New IEC design from University of Sydney

Point out news stories, on the net or in mainstream media, related to polywell fusion.

Moderators: tonybarry, MSimon

Post Reply
Skipjack
Posts: 6805
Joined: Sun Sep 28, 2008 2:29 pm

New IEC design from University of Sydney

Post by Skipjack »

Please forgive me if this has been covered before and I simply missed this.
John Hedditch from the University of Sydney Plasma Fusion Group has proposed a new design for an IEC device. This device seems much like a polywell with only 2 coils and I am not quite sure I understand why they think that it would work better than a polywell with 6 coils. I would have maybe dismissed this as being rather odd for many reasons (why has nobody thought of this before?) if it was not that the US has lots of experience with IEC devices and Joe Khachan who oversaw this project is also behind the polywell at the US.
Here is the paper:
http://scitation.aip.org/content/aip/jo ... /1.4933213
And here is a video of a presentation describing their device:
https://www.youtube.com/watch?v=G2ijovauYQk

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

Re: New IEC design from University of Sydney

Post by hanelyp »

Just skimming, it looks more like a shielded grid fusor.
The daylight is uncomfortably bright for eyes so long in the dark.

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Re: New IEC design from University of Sydney

Post by D Tibbets »

Some interesting interactions pushed. Problems may be m ostly scaling. An ion rich core contained by physical magnetically shielded cathodes seems reasonable, but...

Bussard stressed that for useful fusion power independant of Q you must have a considerable density of ions. This is one of the keys of the Polywell Wiffleball trapping factor. Even with magnetic shielding there are limits to the density/ charge imbalance of mobile charged particles that can be isolated before pashin arc discharge is unavoidable. This limits the charge imbalance in an exponential manner related to density. They mostly ignor electron losses with the assumption (I think) that there are few mobile electrons. This directly limits the density of ions that can be obtained with any reasonable containment method- Brillion limiting conditions.

IE: Small densities of fuel ions only, and low fusion rates scaling as density squared. The say the system should be scalable to MWs of fusion power, but it is not apparent to me without being near quasi neutral- in Bussard's Polywell there is a difference of ~ 1 ppm, thus lots of electrons that are poorly contained in this low Beta design (I think they stipulated this). They mention pressures of ~ 1 Micron and this might be 100-1000 times less than the Wiffleball trapping factor limited internal density of the high Beta Polywell. As fusion scales as the square of the density, if the comparable polywell produced 100,000,000 Watts of fusion power, this design might generate 100 to 10,000 Watts. If you could maintain input power to less than this Q would be greater than one, but still not much useful power, unless you scaled up the size of the machine to huge proportions. There might be modest power applications that might benefit from this, but the machine weight would probably be huge- got to power those (I assume) very strong magnets.

They also mostly ignor ExB diffusion issues except to say that keeping the surface area of the magnets small helps- sort of like the pre WB6 EMC2 fallacy of assuming the magnet can surface area was essentially zero. In the Polywell this allowed for ignoring electron ExB losses. In this design with a positive charge ion core virtual anode the ExB loss issues will be at least 62 times greater than the comparable electron losses in the negative potential well of the Polywell- where the ions are electrostatically contained without any (ideally) ion ExB diffusion issues. In the Lockheed design the plasma is neutral and the ions undergo their share of ExB diffusion but the dodge is that it is at high Beta so only the ions in the edge region are undergoing this random walk process and this is a small percentage of their lifetime (most of their time is spent in the B field free interior away from the edge .So long as the density driven Coulomb collision frequency (1/MFP) is low enough enough this would limit ExB diffusion on the edge. With low Beta I do not think this can be claimed.

This might make a good neutron source. If superconducting magnet technology can make tremendous advances, this might make a modest power source- such as for a satellite or deep space probe.

Dan Tibbets
To error is human... and I'm very human.

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Re: New IEC design from University of Sydney

Post by D Tibbets »

D Tibbets wrote:Some interesting interactions pushed. Problems may be m ostly scaling. An ion rich core contained by physical magnetically shielded cathodes seems reasonable, but...

Bussard stressed that for useful fusion power independant of Q you must have a considerable density of ions. This is one of the keys of the Polywell Wiffleball trapping factor. Even with magnetic shielding there are limits to the density/ charge imbalance of mobile charged particles that can be isolated before pashin arc discharge is unavoidable. This limits the charge imbalance in an exponential manner related to density. They mostly ignor electron losses with the assumption (I think) that there are few mobile electrons. This directly limits the density of ions that can be obtained with any reasonable containment method- Brillion limiting conditions.

IE: Small densities of fuel ions only, and low fusion rates scaling as density squared. The say the system should be scalable to MWs of fusion power, but it is not apparent to me without being near quasi neutral- in Bussard's Polywell there is a difference of ~ 1 ppm, thus lots of electrons that are poorly contained in this low Beta design (I think they stipulated this). They mention pressures of ~ 1 Micron and this might be 100-1000 times less than the Wiffleball trapping factor limited internal density of the high Beta Polywell. As fusion scales as the square of the density, if the comparable polywell produced 100,000,000 Watts of fusion power, this design might generate 100 to 10,000 Watts. If you could maintain input power to less than this Q would be greater than one, but still not much useful power, unless you scaled up the size of the machine to huge proportions. There might be modest power applications that might benefit from this, but the machine weight would probably be huge- got to power those (I assume) very strong magnets.

They also mostly ignor ExB diffusion issues except to say that keeping the surface area of the magnets small helps- sort of like the pre WB6 EMC2 fallacy of assuming the magnet can surface area was essentially zero. In the Polywell this allowed for ignoring electron ExB losses. In this design with a positive charge ion core virtual anode the ExB loss issues will be at least 62 times greater than the comparable electron losses in the negative potential well of the Polywell- where the ions are electrostatically contained without any (ideally) ion ExB diffusion issues. In the Lockheed design the plasma is neutral and the ions undergo their share of ExB diffusion but the dodge is that it is at high Beta so only the ions in the edge region are undergoing this random walk process and this is a small percentage of their lifetime (most of their time is spent in the B field free interior away from the edge .So long as the density driven Coulomb collision frequency (1/MFP) is low enough enough this would limit ExB diffusion on the edge. With low Beta I do not think this can be claimed.

This might make a good neutron source. If superconducting magnet technology can make tremendous advances, this might make a modest power source- such as for a satellite or deep space probe. Or, as Dr Mills proposed- an IEC rocket thruster with efficiencies and power increased over current ion thrusters. It is both the rocket engine and it's own power source.

Dan Tibbets
To error is human... and I'm very human.

D Tibbets
Posts: 2775
Joined: Thu Jun 26, 2008 6:52 am

Re: New IEC design from University of Sydney

Post by D Tibbets »

PS: I am getting in the habit of using the term of diffusion for ExB losses. This is because it is much like gas diffusion except for the quantification due to the B field strength and associated gyroradius jumping. Transport is also commonly used, and I think it is included in the group of 'classical' magnetic containment issues. In his Google talk, Dr Bussard stressed that magnetic fields are poor at containing ions due to the magnitude of the gyroradius jumps, at least with densities of interest for fusion power generation.

Also note that the cusp loss area is directly related to the charged particle gyro radius and it is about one gyroradius at high Beta. This may contain electrons sufficiently, even without recirculation- my impression is that Dr Parks beleives this. With ions this is not the case. With the Lockheed model both electron and ion recirculation seems to be key. In the Polywell only direct electron magnetic containment, perhaps with electron recirculation assist is nessisary. With a positive charge magrisd, ion recirculation is not possible in the Polywell, but only upscattered ions escape the potential well so this is not limiting and may even be benificial. With a grounded magrid things change some. This machine appears to operate with ion magnetic containment replacing the electrostatic ion containment in the Polywell. And this is at low Beta, so essentially mirror containment, not even low bata cusp containment. Assumption of high recirculation of the escaped ions may provide for adequate containment, except again for ExB loss considerations. Instabilities in the magnetic recirculation (B fields concave towards the plasma) areas outside of the two ring magnets would cause instabilities. The magnitude of these instabilities for the relatively high densities of recirculating ions (relative to the number of ions inside) means that if you can push the internal density, the exterior density goes up proportionately and instabilities grow accordingly. The Polywell and to a lesser extent the Lockheed design limits the relative exterior densities compared to the internal densities due to the high beta (Wiffleball trapping factor) cusp containment, so external density driven instabilities due to concave B fields are constrained. I do not see this with this design with only mirror magnetic confinement .

Dan Tibbets
To error is human... and I'm very human.

Skipjack
Posts: 6805
Joined: Sun Sep 28, 2008 2:29 pm

Re: New IEC design from University of Sydney

Post by Skipjack »

Dan, in the video Hedditch particularly mentions that this design differs from a magnetic mirror.

crowberry
Posts: 672
Joined: Sun Sep 08, 2013 6:34 am

Re: New IEC design from University of Sydney

Post by crowberry »

Skipjack wrote:Please forgive me if this has been covered before and I simply missed this.
John Hedditch from the University of Sydney Plasma Fusion Group has proposed a new design for an IEC device. This device seems much like a polywell with only 2 coils and I am not quite sure I understand why they think that it would work better than a polywell with 6 coils. I would have maybe dismissed this as being rather odd for many reasons (why has nobody thought of this before?) if it was not that the US has lots of experience with IEC devices and Joe Khachan who oversaw this project is also behind the polywell at the US.
Here is the paper:
http://scitation.aip.org/content/aip/jo ... /1.4933213
And here is a video of a presentation describing their device:
https://www.youtube.com/watch?v=G2ijovauYQk
Thanks Skipjack for the the pointer to the video. It is quite interesting and complements nicely their paper. Actually the paper was already pointed to by Mattman in this post (points 11 and 12) viewtopic.php?f=10&t=6082.

Their device is described in their paper title:
Fusion energy in an inertial electrostatic confinement device using a magnetically shielded grid
. So this is not a polywell as this works in a different way.

It will be interesting to see data from the device they are building and to read their next paper on high density behaviour in the device.

Post Reply