A levitated dipole is a nuclear fusion experiment using a superconducting torus which is magnetically levitated inside the reactor chamber. It is believed that such an apparatus could contain plasma more efficiently than other fusion reactor designs. The superconductor forms an axisymmetric magnetic field of a nature similar to Earth's or Jupiter's magnetospheres. The machine was run in a collaboration between MIT and Columbia University.
The Levitated Dipole is designed to be stable against "gentle" changes in the electric or magnetic field. This makes the Levitated Dipole unique when compared with other magnetic confinement machines. In those experiments, small fluctuations can cause significant energy loss. By contrast, in a dipolar magnetic field, fluctuations tend to compress the plasma, without energy loss. This compression effect was first noticed by Akira Hasegawa (of the Hasegawa-Mima equation) after participating in the Voyager 2 encounter with Uranus.
Adapting this concept to a fusion experiment was first proposed by Dr. Jay Kesner (MIT) and Dr. Michael Mauel (Columbia) in the mid to late nineties. The pair assembled a team and raised money to build the machine. They achieved first plasma on Friday, August 13, 2004 at 12:53 PM. First plasma was done by (1) successfully levitating the dipole magnet and (2) RF heating the plasma. The LDX team has since successfully conducted several levitation tests, including a 40-minute suspension of the superconducting coil on February 9, 2007. Shortly after, the coil was damaged in a control test in February 2007 and replaced in May 2007. The replacement coil was inferior, a copper wound electromagnet, that was also water cooled. Scientific results, including the observation of an inward turbulent pinch, were reported in Nature Physics.
This experiment needed a very special free-floating electromagnet, which created the unique "toilet-bowl" magnetic field. The magnetic field was originally made of two counter-wound rings of currents. Each ring contained a 19-strand niobium-tin Rutherford cable (common in superconducting magnets). These looped around inside a Inconel magnet; a magnet that looked like an oversized donut. The donut was charged using induction. Once charged, it generated a magnetic field for roughly an 8-hour period. Overall, the ring weighed 450 kilograms and levitated 1.6 meters above a superconducting ring. The ring produced roughly a 5-tesla field. This superconductor was encased inside a liquid helium, which kept the electromagnet below 10 kelvins. This design is similar to the D20 dipole experiment at Berkeley and the RT-1 experiment at the University of Tokyo.
The dipole was suspended inside a mushroom-shaped vacuum chamber, which was about 5 meters in diameter and ~3 meters high. At the base of the chamber was a charging coil. This coil is used to charge the dipole, using induction. The coil exposing the dipole to a varying magnetic field. Next, the dipole is raised into the center of the chamber. This could be done with supports or using the field itself. Around the outside of this chamber were Helmholtz coils, which were used to produce a uniform surrounding magnetic field. This external field would interact with the dipole field, suspending the dipole. It was in this surrounding field that plasma moved. The plasma forms around the dipole and inside the chamber. The plasma is formed by heating a low pressure gas. The gas is heated using a radio frequency, essentially microwaving the plasma in a 17-kilowatt field.
The machine was monitored using diagnostics fairly standard to all of fusion. These included:
Single particles corkscrew along the field lines, flowing around the dipole electromagnet. This leads to a giant encapsulation of the electromagnet. As material passes through the center, the density spikes. This is because lots of plasma is trying to squeeze through a limited area. This is where most of the fusion reactions occur. This behavior has been called a turbulent pinch.
In large amounts, the plasma formed two shells around the dipole: a low density shell, occupying a large volume and a high density shell, closer to the dipole. This is shown here. The plasma was trapped fairly well. It gave a maximum beta number of 0.26. A value of 1 is ideal.
There were two modes of operation observed:
In the case of deuterium fusion (the cheapest and most straightforward fusion fuel) the geometry of the LDX has the unique advantage over other concepts. Deuterium fusion makes two products, that occur with near equal probability:
|D + D||→ T||+ 1H|
|D + D||→ 3He||+ n|
In this machine, the secondary tritium could be partially removed, a unique property of the dipole. Another fuel choice is tritium and deuterium. This reaction can be done at lower heats and pressures. But it has several drawbacks. First, tritium is far more expensive than deuterium. This is because tritium is rare. It has a short half-life making it hard to produce and store. It is also considered a hazardous material, so using it is a hassle from a health, safety and environmental perspective. Finally, tritium and deuterium produces fast neutrons which means any reactor burning it would require heavy shielding.