Figure 9.2 is a more realistic drawing of the ITER machine than shown in Chap. 8. The plasma will occupy the D-shaped vacuum space surrounded by tiles. These tiles are the plasma facing components (PFCs), commonly called the “first wall.” They have to withstand a tremendous amount of heat from the plasma and yet must not contaminate the plasma and be compatible with the fusion products that impinge on them. Early tokamaks used stainless steel, but clearly this is not a high-temperature material. Current tokamaks use carbon fiber composites (CFCs), a light, strong, high — temperature material that is used in bicycles, racing cars, and space shuttles. Just as rebars are used to strengthen concrete, carbon fibers are used to strengthen graphite. However, carbon cannot be used in a reactor because it absorbs tritium, which would not only deplete this scarce fuel but also weaken the CFC. After all, hydrocarbons like methane and propane are very common, stable compounds; and tritium is just another form of hydrogen and can be captured by the carbon to form hydrocarbons.

Tungsten is a refractory metal, but it is high-Z; that is, it has a high atomic number and therefore has so many electrons that it cannot be completely ionized. The remaining electrons radiate energy away, cooling the plasma. The good thing about hydrogen and its isotopes is that they have only one electron, and once that electron is stripped free of the nucleus by ionization, the atom can no longer emit light. Beryllium is a suitable low-Z material, but it has a low melting point, and so has to be cooled aggressively. In preparation for ITER, the European tokamak JET is being upgraded with a beryllium first wall. In short, the first-wall material must not absorb tritium and must have a low atomic number, take high temperatures, and be resistant to erosion, sputtering, and neutron damage.

ITER, of course, is only the first step. There are large steps between ITER and DEMO and between DEMO and a full reactor. Some large numbers on the first wall

are given in Table 9.1. We see that the step between ITER and DEMO is much larger than between DEMO and Reactor. Hence the call for a materials-testing facility intermediate between ITER and DEMO.

The fusion power is given in gigawatts. A typical power plant generates 1 GW of electricity; and perhaps 5 GW of fusion power is needed to give that, since the tokamak needs power to run, and there is still a heat cycle in a steam plant to produce electricity. The heat flux impinging on the first wall is about 0.5 MW/m2. This translates to 50 W/cm2 or about 300 W/sq. in. This is not much more than the surface of an electric iron, though the total heat is considerable. The real problem is in the divertor, which has to handle most of the heat from the plasma. Divertors will be covered later.

The neutron load is the energy of the 14-MeV neutrons from the D-T reaction which pass through the first wall. This energy is not deposited in the first wall, but the neutrons damage the wall. The neutron load summed over the life of the wall is what matters. This is much larger for a reactor than for ITER, since ITER is just an experiment, while a reactor should last about 15 years before it has to be revamped. The neutron damage is measured in displacements per atom (dpa). The longer the material is exposed to a neutron flux, the more times one of its atoms will be knocked out of place by a neutron. After many dpa’s, the material will swell or shrink and become so brittles as to be useless.

Beryllium melts so easily that it cannot be used in a reactor. Boron coating has been tried successfully, but also cannot take high temperatures. Tungsten seems to be the best available wall material because it does not erode or sputter easily and has a high melting point of 3,410°C. However, it is a high-Z material and also cannot be machined easily. A liquid lithium first wall has been considered, but it is no longer proposed.1 Silicon carbide (SiC) is a promising material that has been studied extensively in the laboratory but does not have a known method for manufac­turing in large quantities [1]. How SiC compares with other materials in operating temperature is shown in Fig. 9.3. These temperature ranges are for irradiated materials so that the swelling and fracture effects caused by neutrons are included. Carbon fiber-reinforced graphite (C/C) can take high temperatures, but carbon cannot be used because of tritium retention. Tungsten and molybdenum are classical refractory metals but will cool the plasma if they sputter into it. Silicon carbide reinforced with layers of SiC fibers (SiC/SiC) seems to be the ideal material for the first wall if it can be made without impurities. It takes high temperature, is quite strong, and is resistant to radiation damage. It can last for the 15-year life of the

reactor. Its properties have been measured in fission reactors [2]. The main drawback is a thermal conductivity lower than for other materials.

The latest high-tech material is a SiC matrix/graphite fiber composite [1], which has increased thermal conductivity in addition to the other good properties. These advanced materials cannot be designed with existing computer programs, which are applied only to metals. Some reactor studies assume that SiC first-wall material will be available. Though SiC composites have tremendous potential, much research and testing remain to be done before they become a reality.