The Farnsworth–Hirsch Fusor, or simply fusor, is an apparatus designed by Philo T. Farnsworth to create nuclear fusion. It has also been developed in various incarnations by researchers including Elmore, Tuck, and Watson, and more lately by George Miley and Robert W. Bussard. Unlike most controlled fusion systems, which slowly heat a magnetically confined plasma, the fusor injects "high temperature" ions directly into a reaction chamber, thereby avoiding a considerable amount of complexity. The approach is known as inertial electrostatic confinement.
Hopes at the time were high that it could be quickly developed into a practical power source. However, as with other fusion experiments, development into a generator has proven difficult. Nevertheless, the fusor has since become a practical neutron source and is produced commercially for this role. It has been assembled in low-power forms by hobbyists.
The fusor was originally conceived by Philo Farnsworth, better known for his pioneering work in television. In the early 1930s he investigated a number of vacuum tube designs for use in television, and found one that led to an interesting effect. In this design, which he called the multipactor, electrons moving from one electrode to another were stopped in mid-flight with the proper application of a high-frequency magnetic field. The charge would then accumulate in the center of the tube, leading to high amplification. Unfortunately it also led to high erosion on the electrodes when the electrons eventually hit them, and today the multipactor effect is generally considered a problem to be avoided.
What particularly interested Farnsworth about the device was its ability to focus electrons at a particular point. One of the biggest problems in fusion research is to keep the hot fuel from hitting the walls of the container. If this is allowed to happen, the fuel cannot be kept hot enough for the fusion reaction to occur. Farnsworth reasoned that he could build an electrostatic plasma confinement system in which the "wall" fields of the reactor were electrons or ions being held in place by the multipactor. Fuel could then be injected through the wall, and once inside it would be unable to escape. He called this concept a virtual electrode, and the system as a whole the fusor.
His original fusor designs were based on cylindrical arrangements of electrodes, like the original multipactors. Fuel was ionized and then fired from small accelerators through holes in the outer (physical) electrodes. Once through the hole they were accelerated towards the inner reaction area at high velocity. Electrostatic pressure from the positively charged electrodes would keep the fuel as a whole off of the walls of the chamber, and impacts from new ions would keep the hottest plasma in the center. He referred to this as inertial electrostatic confinement, a term that continues to be used to this day.
Various models of the fusor were constructed in the early 1960s. Unlike the original conception, these models used a spherical reaction area but were otherwise similar. Farnsworth ran a fairly "open" lab, and several of the lab techs also built their own fusor designs. Although generally successful, the fusor had a problem being scaled up: as the fuel was delivered via accelerators, the amount of fuel that could be used in the reaction was quite low.
Things changed dramatically with the arrival of Robert Hirsch at the lab. He proposed an entirely new way of building a fusor without the ion guns or multipactor electrodes. Instead the system was constructed as two similar spherical electrodes, one inside the other, all inside a larger container filled with a dilute fuel gas. In this system the guns were no longer needed, and corona discharge around the outer electrodes was enough to provide a source of ions. Once ionized, the gas would be drawn towards the inner (negatively charged) electrode, which they would pass by and into the central reaction area.
The overall system ended up being similar to Farnsworth's original fusor design in concept, but used a real electrode in the center. Ions would collect near this electrode, forming a shell of positive charge that new ions from outside the shell would penetrate due to their high speed. Once inside the shell they would experience an additional force keeping them inside, with the cooler ones collecting into the shell itself. It is this later design, properly called the Hirsch–Meeks fusor, that continues to be experimented with today.
New fusors based on Hirsch's design were first constructed in the later 1960s. Even the first test models demonstrated that the design was effective; soon they were showing production rates of up to a billion neutrons per second, and rates of up to a trillion per second have been reported.
All of this work had taken place at the Farnsworth Television labs, which had been purchased in 1949 by ITT Corporation with plans of becoming the next RCA. In 1961 ITT placed Harold Geneen in charge as CEO. Geneen decided that ITT was no longer going to be a telephone/electronics company, and instituted a policy of rapidly buying up companies of any sort. Soon ITT's main lines of business were insurance, Sheraton Hotels, Wonderbread and Avis Rent-a-Car. In one particularly busy month they purchased 20 different companies, all of them unrelated. It didn't matter what the companies did, as long as they turned a profit.
A fusion research project was not regarded as immediately profitable. In 1965 the board of directors started asking Geneen to sell off the Farnsworth division, but he had his 1966 budget approved with funding until the middle of 1967. Further funding was refused, and that ended ITT's experiments with fusion.
The team then turned to the AEC, then in charge of fusion research funding, and provided them with a demonstration device mounted on a serving cart that produced more fusion than any existing "classical" device. The observers were startled, but the timing was bad; Hirsch himself had recently revealed the great progress being made by the Soviets using the tokamak. In response to this surprising development, the AEC decided to concentrate funding on large tokamak projects, and reduce backing for alternative concepts.
Farnsworth then moved to Brigham Young University and tried to hire on most of his original lab from ITT into a new company. The company started operations in 1968, but after failing to secure several million dollars in seed capital, by 1970 they had spent all of Farnsworth's savings. The IRS seized their assets in February 1971, and in March Farnsworth suffered a bout of pneumonia which resulted in his death. The fusor effectively died along with him.
In the early 1980s, disappointed by the slow progress on "big machines", a number of physicists took a fresh look at alternative designs. George Miley at the University of Illinois picked up on the fusor and re-introduced it into the field. A low but steady interest in the fusor has remained since then. An important development was the successful commercial introduction of a fusor-based neutron generator. From 2006 until his death in 2007, Robert W. Bussard gave talks on a reactor similar in design to the Fusor, now called Polywell, that he stated would be capable of useful power generation.
Nuclear fusion refers to reactions in which light nuclei are combined to become heavier nuclei. Several such reactions release energy that can, in principle, be harnessed to provide fusion power. The lowest energy reaction occurs in a mix of deuterium and tritium, when the ions have to have a temperature of at least 4 keV (kiloelectronvolts), equivalent to about 45 million kelvins. At such temperatures, the fuel atoms are ionized and constitute a plasma. In a practical fusion power plant, fusion reactions have to occur fast enough to make up for energy losses. The rate of reaction varies with the temperature and the density of the fuel and the loss rate is characterized by the energy confinement time τE. The minimum conditions required are expressed in the Lawson criterion. In the most successful approach, magnetic confinement fusion, the necessary conditions are approached by heating a plasma contained by magnetic fields. This has proven to be very difficult in practice. The complexity of the systems applied detracts from the usefulness of the design for a practical generator.
In the original fusor design, several small particle accelerators, essentially TV tubes with the ends cut off, inject ions at a relatively low voltage into a vacuum chamber. In the Hirsch version of the fusor, the ions are produced by ionizing a dilute gas in the chamber. In either version there are two concentric electrodes, the inner one being charged negatively with respect to the outer one to about 80 kV. Once the ions enter the region between the electrodes, they are accelerated towards the center.
In the fusor, the ions are accelerated to several keV by the electrodes, so heating as such is not necessary (as long as the ions fuse before losing their energy by any process). Whereas 45 megakelvins is a very high temperature by any standard, the corresponding voltage is only 4 kV, a level commonly found in such devices as neon lights and televisions. To the extent that the ions remain at their initial energy, the energy can be tuned to take advantage of the peak of the reaction cross section or to avoid disadvantageous (for example neutron-producing) reactions that might occur at higher energies.
The ease with which the ion energy can be increased appears to be particularly useful when "high temperature" fusion reactions are considered, such as proton-boron-11, which has plentiful fuel, requires no radioactive tritium, and produces no neutrons in the primary reaction.
Because an electrostatic potential well cannot simultaneously trap both ions and electrons, there must be some regions of charge accumulation, which will result in an upper limit on the achievable density. The corresponding upper limit on the power density, even assuming D-T fuel, may be too low for power production.
When they first fall into the center of the fusor, the ions will all have the same energy, but the velocity distribution will rapidly approach a Maxwell-Boltzmann distribution. This would occur through simple Coulomb collisions in a matter of milliseconds, but beam-beam instabilities will occur orders of magnitude faster still. In comparison, any given ion will require a few minutes before undergoing a fusion reaction, so that the monoenergetic picture of the fusor, at least for power production, is not appropriate. One consequence of the thermalization is that some of the ions will gain enough energy to leave the potential well, taking their energy with them, without having undergone a fusion reaction.
There are a number of unsolved challenges with the electrodes in a fusor power system. To begin with, the electrodes cannot influence the potential within themselves, so it would seem at first glance that the fusion plasma would be in more or less direct contact with the inner electrode, resulting in contamination of the plasma and destruction of the electrode. However, the majority of the fusion tends to occur in microchannels formed in areas of minimum electric potential , which present themselves as visible "rays" penetrating the core. These form because the forces within the region correspond to roughly stable "orbits". 40% or so of the high energy ions in a typical grid operating in star mode may be within these microchannels Nonetheless, grid collisions remain the primary energy loss mechanism for Farnsworth-Hirsch fusors. Complicating issues is the challenge in cooling the central electrode; any fusor producing enough power to run a power plant seems destined to also destroy its inner electrode. As one fundamental limitation, any method which produces a neutron flux that is captured to heat a working fluid will also bombard its electrodes with that flux, heating them as well.
Attempts to resolve these problems include Bussard's Polywell system, D. C. Barnes' modified Penning trap approach, and the University of Illinois's fusor which retains grids but attempts to more tightly focus the ions into microchannels to attempt to avoid losses. While all three are IEC devices, only the latter case is actually a "fusor".
One oft-presented concern is Bremsstrahlung (German for "braking radiation"). In Fundamental limitations on plasma fusion systems not in thermodynamic equilibrium, Todd Rider shows that a quasineutral isotropic plasma will lose energy due to Bremsstrahlung at a rate prohibitive for any fuel other than D-T (or possibly D-D or D-He3). This paper is not applicable to IEC fusion, as a quasineutral plasma cannot be contained by an electric field, which is a fundamental part of IEC fusion. However, in a further paper, "A general critique of inertial-electrostatic confinement fusion systems", Rider addresses the common IEC devices directly, including the fusor. In the case of the fusor the electrons are generally separated from the mass of the fuel isolated near the electrodes, which limits the loss rate. However, Rider demonstrates that practical fusors operate in a range of modes that either lead to significant electron mixing and losses, or alternately lower power densities. This appears to be a sort of catch-22 that limits the output of any fusor-like system.
Regardless of its possible use as an energy source, the fusor has already been demonstrated as a viable neutron source. Fluxes are not as high as can be obtained from nuclear reactor or particle accelerator sources, but are sufficient for many uses. Importantly, the neutron generator easily sits on a benchtop, and can be turned off at the flick of a switch. A commercial fusor was developed as a non-core business within DaimlerChrysler Aerospace - Space Infrastructure, Bremen between 1996 and early 2001 After the project was effectively ended, the former project manager established a company which is called NSD-Fusion
Each electrode is spot-welded from hoops of stainless-steel wire (often welding rod) at right angles. The fusor's electrode dimensions are not very critical. The outer electrode can range from baseball to beach-ball size (100 to 600 mm diameter), and the inner from ping-pong ball to baseball size (40 to 100 mm diameter). Usually such projects use the high-voltage transformer from a neon sign or x-ray machine, and high voltage rectifier from a hobby shop. Spark plug wires carry the power, with spark plugs or similar ceramic insulators to pass it into the vacuum chamber. Deuterium is available in lecturer bottles and is not a controlled nuclear material. Neutrons can be detected by measuring induced radioactivity in aluminium, silver or indium foil after moderating the neutrons with wax, water or plastic, or a plastic neutron luminescent material can be used with a photodetector. Advanced and sensitive neutron detectors using boron trifluoride or helium-3 filled tubes are becoming increasingly common, but a functioning neutron counter is very hard to find on the surplus market. The major expense is usually the vacuum pump.
The voltages are extremely dangerous (exceeding 20,000 volts), and neutron emissions can present a hazard if voltages above 40 kilovolts are used. The X-ray emissions are the greatest radiological hazard associated with the Fusor, and proper steps must be taken to shield X-ray transparent regions such as viewports.