Heavy ion fusion revisited The two schemes which remain under study as 'drivers' to achieve controlled thermonuclear fusion via heavy ion bombardment of deuterium-tritium pellets. On the left radio-frequency linacs and storage rings produce the high energy, high intensity bombarding ion beams. On the right, an induction linac, accelerating several beams simultaneously, produces the required beams. Enthusiasm for the heavy ion route to thermonuclear fusion has had its ups and downs since it first flourished in 1976. Now it seems to be on the way up, judging by the discussions at the 'Heavy Ion Accelerators and their Applications to Inertial Fusion' Symposium held at the Institute for Nuclear Study in Tokyo in January. (We are grateful to Helmut Reich for communicating information from this Symposium.) The basic idea is to bombard deu­ terium-tritium pellets with intense beams of high energy heavy ions so as to compress the pellets to the point where fusion occurs (and ulti­ mately, of course, to use the energy liberated in the deuterium-tritium reaction to produce electricity). De­ spite the fiendish complications in­ volved in achieving this objective, there are those who advocate this technique as being more likely to achieve controlled thermonuclear fusion for power reactors than any other currently under investigation (including the conventional magnetic confinement technique studied, for example, at the Tokamak Fusion Test Reactor at Princeton in the US and the Joint European Torus at Cul- ham in the UK). To set the scale of the complica­ tions — the heavy ion beam energy would need to be in the region of 10GeV and have an intensity of some 30 kA, bombarding the fuel pellets from both directions. Infor­ mation on appropriate pellet design dropped after it was realized that lets. This is the scheme under study has become more open than it was high currents and lower energies are predominantly in Japan and in the some years ago, but a lot of thorough inappropriate. Federal Republic of Germany. work on the physics of these target One driver uses conventional tech­ Work on ion sources to produce pellets remains to be done (stopping nologies, although aimed at limits yet intense beams and on radio-frequen­ power, effect of pre-heating, energy to be achieved. It has a cascade of cy quadrupoles for the linac inputs is transport parameters, possible in­ r.f. linacs progressively funnelling quite well advanced. At the Tokyo stabilities, etc.). ions from many sources into a single Institute for Nuclear Study, a small To provide the necessary ion beam as the energy is increased. storage ring, TARN (Test Accumula­ beams, two schemes (usually known Storage rings subsequently get the tion Ring for NUMATRON — a heavy as 'drivers') remain under study, and beam to the right intensity and the ion project), and LITL (Lithium Ion all synchrotron schemes have been right form for bombarding the pel­ Test Linac) are being used for beam 188 CERN Courier, June 1984 studies. A larger (25 m diameter) (1985—90) with a storage ring at a 'High Temperature Experiment' by version, TARN II, is proposed to ex­ Darmstadt and with target experi­ the end of the decade with 6.5 A tend the ion energies to 450 MeV per ments. sodium ion beams taken down nucléon to investigate such aspects Before leaving the r.f. linac scheme 500 m of induction linac to an energy as multiturn injection at high cur­ for the driver of a fusion reactor, it is of 125 MeV. Amongst the main pre­ rents, beam bunching and instabili­ worth mentioning the work at Wes- paratory studies, cores of the new ties, and the possibilities of beam tinghouse in the US, where the think­ Metglas material are being tested. cooling. The Japanese have also pre­ ing has been carried all the way There was one new idea from AI pared a conceptual design known as through to the eventual electrical Maschke presented at the Sympo­ fHIBLIC (Heavy Ion Beam and Lithium power generator. Other studies at sium under the name 'momentum Curtain). It has 16 ion sources for the Rutherford Appleton Laboratory rich beams'. It suggests trying to singly charged ions of lead 208 and in the UK have shown that their Spal­ achieve the necessary energy trans­ radio-frequency quadrupoles pre­ lation Neutron Source, shortly to fer to the deuterium-tritium pellets, ceding the linac cascade. Alvarez come into operation, will be useful not by energy deposition, as the ions linacs complete acceleration to (operating in storage ring mode) for are brought to a halt in the pellets, 15 GeV prior to bunchers and stor­ some important simulation studies but simply by momentum transfer on age rings. on heavy ion beam dynamics. collision. The need is then for a beam In Germany, a six-year research The second driver scheme under of neutral particles of total mass over and development programme began study is based on the use of linear one milligram (for example 1 MV, in 1979, based at the GSI Labora­ induction accelerators (where the ac­ 100 kA caesium beams) transferring tory, Darmstadt (the home of the celerating field gradients are pro­ some 25 per cent of their energy on UNILAC heavy ion linac), Garching, vided by changing the flux in cores collision. Karlsruhe, and the Max Planck Insti­ around the beam) has been success­ The Symposium showed that the tute of Quantum Optics, together fully used in electron acceleration but interest in pursuing the concepts of with several universities (including not yet for heavy ions other than in heavy ion drivers for fusion is still Wisconsin). They have also pre­ single modules. The scheme has re­ very much alive and a number of pared a conceptual design know as ceived most attention at Berkeley in research programmes will contribute HIBALL with eight bismuth ion the US and the requirements for a more information in the coming isources, linacs and storage rings full-scale driver have been sketched years. Schemes to reach the stage of finishing up with 10 GeV, 2.5 kA out. Four 3 MeV ion beams are en­ a High Temperature Experiment, pulses being fired at the pellets. visaged at the linac input and the out­ where ion interactions in a hot plas­ There are plans to extend the work put is some 3 kA of ions at an energy ma can be confined, are on the table beyond the present programme of about 10 GeV. There are plans for in Germany, Japan and the US. Around the Laboratories proton-antiproton Collider are never­ gramme now approved (see Janua­ CERN theless six years old. ry/February issue, page 23), sub­ UA1 gets a refit Since it recorded its first events stantial increases in the proton-anti- back in 1981, the mammoth detec­ proton collision rate can be expected tor has operated to its design speci­ over the next few years. The detec­ Although still basking in the glory of fications and schedules, and most of tor itself is looking towards new phy­ last year's epic discoveries of the W the physics goals foreseen in the ini­ sics. Thus the UA1 team has em­ and Z particles, and with recent in­ tial design are now science history. barked on an ambitious improve­ dications of still more interesting However techniques at the SPS ment programme, the first phase of physics to come (see May issue, proton-antiproton Collider have which is scheduled to be ready for page 139), the design concepts be­ evolved over the years, and with this year's Collider run which begins hind the big UA1 detector at the SPS the Antiproton Improvement Pro­ in September. CERN Courier, June 1984 189 .