A letter to Newbie about the Polywell

Discuss fusion-related developments, personalities, and events. Explore how we got to where we are today.

Moderators: tonybarry, MSimon

Post Reply
Posts: 459
Joined: Tue May 27, 2008 11:14 pm

A letter to Newbie about the Polywell

Post by mattman »

Hello All,
I wrote a friend a letter about the Polywell. They know nothing about the machine. I thought parts of this were good enough to put up on the forum. Please let me know what you think of my description of the state of things....



I have been studying an idea known as the Polywell. It is a new idea for fusion power. It is currently being researched by: the United States Navy, the University of Sydney, the Atomic Energy Organization of Iran, and a community of amateurs. The idea is the merger of two previous technologies: the fusor and magnetic mirrors.

Fusors have been around since the early 1960's. They are very easy devices to build. Many high school students have built a fusor in their homes. A fusor uses two wire cages, one inside the other, inside a vacuum chamber. A voltage is applied between the two cages. Charged ions are injected into this. They see this voltage like a hill and fall towards the center. when they reach the center, they have gained kinetic energy. They can hit and they may fuse. The minimum voltage needed for fusion is 10,000 volts. You can fuse isotopes of hydrogen at that voltage. Magnetic mirrors have also been around since the 1960s. When an electron is corkscrewing into a dense magnetic field, it will hit a "mirror" and turn around. Magnetic fields are used inside tokomaks to contain plasmas. In the 1970's and 1980's there was an extensive mirror program funded at the US national labs. The biggest machine they built was the MFTF. This machine was shut down by the Reagan administration the day that it opened, in the mid 80's. The excuse given was that they needed to balance the budget.

The problem with the fusor is, though they can do fusion, they cannot make net power because the loss rate are so high. There are three ways the machine loses power. The first is conduction from the wire cage. Charged particles touch the inner and outer wires and are conducted away from the center. The second method is through energy capture, like all power plants, the device is limited by energy capture efficiency. The last method is through light radiation. Light leaves the cloud of electrons and ions in the center through several means. It comes off as visible, IR and X-rays. The X-ray radiation being the greatest. These ideas come from John Lawson's 1950's paper on fusion power. In the middle 1990's a single PhD student at MIT did his doctoral thesis on all fusion devices not at thermodynamic equilibrium and stated that they were fundamentally limited. The major limitation being that the X-ray losses were so great that they would be higher than the energy produced by fusion. It was highly theoretical work and I have been looking at it. Very little research has been done in this direction since.

The Polywell combines the fusor and the magnetic mirror machines together. In doing so, it reduces the conduction losses. My big idea is to reflect the X-rays. By reduces these losses the machine would move closer to net power at low temperatures. At present, there is no clear way to do this. Surfaces which reflect X-rays are available, but the plasma will need to be dense enough to re-absorb the reflected light. It is unclear the best way to do this - or even if this is the right path to take.

The polywell uses six rings in a cube. This ring structure is put inside a cage. The rings are about a foot a side. The cage is about three feet a side. This is all put inside a vacuum chamber. The rings are electromagnets, in a box. They are switched on. They act like 6 magnets, where all the poles point inward. Outside the rings, an electric voltage is applied. The positive voltage points from the cage to the rings. An electron emitter is roughly halfway between the rings and the cage. Electrons are emitted. They fall down the voltage to the rings in the center. At some point the magnetic Lorentz force overtakes the electric force - the electrons are "caught" by the rings. They take a corkscrewing path towards the rings. They enter the ring structure, losing energy as they move. They are then, internally reflected. The magnetic fields trapped them center. They fly around this center, like marble rolling around in a bowl. This supposedly traps 2E12 net electrons in the center. This cloud supposedly looks a 14 point "star". Eventually, these electrons will be lost. Finding ways to extend electron confinement will also cut into energy losses.

If you trap that many electrons in the center, they collectively act like a point charge. Now you puff hydrogen gas towards the rings. The gas is uncharged, so it can pass through the voltage drop with no problems. When the gas reaches the edge of the electron cloud it exchanges energy with "hot" electrons. If the electron is hotter than 16 eV, it ionizes. The gas breaks apart into 2 electrons and 2 ions. Classically, the ion is ~18,000x times bigger than the electron. The two ions are positive, they fall towards the big point charge in the center. They build up speed, or kinetic energy, as they fall. When they reach the center they can have as much as 10,000 eV. If they hit in the center, they may fuse. This fusion - generates a high energy particles and neutrons. The particle cannot be contained by any of the fields. The products rapidly exit the machine. The stated goal of NIF is to get plasma at an average energy of 10,000 eV, under dense confinement.

Obviously, there are many competing physical mechanisms. Part of making this work, means optimizing it. How to get the desired fusion mechanism, while minimizing everything else. These include but are not limited to: x-ray radiation, plasma instabilities, neutron embrittlement, columbic repulsion without fusion, electron loss, arching, all conduction losses and all other radiation losses. Also a practical method for energy capture must be found. Two methods have been purposed: the first is just energy capture by heating a fluid the other is direct conversion. Direct conversion was an idea from the national labs in the late 70's and early 80's. It was expressed theoretically by a physcist, William Bar. He got funding for it and built an experimental system in the 80's to test the idea.

These machines are very complex. There are many variables to consider: the electron to ion ratio, the reactor volume, the strength of the magnetic fields, the voltage drop, the fusion fuel used, the placement of the emitters, the vacuum pressure, the amount of space around the rings, ect... This problem needs to be simplified, and a myriad of designs need to be tested in simulation before an experimental reactor is built. With so many competing variables, there are many ways the machine can be operated. There are probably modes of operation, just as there are in the fusor. There may even be an operational sweet spot. It remains to be seen is there is a way to build it, so that net power is possible.

Aside from those already in the field, it seems the traditional channels are closed at the moment. The department of energy is not interested in funding technologies that compete with their flagship fusion machines: ITER, NIF, Z-Machine. Since the 1970's they have sunk billions into ICF and the tokamak and there is a whole established network of career scientists and bureaucrats working on these projects. They are not interested in funding radical new ideas. Moreover, why would they? The MIT work provides a good - albeit wholly theoretical - basis for this machine being limited. The Polywell has been around a long time and has remained "the black sheep" of fusion ideas. It is hard to publish with no funding, it is hard to get funding without papers.

Up close, things are changing. Over the past year, the machine has been presented at two technology conferences, a forum has grown online, the machine was mentioned on CNN and a technical conference has grown up around it. From far away, this is hard to see, because the mainstream has taken no notice of this. What really concerns me is if the Asian economies got their hands on this technology. If a budding nation like China were to put its monetary and technically trained people on this project I am concerned they could go very far with this idea. I do not think America realizes the opportunity that it is passing up here.

There maybe some promising technology here. A community of people has strung up online discussing the potential. Even if the machine cannot be used for power production, there may be other interesting commercial uses. I do not know. We will see what happens next.


Tyler Jordan
Posts: 30
Joined: Sat Mar 29, 2008 2:34 pm
Location: Tasmania

Re: A letter to Newbie about the Polywell

Post by Tyler Jordan »

I know this is an old thread, but only one post in it ... so fair game?

I've posted some wild ideas on talk-polywell from time to time - just to let people know, I'm the 'admin' at protonboron.com ... so, don't think I'm designing reactors or anything critical over there! :lol: I'm just their village idiot ... who reads about things and tries my best to contribute something interesting.

So ...

after reading a lot of papers that I mostly didn't understand, I'm sceptical of the approach of reflecting x-rays back into the machine. Still we will most likely have an abundance of x-rays to deal with and finding some proposition to solve the problem might be useful.

If the vacuum chamber were made of beryllium which from what I've read is x-ray transparent to a large extent, then we could use a tough phosphor (term used loosely to apply to any type of phosphor that could conceivably handle hard x-rays) and then embedded in this phosphor some high efficiency photovoltaics. Perhaps layers of phosphor and photovoltaic. I've seen heavy duty photovoltaics being used experimentally before on large parabolic dishes that concentrate sunlight. In fact there is an experimental power station here in Australia that uses such a system. I'm not certain what frequency range these cells absorb, but they are very heat resistant.

I grant that normal photovoltaics aren't very efficient and certainly doing a conversion with some sort of phosphor is going to make it worse, however these photovoltaics are, I believe, much more efficient - over 40% @25C - which would imply a hell of a lot of cooling. http://www.altenergy.org/renewables/sol ... stone.html.

Additionally, unlike solar, this would be always on! - unlike the sun and in a controlled environment, not outdoors.

Maybe this is just another crazy idea, I'll leave it to the geniuses here to be the judge.


EDIT: just read that Beryllium is a neutron multiplier with input neutrons over about 1.9 MeV - that doesn't seem like a good thing. So wondering what the energies would be from the neutron producing side-reaction in p-B11 fusion? Maybe this is an issue, maybe not, will keep researching ...

EDIT: Posted question on physics stack exchange regarding neutron energy levels in side-reactions in proton-boron fusion. http://physics.stackexchange.com/questi ... ron-fusion

EDIT, just started looking at the cost of Beryllium - I did realize it would be expensive, but whoa! Thinking that making a large vacuum chamber out of this stuff would significantly add to the build cost of the reactor.

So, what other strong materials would be mostly transparent to hard x-rays?

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

Re: A letter to Newbie about the Polywell

Post by D Tibbets »

P-B11 neutrons are mostly a non issue. The flux of neutrons may be less than 1 /10,000,000 that of deuterium- deuterium fusion. If neutron damage limits a D-D fusion reactor to 1 day of operation before it had to be rebuilt, then with P-B11 and everything else being equal, the reactor would last ~ 10 million days or 30,000 years.

Photo voltaics for x-ray conversion may indeed be a solution. Eric Learner of LPP has a patent for such a device. I think the key is that there are multiple photovotaic layers that eventually absorb the highly penetrating X-rays. I don't know the durability or efficiency of the idea. Also, I don't think it has actually been built. The X-ray problem of cooling the plasma (Bremsstruhlung) can be addressed by decreasing Bremsstruhlung - there are two mechanisms that may apply to the Polywell, reflecting and reabsorbing the X-rays as in the title of this thread, or recovering the energy lost through the X-rays- conversion to electricity and recycling back into the reactor. The last, if it works at high efficiency would be a real game changer.

X- ray reflection in hydrogen bombs is practical but very destructive. Only the inertia of the heavy uranium shell/ tamper allows survival for long enough times. Reabsorption is another issue. With the plasma at densities of perhaps 10^28 high Z ions per cubic meter in a bomb, a large portion can be reabsorbed. With densities of 10^22 of low Z ions the absorption path may be many millions of times longer.

PS: If not directly converted to electricity, the X-rays end up as heat in the shell / coils of the reactor. This can be converted by a steam cycle at perhaps 30% efficiency. Or perhaps at even ~ 45 % efficiency if a supercritical CO2 steam cycle is used. For photovoltaics to be useful they would have to convert at efficiencies well above this, or perhaps supplement the thermal conversion. The high capacity cooling will be needed anyway so it is not like you are avoiding a cooling plant. The question is if you can avoid an additional steam/ turbine plant (expensive) and if the conversion efficiency by whatever means is enough to overcome the losses.

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

Post Reply