Ureka, I think I understand the Polywell and it's parent better.
First an apology to those better versed in the math and physics involved, that have to suffer my ruminations, but...
I had conceived that the Elmore Tuck and Watson( ETW) version of the Fuser, which the Polywell is based on, consisted of a peripheral positive grid that 'pushed' the ions towards the center. Thanks to what I learned in another thread-
viewtopic.php?t=966
I now realize that view was silly. The positively charged wire grid in the ETW and the positively charged magrid in the Polywell are located towards the center of the machine. The periphery/ outside is the vacuum shell . In the ETW machine the electrons are introduced outside of the pos. charged wire grid and are accelerated past the grid towards the center. They would then decelerate, not because of the pos. grid now outside of the electron, but due to mutual repulsion. Once they had passed through the grid on the other side (or reversed direction in the core and came back through the same side) they would decelerate due to the pos. grid behind (and inside) of them till they came to rest at their starting distance from the center and started a new cycle. They would be contained well, except for two problems. Up scattered electrons could possibly reach the vacuum vessel wall/shell, of more likely (?) the electron could hit the wire grid on its' way towards the center(or on the way out). This is the same problem with Hirsch Farnsworth type Fusors, except the central wire grid is negatively charged, with the ions being introduced outside of it.
In the ETW Fusor the ions are introduced inside of the pos. charged grid and are not directly affected by it. But the outside introduced electrons that are attracted towards the center by the pos. grid serve to pull the ions towards the center. Because the electrons are converging towards the center ( and slowing once past the grid due to mutual repulsion) they are concentrated, both due to the increased density due to convergence, and because they are traveling slower in this region ( at least slower than the average speed outside the grid?). This concentration of electrons near the center is the driving force for the ions. The ions, which are 'born' within the pos. charged grid would never reach it as their starting position represents the top of the potential well. Of course upscattering of the ions changes this to some significant degree.
I perceive the Polywell as having two major advantages over the ETW (plus recirculation). The magnetic shielding prevents (or at least greatly decreases) the electrons from being lost by hitting the grid. And , due to the magnetic field 'confining' the electrons within the maggrid once they are introduced, the concentrated volume of the electrons are more localized towards the central regions. This would create a steeper well, which would stop a higher percentage of the up scattered ions from reaching the grid.
The magnetic fields cannot perfectly shield the interior from electron egress due to cusps, but this is not necessarily bad.
In the Polywell, the electrons are injected into the core with electron guns (as opposed to the ETW where they are accelerated solely towards the center by passing the pos. grid). Aligning the guns at these cusps allows the injection of the electrons without having to overcome the strong resistance of the magnetic fields elsewhere. And, because the electron guns are focused on these hopefully tiny cusps, an electron has a near 100% chance of passing through, where they then bounce around after being deflected off their initial path by other electrons and ions(once they are introduced)( if they stayed on their original path they would immediately exit through the opposite cusp) so that they take up a random path. If the surface area of the cusps is only ~1/10,000th of the total surface area of the blocking magnetic bottle, then the electron would be contained for 10,000 cycles.
Those electrons that do escape through a cusp have a chance to be recirculated if they are not traveling to fast. I have, perhaps foolishly, believed that they could continue on the magnetic field line till they could follow it back inside the magrid through another cusp. I accept now, that the electron may bounce back through the same cusp, but couldn't 'orbit' through another cusp because the magnetic field lines that make up the borders of the cusp are looping around such huge distances, that the vacuum vessel walls would have to be very far away to avoid collision.
But, remember the ETW grid (magrid in the Polywell) is positively charged so the electrostatic force would stop the electron at some presumably short distance from the grid, and the electron would then accelerate back into the center again (through the same cusp it exited from?). Say 9/10 electrons are recovered in this way. This added to the containment afforded by the tiny cusp holes ( 1/10,000th of surface area of the magnetic bottle in my example) that are present due to the Wiffle Ball effect, and you get the 100,000 cycle lifetimes of the electrons that are advertised for WB6.
Dan Tibbets
Ureka, my new understanding of the Polywell
Ureka, my new understanding of the Polywell
To error is human... and I'm very human.
I had reached the conclusion that the positively charged grid did not effect the ions within it ( no internal electrical field within a hollow sphere). Only the magnetically confined electrons provided an accelerating force on the ions. But, does the cloud of electrons near the center of the space within the pos. magrid shield the ions from the opposite side of the 'spherical grid'? This would allow the net positive charge on the near side grid to repell the ions on that side of the core- thus decelerating them (force proportional to the size of the 'electron shadow' relative to the diameter of the magrid), especially those upscattered ions that would otherwise hit the grid, or follow the electrons through the cusps. Would this counter Art Carlson's contention that the ions would follow the electrons through the cusps? Does this still apply if the core is almost neutral (mix of ions and electrons)? Does the ~ 1/1,000,000 excess of electrons play a role?
I'm guessing that this effect must be real, otherwise how could a positively charged grid near the vacuum vessel wall (essentially a hollow sphere itself)decellerate fusion alpha particles in order to capture their energy.
Dan Tibbets
I'm guessing that this effect must be real, otherwise how could a positively charged grid near the vacuum vessel wall (essentially a hollow sphere itself)decellerate fusion alpha particles in order to capture their energy.
Dan Tibbets
To error is human... and I'm very human.
You could say, if you wanted, that the electrons shield the ions from the positive charge on the far side. But then you'd have to subtract that effect from the accelerating effect of the electrons. The upshot is the same; the correct way to look at it (to first order) is that the ions see the ball of electrons but not the grid.
As far as direct conversion is concerned, the important part is not the charge on the outer end of the direct conversion system (the collectors); it's the charge on the inner end (what I call the 'trap grid'). This is a grid in between the magrid and the collectors, at a potential at least as low as that of the electron guns (ie: magrid minus drive). It allows electrons to recirculate because the magrid is still positively charged, and the potential difference between the magrid and the trap grid is high enough that electrons mostly can't get out that far, but since the trap grid has a massive negative charge, alphas decelerate once they've cleared it. The potential difference between the trap grid and the wall is just enough (ideally) to stop the alphas.
As far as direct conversion is concerned, the important part is not the charge on the outer end of the direct conversion system (the collectors); it's the charge on the inner end (what I call the 'trap grid'). This is a grid in between the magrid and the collectors, at a potential at least as low as that of the electron guns (ie: magrid minus drive). It allows electrons to recirculate because the magrid is still positively charged, and the potential difference between the magrid and the trap grid is high enough that electrons mostly can't get out that far, but since the trap grid has a massive negative charge, alphas decelerate once they've cleared it. The potential difference between the trap grid and the wall is just enough (ideally) to stop the alphas.
OK, that makes sence, except- the electrons have a negative charge, and attract the pos. charged ions towards the center. The near side pos. charged magrid is positively charged and would repell the ions towards the center. Would not the effect be addative?93143 wrote:You could say, if you wanted, that the electrons shield the ions from the positive charge on the far side. But then you'd have to subtract that effect from the accelerating effect of the electrons. The upshot is the same; the correct way to look at it (to first order) is that the ions see the ball of electrons but not the grid.
Dan Tibbets
To error is human... and I'm very human.
You've misunderstood, which is what I was afraid of. Never mind what I said.
Electrons do NOT block the effect of the far grid. It's that simple.
Are you familiar with the Principle of Superposition? You take the net effect of the magrid (zero, at least for an ideal spherical shell), the effect of the electrons, and the effect of the ions, and add them. That gives you the effect the whole configuration would have on a given test charge in a given location.
Adding together the electron and ion effects is as simple as finding the differential between the ion and electron charge densities and using that as a net space charge density, at any given point in the reactor volume. Ionic charge in a given location cancels out an equal amount of electronic charge, so what you're left with is the difference.
Electrons do NOT block the effect of the far grid. It's that simple.
Are you familiar with the Principle of Superposition? You take the net effect of the magrid (zero, at least for an ideal spherical shell), the effect of the electrons, and the effect of the ions, and add them. That gives you the effect the whole configuration would have on a given test charge in a given location.
Adding together the electron and ion effects is as simple as finding the differential between the ion and electron charge densities and using that as a net space charge density, at any given point in the reactor volume. Ionic charge in a given location cancels out an equal amount of electronic charge, so what you're left with is the difference.