' acceptably mitigated within critical regions of Polywell - vis:

'...according to Ligon the bremsstrahlung problem is solved by the fact that the electrons only maxwellianize at the top of the well as they bunch up, which serves to greatly reduce those losses...' - ( http://lists.powerblogs.com/pipermail/d ... 11452.html )

but:

'...On the other hand, the polywell has electrons and ions existing in the same volume, reintroducing the Bremsstrahlung that the fusor can avoid...' - ( http://www.answers.com/topic/polywell?cat=biz-fin ).

Does seem to me that a 'map' of Brem losses (probability density function) could be really useful. Or do I also need to go to school? :)

Post Reply