Sandia researcher Bill Wampler (1111) and a Labs/industry/university team have resolved one of many issues impeding successful magnetic confinement fusion. They discovered a way to keep the fusion plasma from eroding divertor walls inside tokamak fusion machines.

Using the DIII-D tokamak magnetic fusion machine at General Atomics in San Diego and the Divertor Materials Evaluation System (DiMES), Bill and colleagues at the University of California at San Diego, General Atomics, and Sandia recently conducted experiments that showed erosion is eliminated during operation with detached plasmas.

Fusion is the combining or fusing of light atoms, such as atoms of hydrogen, to form heavier atoms. In the process some of the mass of hydrogen is converted into energy the same way the sun creates energy. The goal is to convert this fusion energy into electric power. While there are several different ways to produce fusion, among the most promising is the tokamak, a large doughnut-shaped magnetic confinement device. The word tokamak is an acronym derived from Russian words meaning "toroidal chamber and magnetic coil."

A large current, up to several million amps, flows through the doughnut-shaped plasma, which in a reactor will be made up of deuterium and tritium. The plasma is confined by a strong magnetic field and is heated to temperatures more than a hundred million degrees Celsius by high-energy particle beams or radio-frequency waves. The hot plasma travels quickly parallel to the magnetic field, but slowly across the field toward the wall.

Plasma at the outer boundary flows into a separate chamber where it strikes the divertor surface and is neutralized. The divertor prevents the plasma from striking and degrading the chamber walls and generating impurities that would cool and contaminate the main plasma. Constructed of graphite to survive high heat loads, the divertor can be eroded by hydrogen plasmas.

Erosion is undesirable because it increases the need for replacement of graphite tiles in the divertor, results in contamination of the plasma, and causes material depositions in undesirable places, all problematic for successful fusion.

Bill, Stuart Van Deusen (1111), Bob Bastasz (8724), Dennis Whyte of the University of California at San Diego, and Clement Wong and Phil West of General Atomics conducted experiments that showed that erosion can be completely eliminated by causing plasmas to detach.

Detached divertor plasmas, which have been studied for several years as a method to spread power deposition over larger areas, are produced by injecting deuterium gas into the plasma in the divertor. The gas cools the plasma near the divertor surface but does not significantly cool the core plasma in the main chamber. This reduces the energies of the ions and neutral hydrogen atoms at the divertor surface, totally eliminating divertor erosion.

"Because today’s fusion machines do not operate for long periods [typically less than 10 minutes a day], erosion has not been a significant problem to this point," Bill says. "But when we eventually have a fusion reactor working all the time, the resulting erosion will be a big issue."

The lower temperature of the detached plasma at the divertor walls reduces the energy of hydrogen atoms and ions impinging on the divertor surface. These new experiments show that erosion by "physical sputtering" (erosion process where incoming atoms have sufficient energy to knock atoms off a surface) is eliminated because the energy of particles hitting the material surface is now below the threshold of erosion.

To determine this, the researchers exposed graphite cylinders and various metal films to divertor plasmas. Bill measured the erosion resulting from these exposures using ion-beam analysis. No erosion was detected at any site with detached plasmas. In contrast, high erosion rates were measured for the normal "attached" divertor plasmas obtained without deuterium injection in the divertor.

"We still have much more work to do, but our preliminary results show that detached plasma operation provides an effective way to eliminate erosion in the divertor. We may have resolved a major problem in fusion," Bill says.

Last modified: August 28, 2000