Thermonuclear fusion

Nuclear fusion reactions occur when gases of light atoms (hydrogen is one example) are heated to high temperatures of several tens of millions of degrees. These reactions are commonly observed, as they are typical of the way stars work, and are, in particular, the source of the heat and light we receive from the Sun. Mastering such reactions on Earth, for energy production purposes, would open the way to virtually unlimited resources, since every liter of seawater from which 3.3 mg of deuterium is extracted would become the energy equivalent of more than 250 liters of oil. We might also add that the product of these reactions, helium, is chemically and radiologically harmless, and does not contribute to the greenhouse effect. So, while the interest in mastering fusion is great, the difficulties to be overcome are just as great. In stars, the force that keeps the reactive medium in equilibrium is gravity; given the masses involved, such a mechanism is impossible to reproduce on Earth.

Other avenues have therefore been explored. The remainder of this text will examine in parallel the two research avenues of magnetic confinement fusion (or FCM) and inertial confinement fusion (or ICF). This examination will begin with the physical bases essential to understanding the phenomena, and will continue with a look at the problems of the fusion reactor.

It's worth pointing out here that, despite the difficulties encountered, fusion machines have recently been able to produce several tens of megajoules by effectively "melting" the nuclei of two hydrogen isotopes. These results, which are well understood and well repeated, are real successes for physics research, but they also allow extrapolations that can now go as far as the power reactor. It is therefore no longer unrealistic to think in terms of a fusion reactor, and to make plans for the years ahead, when we already know how hungry we will be for energy.

Let's be clear on one point that we won't return to later. The subject is so vast and diverse that we'll confine ourselves, on the one hand, to the basics essential to a first approach to the subject and, on the other hand, without concern for detail, to the orders of magnitude necessary to anchor the subject in the engineer's reality.