Site logo

Magnetic Fusion

Magnetic traps for Fusion plasmas.

Stacks Image 579
Spin tight and fast.
1. The Spinner
Stacks Image 597
Mirror + straight Cusp shapes a magnetic humbug.
2. Magnetic bottles
Stacks Image 647
World's largest superconducting Mirror coils. Scrapped by Reagan for Star Wars.
(3) Tennis Ball Coil
Stacks Image 644
JET Torus and Coils.
(4) Magnetic Torus
Stacks Image 860
Terrorists put JET at Culham.
(5) Bas Pease
Stacks Image 573
The ITER Reactor and technician.- page background.
(6) ITER Torus
Stacks Image 581
Sculpted Mirror-Stellarator coils.
(7) Stellarator Torus
Stacks Image 584
DII-D scale for small SphericalTokamak Reactor. The man is 5'6"
(8) Glowing plasma surface in START Torus
Stacks Image 587
Ivanov: GDT Cutaway. Novosibirsk, Russia
(9) Mirror trap as Neutron source

Using a magnetic field to hold a fusion plasma, a gas at 100 million degrees, is not as simple as picking up a nail with a solid magnet. We first need to know how a single nucleus moves in a magnetic field and then design some kind of magnetic bottle, sphere, or closed toroidal tube to contain a plasma.
The story is illustrated in the left side gallery and covers much of the history of Magnetic Fusion. The right hand gallery takes you to discussions on the various magnetic fusion machines.

The Spinner (1)
There is a neat way to visualise how the charged particles move in a magnetic field as it is very similar to the mechanical movements of a girl on a playground spinner.
The green pole represents the vertical direction of the magnetic field. Stand on the spinner and nothing happens. Jump up an down and still nothing happens. Similarly, a charged particle can move along a straight magnetic field without hindrance.
Now, the girl holds onto the pole and pushes forward from it to spin slowly round in a circle. When the girl pulls inwards she spins much faster. Click to see the result.
If she now jumps up and down then her motion will be in up and down spirals.
So, single charged particles in magnetic fields move in spirals, a bit like flying in a continuous barrel roll. Controlling them is like picking up spiral spaghetti on the back of a fork.

Particles in a Magnetic Field.
An electron or nucleus trying to cross a magnetic field instead spins around. Its path unfolds as a spiral due to its motion parallel to the field. The field does not trap the particles along the field lines.
Of course, magnetic field lines are not real lines but just map the direction of the magnetic field, like the lines of pressure on a weather chart. Magnetic fields arise from permanent natural magnets or from electrical currents in conducting wires. Computer codes, using Maxwell's beautiful equations for the interactions of electric and magnetic fields and currents and charges can map the complicated geometries of the resulting magnetic fields.

You will see many variations in magnetic trap designs.

Closing the Trap: Magnetic Bottles (2)
We need to stop the particles from running away up the field line. The simplest way to do this is to compress and strengthen the magnetic field, converting the parallel motion into spin. The compressed field can act as a mirror, reflecting the particles back. This layout did not work since the plasma as a whole could wobble out sideways. Addition of a set of parallel coils stabilised the plasma. As shown in the next picture, a pair of circular coils, carrying a large current,can provide a 3D plasma trap.

Gross Instability due to Bad Field Curvature.
The simple magnetic trap we showed suffers from a gross instability like an aneurism, the whole plasma wobbling out into the wall. It happens because the field lines already bulge outwards and that is simply amplified.
We have learned how to sculpt magnetic coils to make the fields curve as we wish to give gross stability. The first success with mirrors was to add a separate set of straight coils which squeezed the bundle of field lines through a vertical ellipse at one end and a horizontal ellipse at the other. The magnetic trap had a humbug shape with concave curvature along most of the bundle.
A keen tennis playing computer modeller realised the coil sets could be combined into a single loop like the seam on a tennis ball (3).
We learned something deeper from the failure of the simple magnetic mirror machine. The magnetic field strength B in the centre of the humbug is lower than at the edges. The mathematics of the plasma equilibrium in any minimum-B configuration is completely stable. This was a very useful
theorem found from failure.

Superconducting Coils
Large magnetic coils need large currents which consume a lot of power. It is essential that the coils for reactors be superconducting.
The final version of the tennis ball coil was to splitting it in two, forming a Yin-Yang coil set. The ultimate set was built at Lawrence Livermore Lab. in 1984, with superconducting, Niobium-Tin, coils, cooled by liquid helium.
Click on the tennis ball to see these coils being hauled away on a truck after Ronald Reagan cancelled the program to divert funds and scientists into the witless Star Wars project. Reagan carried on to cut Fusion funding by 80% which is the only reason success is still 40 years away.

Dirt: A big obstacle to reaching Fusion Temperatures?

As we heat matter it changes state from solid to liquid to gaseous forms. At the 100 million degrees needed for sufficient Deuterium-Tritium fusion reactions to take place these hydrogen atoms are torn apart into a
plasma of their electrons and nuclei. (Plasma is a Greek word for stuff!).
In fact, most atoms lose some or all their electrons at a temperature of 1 million degrees. Heavier atoms, from carbon upwards, keep recombining with an electron, emitting hot electromagnetic radiation, and cooling the plasma. Impurity radiation heat loss was the main barrier to all early fusion experiments. The world programme seemed to be failing on this issue.

All surfaces in the plasma chamber must be extremely clean. The chamber must be pumped down to the vacuum of outer space, removing air, water and other gases.

Nature did not cooperate willingly in achieving fusion on earth. In 1958, a meeting between US, British, French and Russian scientists in Geneva declassified the Fusion Energy project. Cooperation would be better than secrecy.

Plasma diffusion losses
This trap still leaks a bit at the ends, losing particles that ran almost parallel to the field line at the centre. They do not spin up enough to reflect. Collisions between particles will knock some into this directional loss cone, so this simple trap leads to a low performance reactor.
The trap can be improved with some additional end stopper or electric fields to cut the losses.
The more obvious alternative is to lead the field lines into closed loops. There is a considerable family of toroidal and spherical devices which do this.
In all traps, particle collisions scatter them from one spiral orbit to another. The plasmas will all diffuse slowly out of the traps.
That is not the only thing to worry about.

Plasma instability
A hot magnetised plasma is a far better conductor than gold and many types of electromagnetic waves are possible. Some can feed on plasma motions to grow and destroy the confinement, ejecting plumes of plasma into the chamber walls. Try getting a grip on a piece of jelly!
Control or elimination of the myriad possible ways that a plasma can become unstable is central to successful reactor design.
Alternatively, we can heat the plasmas with microwaves resonating with the spiralling of the ions or the much faster spiralling of electrons. Heating just the electrons or just the ions can be used to drive large currents in plasmas.

Reaching Fusion Energy
The only way to heat a static plasma without large currents or rapid compression is by injecting the D-T fuel at high enough current and energy, > 10keV. The Livermore Mirror Program contracted Kunkel and Pyle at UC Berkeley to invent something to do that. The charged beam from a simple ion accelerator is pushed apart by its own electric fields. Kunkel's team neutralised the beam of Deuterium by passing it through a neutral gas. The fast ions pass out as a neutral beam straight into the plasma chamber. The 2XII Mirror at LLNL was the first to reach fusion energies and densities. All other magnetic fusion machines now use neutral beams as a primary heating source. As the beams re-ionise in the plasma they can also drive a net electrical current in a Tokamak.

The Tokamak: The most successful Toroidal magnetic trap.

The simplest idea for a closed field line device would be a set of small circular coils arranged around a large circle to make a donut shaped trap. Every field line would then make a single closed field line the long way round the
torus. Unfortunately a plasma has enough options to just separate into set of quoit like rings and move off into the walls. Completely unstable.
One solution is to drive a large current round the plasma, making it a second part of the overall coil set. It adds a magnetic field component circling the short way round. The combined effect gives a small helical twist to the field lines. A field line does not close on itself but can wrap around an infinite number of times on the same surface, which is just as good. All the particles on a magnetic surface remain on it till a collision knocks it off. A single field line threading a compact magnetic torus is shown (

Sakharov & Tamm, the theorists behind the Russian Fusion bomb, invented the most successful version of this device, the
Tokamak, now leading world fusion development. Precise calculations of the magnetised plasma configuration found conditions for strong stability.

In 1969, the Kurchatov Laboratory announced that they had reached temperatures of 10 million degrees. Others had made false claims before, including Harwell, so we did not believe them. Still, a team from the Culham Fusion Laboratory was dispatched to make accurate laser temperature measurements and they confirmed the achievement. Many efforts were abandoned around the world and Tokamaks built.

Making Politics work.
By 1972, (5) Bas Pease at Culham led other European Labs to propose a joint effort to build a prototype reactor,the Joint European Torus,
JET (4). The design work was run on optional funds. The EU Commission had to agree to the project but politicians fought for 5 years over where to build it. It was decided by terrorists. The Bader-Meinhoff gang hijacked an airliner and held it hostage in Stuttgart. A UK SAS team stormed the plane, shot the terrorist, and saved all the passengers and crew (5). Germany yielded the site to Culham.

The JET project, and similar size machines in Russia, China, Japa and the USA, solved most of the remaining physics and plasma stability problems. Smaller machines across the world contributed understanding at every level. The follow on international project is ITER (6), being constructed in France, and expected to produce 500MW of near continuous or long pulse fusion power by 2030. An electric power plant is expected to follow.

Engineering and Breeder Blankets
Most of the problems for ITER are in the engineering of such a large device running at high power loads for every component. New materials able to withstand high neutron fluxes have to be developed.
The steel chamber walls are almost transparent to neutrons which are absorbed in a blanket containing Lithium-6. Absorb a neutron and it disintegrates into a Tritium and a Helium nucleus. The Tritium fuel burned in the plasma is replaced by a new one bred in the blanket.
Whatever the outcome, the ITER project will have solved most of the materials and engineering problems for centuries of Fusion power.

The Tokamak uses the plasma as one of the principal magnetic coils. The current has to be induced for startup and then driven by powerful beams of particles or microwaves to keep it running.
An earlier scheme, the
Stellarator (7), is still in business. The idea by Spitzer at Princeton was to make the external coils themselves into three separate helices. This provides the same type of magnetic surfaces as the Tokamak but without a toroidal plasma current drive. It only needs fuel to be injected and cooling alpha particles, the ash from fusion, to be removed.
A cunning set of sculptured coils is now preferred. These are like the Yin-Yang coils except that they are also mounted in a twisted pattern around the torus. This is essentially a set of stable mirror traps linked to make closed magnetic surfaces for the plasma to ride on. We show a cutaway of this German design.

All this is taking too long and will not contribute to global energy supplies till the second half of this century.
The physics of magnetic confinement is well understood. Smaller scale machines can therefore be built today, using existing materials and industrial strength versions of the technologies to produce Fusion power in the low 20-100MW range. Not much use as an electricity provider, but a huge source of high energy neutrons and what they can do. Small machines can breed fission reactor fuel from depleted Uranium, burn nuclear wastes, generate startup fuel for Thorium reactors, make new medical isotopes, test advanced materials, and much more.
Only private enterprise can carry these plans forward rapidly but the rewards, in about 4 election cycles will be colossal. The Economics Avenue will explore the commercial benefits of innovation.

And Now ….

Go and look at some of the overviews and key lectures which expand on all these themes, or look up the Politics and Economics Avenues to see what non-scientists think.

There is a pot of gold at the end of this book.
Stacks Image 31
Fusion Overviews
Magnetic Fusion Library
Stacks Image 271
JET robot cartoon
Magnetic Fusion Reports
Brazil Fusion Talk
Magnetic Fusion Lectures