Page 2 of 3

Re: Positive Electrostatic Fusion Reactor

Posted: Tue Sep 01, 2015 7:17 am
by Rich Feldman
To begin with, I'm concerned about the scaling of arrows which render the computed E-field strength and direction.
In some cases, they falsely suggest a substantial radial field extending inside the inner grid!
The arrow sizes & directions are based on E-field at their tails, which are located in a regular Cartesian array.

I agree that this is a three-body problem, not a two-body problem. The outer shell has some potential, whether or not it's an electrical insulator. Its voltage with respect to the inner and outer grids matters a lot. Garrett's pictures show that shell voltage can make the difference between a roughly equipotential surface at outer grid radius, and a potential field that dips to around mid-voltage between the outer grid wires.

Garrett: It looks like your FEA model (mesh) outer boundary corresponds to the fusor vacuum chamber wall.
For the insulating chamber case, what's the boundary condition? How come the whole circumference seems to end up at 0 volts?
If a real, finite insulating shell had no static charge, the radial E-field from a positive outer grid would continue smoothly to infinity. ( See "capacitance of sphere" in an E&M textbook. ) FEA result inside the shell would be accurate if shell potential matched the potential at that radius in the real, infinite world. Can we see what happens if ratio of FEA boundary radius to outer grid radius is 10:1 instead of barely 2:1?

Of course a real insulating shell (or insulated metal shell) could have static charge of arbitrary quantity and polarity. Mightn't they come to equilibria at voltages not much different than that of outer grid? In real fusor lab, a floating metal shell should have its voltage measured; otherwise that voltage would be an important uncontrolled variable.

Oops, almost forgot to ask.
Are the simulations 2 dimensional, as if the grid wires and shell were parallel to each other and perpendicular to the drawing plane?
Are they quasi-3d, where the left edge of drawing is an axis of rotation? Then shell is a sphere. Each grid is made of 1 large circle, 2 smaller circles, and 2 small balls at the poles. That could explain difference between contour shapes at North and East grid regions.
Or are the simulations really 3d, where the drawing omits critical details of the grid geometry?

Re: Positive Electrostatic Fusion Reactor

Posted: Tue Sep 01, 2015 9:27 am
by Frank Sanns
Ignore the inner grid for a moment to get a better idea of what is going on with the outer grid, the insulator and the chamber wall.

Consider what the outer grid and chamber condition are. Without the outer shell wall, the E field will put charge onto the insulator and continue to fall of past the insulator. The charge will be contained throughout the insulator and fall off beyond that.

When an outer chamber conductor is present, the E field aligns the dielectric molecules of the insulator so the apparent so the voltage on the inside of the insulator climbs to the potential of the outer grid. Simultaneously, the outer surface climbs to the potential of the chamber wall. This is exactly a capacitor. Consider the configuration of a Leyden jar to understand what the outer grid and chamber wall must look like.

In this condition, the voltage on the inside of the insulator is at outer grid potential. It is potential that moves charges. Charge cannot move without a voltage gradient.

Re: Positive Electrostatic Fusion Reactor

Posted: Tue Sep 01, 2015 2:15 pm
by Garrett Young

Here is an image of the same simulation with the pyrex bowl highlighted (green). The boundary is the outer vessel and is grounded (the floating condition was done in a previous simulation)
This is a quasi-3D simulation so you are correct about the structure of the components. You are also correct about the arrows in that the length is the magnitude and the origin of the tail is the location.

In general, as amateurs experimentation is important, but I think we would all benefit from more theoretical testing to guide and focus our efforts.

Re: Positive Electrostatic Fusion Reactor

Posted: Tue Sep 01, 2015 3:10 pm
by Frank Sanns

I totally agree that we need more modeling but there are some problems that I see.

The first is access and competent knowledge of the software. We old timers see software come and go and never really know what to use and how to get good simulations. Garbage in garbage out is always my fear and is well justified from good experiments and bad computer modeling back in the early 1980's. I would be first in line to have software and know how to do it as there are all kinds of configurations I would like to test out. I would also like to see a good explanation of how the star rays make it beyond the outer grid and even beyond the outer shell and into large side ports like I have in my fusor.

The second concern is items where insulators are put in paths. While I totally agree with your images, including the last one, I am still not convinced that the last one is totally accurate for what an ion or an electron will see since it is confined and the surface of the insulator will surely develop a charge due to collection in addition to the electric field. Not sure how to account for that.

Lastly, how does this apply to the original poster's comments about confinement within a cylinder and the fusion that might or might not occur.

Re: Positive Electrostatic Fusion Reactor

Posted: Fri Sep 18, 2015 5:13 pm
by Royce Jones
Sorry for the delay. An alternative to getting a positive charge around the chamber would be to use permanent magnets. If the goal is to control strays this should work.

Re: Positive Electrostatic Fusion Reactor

Posted: Fri Sep 18, 2015 5:32 pm
by Royce Jones
The Plates: I seems that the plate would work, however the charge would be on both side of the plates so the electrical input would have to be doubled to account for this.

Re: Positive Electrostatic Fusion Reactor

Posted: Sun Oct 18, 2015 6:22 pm
by Royce Jones
A few points. The proposed design has a greater concentration of focused ions.
These ions should move in only two directions resulting in more fusion reactions.
Charged ions are "attractive", so little confinement should be necessary.
An "floating positive grid" as discussed by would be an interesting addition to the concept. Since it only needs to control strays the positive grid should not need much charge. Alternatively, small permanent magnets might also do the job.

Re: Positive Electrostatic Fusion Reactor

Posted: Sun Oct 18, 2015 8:43 pm
by Dan Tibbets
I recomend that you have a good understanding of Gauss law. Basically, in a conductive sphere or wire mesh with a given radius will not effect charged particles inside of it (at a lesser radius). Only charged particles outside of the sphere are effected. That is why discussion often mention accelerating charged particles past a grid towards the center, once inside the grid only thei inertia and local coulomb collisions between particles have effect. The grids do not play a role again untill the particle passes outside of it or hits it. The chamber walls by definition are always outside of the charged particles, till they hit it, so it does not have any contribution to the properties of the charged particles inside of it, just like a grid at a greater radius than the particle. The grid does not need to be spherical or continuous (solid surface for Gauss law to be dominate. Eventually, things start to break down once the openings in the grid grow large enough, but for the most part this is only a miner consideration in fusors, and ideally in Polywells, though there are qualifiers and gotches.

For what I think is a good introduction to Gauss Law, watch this video by an MIT professor:

Concerning a beam of positive ions (like charged mobile charged particles) when one dimension is considered the charges can only be located ahead or behind each other and I could see this as having a cancelling effect. But with two or more dimensions, the beam will also have repulsion away from the axis of the beam- the beam will spread/ defocus and will do so exponentially with increased density. I think this is what you are saying when you mention the 14,000 volts at the surface of a 1 cm diameter beam. This is a lateral acceleration voltage or force and leads to beam spreading. It is not like a beam of light where there is no mutual photon repulsion. The charged particle beam will only maintain collimation with magnetic and/ or electrostatic continous external application of focusing forces. With the Polywell and other quasi spherical designs there are some consequences that may be interesting, but I do not see how they would apply when considering two parallel plates. Neutral beams, or near neutral beam where space charge spreading is under some minimum is not a free ride either. Because ions and electrons have different masses, they have different speeds at the same energy. The electrons leave the ions behind and space charge builds up leading to beam spreading. I think this is the basis of the two stream instability.

An equal mixture of positive and negative ions would be interesting, but
I don't know how complicated things could get, because the orbital electrons need not be well behaved and stick to single ions. There could be charge exchange, and a certain percentage of free electrons always being decoupled from the ions at any given time, especially when you are at energies well above the ionization potential of the ions (like thousands of eV)...

Dan Tibbets

Re: Positive Electrostatic Fusion Reactor

Posted: Sun Oct 18, 2015 8:56 pm
by Dan Tibbets
PS: Using a positive charge on a grid will accelerate opposite charges past it (or hit it), so long as Gauss Law considerations are accounted for. This is the idea for the Elmore Tuck and Watson variation on the fusors. Electrons are introduced outside of the positive grid and accelerated towards the center of the spherical space. With a spherical space, as the center is approached the volume decreases so that there is a concentrating effect of the electrons. This creates a virtual cathode near the center. If ions are introduced inside the positive grid (so they are not immediately accelerated towards the walls and lost) they will be attracted to this virtual cathode located at a smaller radius. They will eccelerate to fusion speeds, and fuse in the center or pass outward till they are stoped by the virtual cathode again and reaccelerate towards the center. Ideally, the ions would ossilate back and forth in this manner till they fused. There is no cathode wire for them to hit. But, the electrons can hit the peripheral positive anode wire because they will not be slowed and reversed till they are beyond it (at a greater radius. Nothing is gained over having the negative wire grid. If the grid, positive or negatively charge is magnetically shielded, though, this transparancy problem is avoided. At least that is the idea. It doesn't work very well because of all of the problems with magnetic shielding/ containment. The Polywell and other approaches have tried to get around this through clever manipulations. The problem though is that for every cleaver manipulation several confounding complications crop up.
These complications are often harmful, but some may actually benefit other aspects so the quest goes on...

Dan Tibbets

Re: Positive Electrostatic Fusion Reactor

Posted: Tue Oct 20, 2015 10:41 pm
by Jason C Wells
Royce Jones wrote:Since the wall has a high positive charge the positively charged ions (protons) will be repelled away from the wall back into the plasma or ion cloud.
This is not true. Any configuration of charge will re-arrange itself to achieve the state of lowest energy, the state with zero field. When you put a single ion inside your vessel, then you won't have zero field. That ion will find a place to stick itself to the inside of your vessel. That is called sputtering.

Your device isn't purely electrostatic though. It has magnets. In fact, your device looks like a pair of Penning traps. Your end grids would need to be within the magnets completely. The curved part of the mag field will happily carry ions directly to the wall if they can get to the gap between the endplates and the magnet. If that happens, your trap will fail to operate. (btdt)

If my assumptions about a Penning trap are correct, your device is not a linear accelerator. A Penning trap is like a little mini cyclotron.

The Penning trap part of your devices will make ions all by itself. You don't need to inject ions at the ends. Wikipedia says Los Alamos has done fusion with Penning traps. ... onfinement

I think you mean that your center grid is -50kV in the picture.

I use a modified Penning trap to create protons. Or... at least I think that's what I'm doing.

Jason C. Wells