Scoping calculations have been performed examining the consequences of fuel melting and pin failures for a reactivity-insertion type accident in a sodium-cooled, pool-type reactor fueled with a metal alloy fuel. The principal gas and vapor species released are shown to be Xe, Cs,and bond sodium contained within the fuel porosity. Fuel vapor pressure is insignificant, and there is no energetic fuel-coolant interaction for the conditions considered. Condensation of sodium vapor as it expands into the upper sodium pool in a jet mixing regime may occur as rapidly as the vapor emerges from the disrupted core (although reactor-material experiments are needed to confirm these high condensation rates). If the predictions of rapid direct-contact condensation can be verified experimentally for the sodium system, the implication is that the ability of vapor expansion to perform appreciable work on the system is largely eliminated. Furthermore, the ability of an expanding vapor bubble to transport fuel and fission product species to the cover gas region where they may be released to the containment is also largely eliminated. The radionuclide species except for fission gas are largely retained within the core and sodium pool.

None

Experiments at LLNL using the 10 TW Novette laser have led to significantly increased understanding of laser/plasma coupling. Tests using 1.06 ..mu..m, 0.53 ..mu..m and 0.26 ..mu..m light have shown increased light absorption, increased efficiency of conversion to x-rays, and decreased production of suprathermal electrons as the wavelength of the incident light decreases. The data indicate that stimulated Raman scattering is the source of the excessive hot electrons and that the effect can be controlled by the proper selection of laser frequency and target material. The control of these effects has led to achievement of higher inertial fusion target compressions and to production of the first laboratory x-ray laser.

Analysis of a single granule on a rotating cone shows that for the 35/sup 0/ half-angle, double-cone-shaped Cascade chamber, blanket granules will stay against the chamber wall if the rotational speed is 50 rpm or greater. The granules move axially down the wall with a slight (5-mm or less) sinusoidal oscillation in the circumferential direction. Granule chute-flow experiments confirm that two-layered flow can be obtained when the chute is inclined slightly above the granular material angle of repose. The top surface layer is thin and fast moving (supercritical flow). A thick bottom layer moves more slowly (subcritical flow controlled at the exit) with a velocity that increases with distance from the bottom of the chute. This is a desirable velocity profile because in the Cascade chamber about one-third of the fusion energy is deposited in the form of x rays and fusion-fuel-pellet debris in the top surface (inner-radius) layer.

Cascade, originally conceived as a football-shaped, steel-walled reactor containing a Li/sub 2/O granule blanket, is now envisaged as a double-cone-shaped reactor containing a two-layered (three-zone) flowing blanket of BeO and LiAlO/sub 2/ granules. Average blanket exit temperature is 1670/sup 0/K and gross plant efficiency (net thermal conversion efficiency) using a Brayton cycle is 55%. The reactor has a low-activation SiC-tiled wall. It rotates at 50 rpm, and the granules are transported to the top of the heat exchanger using their peripheral speed; no conveyors or lifts are required. The granules return to the reactor by gravity. After considerable analysis and experimentation, we continue to regard Cascade as a promising reactor concept with the advantages of safety, efficiency, and low activation.

Pulse Star is a pool-type ICF reactor that emphasizes low cost and high safety levels. The reactor consists of a vacuum chamber (belljar) submerged in a compact liquid metal (Li/sub 17/Pb/sub 83/ or lithium) pool which also contains the heat exchangers and liquid metal pumps. The shielding efficiency of the liquid metal pool is high enough to allow hands-on maintenance of (removed) pumps and heat exchangers. Liquid metal is allowed to spray through the 5.5 m radius belljar at a controlled rate, but is prohibited from the target region by a 4 m radius mesh first wall. The wetted first wall absorbs the fusion x-rays and debris while the spray region absorbs the fusion neutrons. The mesh allows vaporized liquid metal to blow through to the spray region where it can quickly cool and condense. Preliminary calculations show that a 2 m thick first wall could handle the mechanical (support, buckling, and x-ray-induced hoop) loads. Wetting and gas flow issues are in an initial investigation stage.