The longitudinal resistive microwave instability in storage rings has been identified in previous workshops as a key problem in heavy ion fusion. Here we show by means of computer simulation that the predictions from the linearized theory of instability are over-pessimistic in the case of non-relativistic ions. This is due to an early nonlinear saturation of the instability caused by the formation of a thin stabilizing tail of the momentum distribution, which provides enough Landau damping to suppress further growth. The main condition for this non-destructive self-stabilization lies in the slow resistive growth rate, as compared with the large coherent frequency shift induced by space charge. The important conclusion from this is a significant extension of the stable storage ring filling time to values that are comfortable from a design point of view. For the HIBALL reactor driver accelerator a safe storage time of 5 to 10 msec is expected, instead of only 1 msec according to the linearized theory e-folding times.

A heavy ion fusion (HIF) system, named HIBLIC (Heavy Ion Beam and LIthium Curtain) is conceptually designed. The driver system consists of RF linacs (RFQ linacs, IH linacs and Alvarez linacs), storage rings (one accumulator ring and three buncher rings) and beam transport lines with induction beam compressors. This accelerator complex provides 6 beams of 15 GeV208Pb1+ ions to be focused simultaneously on a target. Each beam carries 1·78 kA current with 25 ns pulse duration, i.e., the total incident energy on the target is 4 MJ, 160 TW per shot. Superconducting coils are used in most parts of the magnet system to reduce power consumption.

In 1982 to 83 a preliminary conceptual design of a heavy-ion-driven inertial fusion (HIF) system, called HIBLIC-I (Heavy Ion Beam and LIthium Curtain) was worked out (The Working Group on ‘HIBLIC-I’ 1984). Its principal aims are: (i) to find out the critical issues in this approach to fusion power plant, (ii) to assess its feasibility and (iii) to promote good collaboration between different fields (such as accelerator physics, plasma science and reactor technology) and of different institutions in Japan. (Obayashi 1981; Hirao 1982).

Our recent estimates are that the energy-gains for cryogenic single-shell targets are close to those of similar double-shelled targets. However, the peak-power requirements for single-shell targets are still about a factor of two higher.

Recently revised estimates of target gain have added additional optimistic inputs to the interface between targets, accelerators and fusion chamber beam transport. But it remains valid that neutralization of the beams in the fusion chamber is useful if ion charge state Z > 1 or if > 1 kA per beamlet is to be propagated. Some engineering and economic considerations favor higher currents.

Filamentary and jet-like structures are observed in the interaction of a laser beam with aluminum and gold targets. The range of irradiance is 1013 to 5 × 1014 W cm−2 with nanosecond pulses. According to theoretical predictions these structures can be attributed to thermal instabilities. We show that the electrothermal instability grows more probably at the 0.35 μm laser wavelength whilst the magnetothermal instability develops at 1·06 μm. The jet-like structures occurring on the rear side of thin targets are discussed.

The hitherto successful hydrodynamic plasma theory needed the simplifying assumption of quasi-neutrality. Earlier known ambipolar fields in plasma surfaces were considered as exceptions and Alfvén's model of a complementary description by plasma currents was criticized. Fields in plasmas were derived from the kinetic theory of turbulence. Following a model for the nonlinear force of laser–plasma interaction, we needed a general description of the plasma without space charge neutrality which succeeded numerically and analytically. High electric fields due to inhomogeneities inside plasmas were derived explaining for the first time quantitatively the reduction of thermal conduction in laser-fusion, the measured inverted double layers including a new type of resonance process, the MeV α-upshift by nonlinear-force driven caviton fields and the radial fields in tokamaks which cause plasma rotation.

The self consistent electrostatic field due to plasma inhomogeneities and associated with free expansion of a plasma into a vacuum is evaluated in the framework of two-fluid hydrodynamics and the corresponding generalized Rieman's self-similar solution found. The case of an adiabatic expansion is also discussed and the results obtained are applied to a laser produced plasma with a characteristic electron temperature of 1 keV and a characteristic time scale of 1 nsec.

The Proto-I experiment for a proof-of-principle experiment is described resulting in a beam divergence for ions of only 14mrad at 1·5 MV—420 KA operation and 6 KA/cm2 anode current density before focusing. This is a better beam divergence than required for Sandia National Laboratories' PBFA I and II installations, for which scaling is discussed. Ion source development is reported and the PBFA II experiment described for 30 MV operation with short pulses (10–15 nsec) of lithium ions. Pulses of 50 nsec width and 130 KJ energy have been achieved from the prototype module for PBFA II.