Modeling Polywell

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

Moderators: tonybarry, MSimon

JohnFul
Posts: 84
Joined: Sat Feb 27, 2010 7:18 pm
Location: Augusta, Georgia USA

Modeling Polywell

Post by JohnFul »

In light of the lack of the ultimate model, the experiment, I am keenly interested in computational modeling of Polywell and the wiffleball effect. One start is the code presented in the link by D tibbets, http://www.mare.ee/indrek/ephi/images.pdf.

Interesting as it is, this is a model of conceptual perfection. I'm looking for a real world model that matches data as we know it. Preferably this can be broken down into components that can run on members of an HPC grid. Why you may ask? Because I just happen to personally have 24 cores with a total of 128GB of RAM and 14 TB of storage that I can apply to the problem. I'm running a virualized environment and can do whatever OS is convenient. My leaning is the Microsoft OS' but I am reasonably proficient in the Unix variants. In particular, I mainly work with BSD.

J

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

Post by KitemanSA »

The recent 2009 IEC conference had an interesting paper regarding a PIC simulation of a polywell. That might be a place to start.

http://fti.neep.wisc.edu/iec2009/agenda

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

Post by TallDave »

I corresponded with Joel Rogers the other day, he has an email list too if you're interested. Seems pretty open to adding people to it,
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

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

Re: Modeling Polywell

Post by Art Carlson »

JohnFul wrote:I'm looking for a real world model that matches data as we know it.
Data?! You mean there is actual data somewhere?

Indrek
Posts: 113
Joined: Wed Jul 04, 2007 1:51 pm
Location: Estonia
Contact:

Post by Indrek »

Lol

ckrucks
Posts: 21
Joined: Mon Jun 23, 2008 4:50 pm
Location: Minneapolis, MN

Post by ckrucks »

Indrek! Good to know you still lurk and post.
________
DURASPARK
Last edited by ckrucks on Wed Feb 16, 2011 12:15 am, edited 1 time in total.

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

Post by rcain »

hail Indrek :)
(any simulated data you would like to disclose?)

kcdodd
Posts: 722
Joined: Tue Jun 03, 2008 3:36 am
Location: Austin, TX

Post by kcdodd »

I have been working on a pic code with cylindrical symmetry. I am currently in the process of adding dynamic EM solver, as well as diffusion. My original objective was to model a mirror machine. But I think it will be possible to get some good ideas for a polywell from a simple spindle cusp model, as the same basic principles are at work (point cusps, line cusps, and convex magnetic curvature). My time to work on it is limited but I will post updates as it progresses.
Carter

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

Post by MSimon »

My understanding is that in a mirror machine you have to make them unfeasibly long to get net power.

No problem for a simulation except for run time.
Engineering is the art of making what you want from what you can get at a profit.

kcdodd
Posts: 722
Joined: Tue Jun 03, 2008 3:36 am
Location: Austin, TX

Post by kcdodd »

Yes, it was more like the Maryland supersonically rotating machine. Well the trick is to get it to run on a PC isn't it, lol.
Carter

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

Post by icarus »

Well the trick is to get it to run on a PC isn't it, lol.
Heard that some people have had good success on a PC with CUDA, takes advantage of the GPU on your video card (also paralleled GPUs) ... model problems only so far, so PIC may not be suitable, might be an interesting sideline though.

http://en.wikipedia.org/wiki/CUDA

PS: Didn't happen to come across that super-conducting bag code that you wrote? Might take another look at adding in the electron sheath now I have time again ...

hanelyp
Posts: 2261
Joined: Fri Oct 26, 2007 8:50 pm

Post by hanelyp »

I'm thinking a full up simulation of the polywell isn't practical. But we might still get useful results from simulations of specific effects. Such as electric well profile assuming an idealized spherical confinement of electrons. Or the particle energy profile given the electric well. Both of these simulations could make very good use of symmetry. Isolated cusp behavior would also be of interest.

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

Post by TallDave »

FWIW, Joel told me he hasn't been able to simulate a WB yet. But Rick has seen it in action, so we know it happens.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

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

Post by icarus »

But Rick has seen it in action, so we know it happens.
Is that the official line or more of your privileged-insider third person hear-say? What did he actually say, and what did he actually measure to verify this? Thnx in advance for any info.

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

Post by TallDave »

icarus wrote:
But Rick has seen it in action, so we know it happens.
Is that the official line or more of your privileged-insider third person hear-say? What did he actually say, and what did he actually measure to verify this? Thnx in advance for any info.
He posted somewhere here words more or less to the effect of "confinement is 1000x times better in WB mode so this is easy to see." I believe this means they see the current drop off dramatically as they approach beta=1.

I should hasten to add we don't really know the WB forms, in the scientific proof sense, because the data hasn't been released and the WB effect hasn't been replicated in labs across the world, but I'm trusting Rick's statement for purposes of speculating on the Internet about the position of this technology in the context of suggesting that Polywell modelling is difficult.
n*kBolt*Te = B**2/(2*mu0) and B^.25 loss scaling? Or not so much? Hopefully we'll know soon...

Post Reply