Fusor Computer Modeling - Call for Contributions

This area is for ADVANCED theory and discussions only. If you just joined fusor.net, chances are this is NOT the area for you to be posting.
Post Reply
Frank Sanns
Site Admin
Posts: 2117
Joined: Fri Jun 14, 2002 2:26 pm
Real name: Frank Sanns

Fusor Computer Modeling - Call for Contributions

Post by Frank Sanns »

The subject of computer modeling comes up from time to time but we have never had a group thread of contributions to treat the subject rigorously. Simple field lines does not do it.

Some questions that would be constructive to answer and illustrative for the newcomers:

1. What is the optimum ratio of inner and out grids.
2. Does having an outer grid and leaving a relatively large distance to the chamber wall gain recirculation time or energy?
3. How do ions seemingly have more energy than they started with when they pass right by the outer grid after passing through the inner grid?
4. If potential is relative, why does it matter the inner grid is negative or the outer grid is positive (large space from the chamber wall or insulated)?
5 Why is there "star" mode?
6. Why is there an upper limit of neutrons at a given current despite the voltage?
7. Why are point electrodes or plasma electrodes as effective or more than grids?
8. Is it really recirculation fusion in the bulk fusor volume, in the inner grid, or beam on target fusion of the grid or the walls?

There are more but I just wanted to get some starting thoughts down. I have seen some papers but I have not seen any that could answer several of these questions.

I think this is a good thread started for this advanced area. I hope you computer people step up to this one and give the group a jump start on these issues. Many of us with experience can answer some of them but to take fusion to the next step we really need a comprehensive understanding of all of the above and then some.

Let the discussions and contributions begin.
Achiever's madness; when enough is still not enough. ---FS
We have to stop looking at the world through our physical eyes. The universe is NOT what we see. It is the quantum world that is real. The rest is just an electron illusion. ---FS
benbartlett
Posts: 145
Joined: Mon Jun 01, 2009 7:34 pm
Real name:
Contact:

Re: Fusor Computer Modeling - Call for Contributions

Post by benbartlett »

This is something I have been interested in modeling for quite some time. Solving Poisson's equations to calculating realistic solutions for potential over space given chamber geometry is simple enough, but the true challenge will lie in simulating the dynamical conditions in the reactor with some degree of accuracy for the number of particles involved. I suspect some clever use of smoothed-particle hydrodynamics would be able to handle the majority of the non-fusing dynamics of the problem in some feasible computing timescale, but the two main problems with this would be that it is completely inapplicable at calculating high-energy individual particle interactions, such as scattering and actually fusing. Making a hybrid algorithm that can identify when high-energy reactions are likely to take place and simulate them separately but also handle the bulk of the dynamics seems pretty infeasible for a small group, as it would be so widely applicable that someone probably would have developed some solutions to this.

Unfortunately, there hasn't really been enough rigorous research on some of the dynamics of how fusors actually work for me to be convinced that this simulation is feasible. For example, the "fusion happens in free space" vs "fusion happens in the top layers of the permeable material in the accelerator grid" debate (I think this is your question #8), which has very little supporting evidence from either side as of the last time I checked. (I have been discussing an experimental design with Scott Moroch that might give solid evidence in favor of one side or the other, but this is just a concept at the moment and since my fusor is no longer in working order will likely need to be executed by him.) If it turns out that the latter is the answer to that debate, the problem likely gets even harder. Even if we could identify when high energy interactions were likely to take place, simulating the actual interactions between the particles and the grid that are necessary to even give a semi-accurate approximation is also likely to be pretty infeasible without liberal use of generalizing statistics. From my own experience, it takes several minutes running on CERN's lxplus clusters to simulate a single very simple Higgs-diphoton interaction in GEANT4 - obtaining enough accurate reaction simulations to be able to generalize without resorting to small-number statistics would take a very long time and would need to be iterated over many different conditions. All of this is ignoring the complete infeasibility of programming something like this in a small group.

That said, I still think there could be a lot of work that could feasibly be done in this area, though much of the simulation will need to be generalized and some aspects of it, particularly calculations involving fusion rates when the mechanism for fusion in fusors is largely unknown, should not be taken too seriously. Perhaps if I manage to stumble across another giant roadblock (read: horribly broken clustering algorithm used in the CMS) to my current research I'll have some time to devote to this.

Finally, here are some thoughts on your questions. Consider them not as answers but simply as educated guesses made after only a few minutes of thought:
1. This depends on the mechanism for fusion. I have no answer if fusion significantly involves material-deuteron interactions, but if it happens in free space, I suspect a needle-like design would be the best design (i.e. ratio is zero).
3. Probably due to scattering effects. If you collide a group of ions from equidistant starting locations, then ignoring energy lost to Bremsstrahlung/other forms of light emission, we'd expect about half of the particles to end up with a higher final energy after the first round of collisions.
4. Because the charge of the deuterons is not relative...? I'm not sure I understand this question.
5. Likely due to increased RMS distance allowing for more energetic collisions, so scattering effects are more noticeable, though this is a very naiive guess and the dynamics of this are likely much more complex.
6. I don't understand this question.
7. Similar to my answer to #1, particularly if the majority of fusion happens in free space (which I suspect it does but have little evidence to support). Needle-like electrodes will cause much much greater accelerations over a equally shortened distance; this would probably eliminate much of the energy loss from low-energy interactions and confine the particles to a steeper/narrower potential well. See the attached (very simplistic 2D) demonstration of this using some simplified geometry and potentials from my reactor.
8. I think this is the key question to give more insight into the previous questions.

~Ben
Attachments
Simple potential models
Simple potential models
User avatar
Richard Hull
Moderator
Posts: 14975
Joined: Fri Jun 15, 2001 9:44 am
Real name: Richard Hull

Re: Fusor Computer Modeling - Call for Contributions

Post by Richard Hull »

I have written to this point related to what is the main source of fusion in the simple amateur fusor as typically built here many times in past postings. There are many possibilities and I feel ALL the possibilities do, indeed, produce fusion. Which is the main one? It is anyone's guess and no one has proved one is any better or worse than the other in the mix. A good mind will have to admit to both Deuterium-target fusion, fusion in velocity space, and some limited IEC fusion as originally so naively envisioned, as well as wall based storage of deuterium and fast neutral fusion are all involved. Other processes may yet be contributing, but not put forward by conjecture at this point.

It is probably an equally naive effort to think an amateur fusor, as presented here, can be definitively modeled. Likewise, definitive, rock solid, empirical measurements to determine details related to the commanding processes in the fusion numbers is most likely beyond the casual or even a committed amateur.

I would imagine there are as many as 5 or 6 processes doing fusion in the fusor. What percentage of each are in play would, indeed, be nice to know. As this device is simple by its very design, there is undoubtedly a chaotic maelstrom of interactions within the device....The good the bad and the ugly.

Still, the discussion within this particular thread is needed. Modeling of the simple fusor is a lofty and idealized goal.

The big boys have modeled their designs to death for years and years and at long last have a good handle on how not to do fusion to advantage.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
The more complex the idea put forward by the poor amateur, the more likely it will never see embodiment
User avatar
Nicolas Krause
Posts: 230
Joined: Fri Sep 30, 2016 7:36 pm
Real name: Nicolas Krause
Location: Canada
Contact:

Re: Fusor Computer Modeling - Call for Contributions

Post by Nicolas Krause »

In my limited spare time at school I've been working on modelling a fusor numerically. Since I'm unable to work on the construction of the physical device while at school (all of my components are in my hometown), I've been trying to apply what I'm learning in my numerical methods class. At this point I have a pretty graph of the voltage potential of a fusor in two dimensions calculated in matlab. The figure is below, and I'm happy to provide the matlab code should anyone request it. It's cobbled together from a number of online examples for calculating various electromagnetic values. After some discussions with a number of people however, I've made the decision to move away from matlab and use Python instead, as a result I'm currently rewriting my code. I plan on attempting a particle in cell simulation where I'll be able to observe the orbits of the ions and hopefully gain some insight into fusor operations.
Voltagepotential.jpg
Daniel Firth
Posts: 36
Joined: Wed Apr 04, 2012 10:31 pm
Real name:

Re: Fusor Computer Modeling - Call for Contributions

Post by Daniel Firth »

There's a Turkish professor, Erol Kurt, who performs research into IEC devices. Here's a paper about ion trajectories:

A stationary multi‐component cathode modeling and ion trajectories for an inertial electrostatic confinement fusion device

Some of his work uses the finite element method (FEM) to model the fields:

An inertial electrostatic confinement (IEC) device modeling and the effects of different cathode structures to the fields
DR_DANIEL_GOODMAN
Posts: 7
Joined: Wed Dec 27, 2017 11:43 pm
Real name: Daniel Goodman

Re: Fusor Computer Modeling - Call for Contributions

Post by DR_DANIEL_GOODMAN »

To all here:

https://arxiv.org/abs/1510.01788

I stumbled across this on Arxiv, a quite useful site for free article preprints, this one is on inertial electrostatic confinement, the ODD is free and downloadable, hope is of interest, if I find more, I'll try to send it also, of course.

Many thanks, enjoy,

Dr. Daniel Goodman.
DR_DANIEL_GOODMAN
Posts: 7
Joined: Wed Dec 27, 2017 11:43 pm
Real name: Daniel Goodman

Re: Fusor Computer Modeling - Call for Contributions

Post by DR_DANIEL_GOODMAN »

To all here:

Here's a Google Scholar search I just did, there's also some stuff under IEC on ResearchGate, if you spell out the acronym, as well, though I did find a good deal here, certainly. Here's the link:

https://scholar.google.com/scholar?hl=e ... =fusor&oq=

Enjoy,

Dr. Daniel Goodman
Post Reply

Return to “Advanced Technical Discussion Area”