reddit: We are nuclear fusion researchers, ask us anything

[Joseph Chikva]

You are only juggling with numbers with lack of sense:

Theoretical and experimental studies of kinetic equilibrium and stability of the virtual cathode in an electron injected inertial electrostatic confinement device
Nebel, R. A.; Stange, S.; Park, J.; Taccetti, J. M.; Murali, S. K.; Garcia, C. E.
Physics of Plasmas, Volume 12, Issue 1, pp. 012701-012701-8 (2005).
This paper explores the electron-electron two-stream stability limit of a virtual cathode in spherical geometry. Previous work using a constant density slab model [R. A. Nebel and J. M. Finn, Phys. Plasmas 8, 1505 (2001)] suggested that the electron-electron two-stream would become unstable when the well depth of the virtual cathode was 14% of the applied voltage. However, experimental tests on INS-e have achieved virtual cathode fractional well depths ~60% with no sign of instability. Here, studies with a spherical gridless particle code indicate that fractional well depths greater than 90% can be achieved without two-stream instabilities. Two factors have a major impact on the plasma stability: whether the particles are reflected and the presence of angular momentum. If the particles are reflected then they are guaranteed to be in resonance with the electron plasma frequency at some radius. This can lead to the two stream instabilities if the angular momentum is small. If the angular momentum is large enough it stabilizes the instability much the same way as finite temperature stabilizes the two-stream instability in a slab.

Dude, did you even read what you quoted?

Bussard reports well depths of >80% drive voltage. Nebel states that WB7 didn't show any instability.

Who am I going to call, you, or them? THEM!!!

Yes, understand.
He says damping at the expense of thermalization: big enough angular momentum.
But so big thermalization is impossible at initial phase.
And avoiding one problem he comes to another.
But at least unlike you and other his followers he does not say that "polywell is not beam machine". On your choice: 2-stream instability or big thermalization when angular velocities comparable with radial. Which one is better?

[KitemanSA]

He says damping at the expense of thermalization: big enough angular momentum.
But so big thermalization is impossible at initial phase.
And avoiding one problem he comes to another.
But at least unlike you and other his followers he does not say that "polywell is not beam machine". On your choice: 2-stream instability or big thermalization when angular velocities comparable with radial. Which one is better?

Sorry, where do you get the assumption that the angular velocities are comparable to the radial? "Big enough" might be VERY SMALL and still not be a beam.

[Joseph Chikva]

Sorry, where do you get the assumption that the angular velocities are comparable to the radial? "Big enough" might be VERY SMALL and still not be a beam.

I have read very much about 2-stream instability and Landau damping.
All depends on retio of plasma frequence and electron beam plasma frequency. For non-relativistic electrons - comparable. So, electron temperature should have tens keVs order. For your refernce, desired temperature for ITER 15 keV.
What do you think why electron and light ion driven inertial confinement experiments were canceled? And Heavy Ions Fusion program is still alive.
Because at this moment only the single way of focusing is considered before beams heat hohlraum: namly, propagation of beams trough plasma column. And for electrons and light ions for damping of longitudial instabilities too high temperature of plasma is required. That is very schematic what I said. But it is only my layman's understanding.

[hanelyp]

2. wiffleball, that is as I understand the compact - much smaller than entire reactor zone clot of electrons, which impossible to create without of intersecting of several beams.

Electron beams were used as a means of injecting electrons and creating plasma inside the machine. Injecting gas into the interior and ionizing it with microwaves would also produce a wiffleball, no beams needed. The initial plasma would, however, have a poor energy profile until annealing took place. The high energy tail of ions would leak, leaving a negative charge.

The wiffleball is formed by plasma pressure distorting the magnetic field of a cusp machine, changing the cusps from a funnel shaped inside into a less lossy small hole in a flat wall shape. The outside of the cusps, without high pressure plasma distorting them, retains the funnel shape allowing electrons to recirculate back inside.

If you take a look at the WB-6 or 7, the electron beams are formed by bare heated filament emitters, accelerated by the potential difference between the emitters and the magrid, and focused by the magnetic field.

[Joseph Chikva]

2. wiffleball, that is as I understand the compact - much smaller than entire reactor zone clot of electrons, which impossible to create without of intersecting of several beams.

Electron beams were used as a means of injecting electrons and creating plasma inside the machine. Injecting gas into the interior and ionizing it with microwaves would also produce a wiffleball, no beams needed. The initial plasma would, however, have a poor energy profile until annealing took place. The high energy tail of ions would leak, leaving a negative charge.

The wiffleball is formed by plasma pressure distorting the magnetic field of a cusp machine, changing the cusps from a funnel shaped inside into a less lossy small hole in a flat wall shape. The outside of the cusps, without high pressure plasma distorting them, retains the funnel shape allowing electrons to recirculate back inside.

If you take a look at the WB-6 or 7, the electron beams are formed by bare heated filament emitters, accelerated by the potential difference between the emitters and the magrid, and focused by the magnetic field.

No beams needed?
How the “right” potential distribution is organized? Or you would like to say that inside reactor you have uniform potential dφ/dr=0?

My understanding of Polywell:
Filament (or other type) emitters produce electron clouds, that cloud is accelerated then by the potential difference in the gap and enters into reaction zone radially through the center of short solenoids calling at Polywell’s slang “MaGrids”.
For cubic shape we have 6 such beams intersecting each other in the center having some radial dimension defining by temperature, self-electric field (electrostatic repulse, magnetic attraction and space charge neutralization by positive space charge of being their ions) and scattering.
Some electrons from those beams being inside the certain cone escape reactor while others (scattered at an angle exceeding certain value) are confined by mag field generated by MaGrid.
Am I correct?

If yes, such a system creates the certain space distribution of potential (dφ/dr not equal to 0) with maximum in the center thus accelerating positively charged ions radially.
Does this not the the desired model how Polywell should work?
If yes, I have already shared to you (and to others) what problems may be arisen.

Thanks.

[KitemanSA]

Maybe the best way to satisfy Joe is to suspect that perhaps there is "beam instability" at the very center of the core and the Polywell RELIES on it to get a relatively uniform tangential distribution of radial electron flow (with dominantly radial velocity) which causes a higher density of electrons in the central region and promotes the desired potential well.

Maybe Joe and Nebel are BOTH right, there is "beam instability", but no beam instability "issue".

It is an interesting way of looking at it.

Hmmm.

[krenshala]

Joseph, you keep mentioning six (or at least, multiple) electron beams pumping electrons into the device. My understanding is only one is really needed, but using additional electron sources reduces the time necessary to reach the desired electron density on the inside.

[Joseph Chikva]

Joseph, you keep mentioning six (or at least, multiple) electron beams pumping electrons into the device. My understanding is only one is really needed,...

And what one changes vs. multiple if we talk about electron beam and background plasma? And not one electron beam interacting with another.

[KitemanSA]

From the photos of WB6, I thought there were 8, one at each virtual magnet (corner of the cube).

[hanelyp]

A point to consider is that electrons remain inside the wiffleball a great many times longer than the time they take to cross the diameter of the wiffleball, bouncing and scattering off the magnetic field at each end of their trip across the interior. An spherical symmetry to the plasma is desired, which a small number of beams crossing the device won't create. Even discounting 2 beam instability, beam behavior would quickly degrade with each reflection, becoming a gas of electrons moving in all directions.

Outside the wiffleball, electron density and time is small enough that 2 beam instability shouldn't be a problem.

[Joseph Chikva]

... beam behavior would quickly degrade with each reflection, becoming a gas of electrons moving in all directions...

I speak English too badly for understanding what do you mean here.
But how I understand what you said this is a big nonsense. Because beams are endless and fresh electrons enter into the center continuously instead of deflected and being confined electrons or not deflected and then escaping electrons.
And confined electrons loose their coherent velocities gaining chaotic directions thus converting into background electrons. And 2-stream instability occurs between beam (fresh electrons in which coherent velocity much exceeded chaotic) from one side and background electrons from another.

[KitemanSA]

You have both stated it poorly in different ways.

Electrons coming from the "guns" come in thru the cusps (wiffleball holes) and meet NEAR the center at which point that cumulation of charge forces them all to flow in ONE relavent direction, RADIALLY out. Since the ideal Polywell has a spherical reflecting mirror with itsy bitsy cusp holes, the electrons that are moving radially OUT from the center are (usually, except those that exit the holes) reflected radially back IN... OUT, IN, OUT, IN... no beams.

With each trip to the center, the ANGLE of that path changes, but the RADIAL nature doesn't, much. In theory.

[Joseph Chikva]

You have both stated it poorly in different ways.

Electrons coming from the "guns" come in thru the cusps (wiffleball holes) and meet NEAR the center at which point that cumulation of charge forces them all to flow in ONE relavent direction, RADIALLY out. Since the ideal Polywell has a spherical reflecting mirror with itsy bitsy cusp holes, the electrons that are moving radially OUT from the center are (usually, except those that exit the holes) reflected radially back IN... OUT, IN, OUT, IN... no beams.

With each trip to the center, the ANGLE of that path changes, but the RADIAL nature doesn't, much. In theory.

Excellent. Have beams, but no beams.
Thank you for discussion.

[KitemanSA]

With each trip to the center, the ANGLE of that path changes, but the RADIAL nature doesn't, much. In theory.

Excellent. Have beams, but no beams.
Thank you for discussion.

STARTS WITH beams, doesn't want them, destroys them, gets radial electron flow. Good stuff.

You are welcome for discussion.

[Joseph Chikva]

With each trip to the center, the ANGLE of that path changes, but the RADIAL nature doesn't, much. In theory.

Excellent. Have beams, but no beams.
Thank you for discussion.

STARTS WITH beams, doesn't want them, destroys them, gets radial electron flow. Good stuff.

You are welcome for discussion.

No. Destroyed beam converts to background electrons and getting not radial but chaotic (random) motion interacting then with new portion of electrons entering into reactor zone as a beam.
You would be right only if duration of beam's pulse will be shorter than characteristic time of instability's development (several microseconds). But beams in Polywell as I know have duration of milliseconds order (thousands times longer).