A complex, and apparently stochastic, character frequently can be seen to occur in the solutions to simple Hamiltonian problems. Such behavior is of interest, and potentially of importance, to designers of particle accelerators--as well as to workers in other fields of physics and related disciplines. Even a slow development of disorder in the motion of particles in a circular accelerator or storage ring could be troublesome, because a practical design requires the beam particles to remain confined in an orderly manner within a narrow beam tube for literally tens of billions of revolutions. The material presented is primarily the result of computer calculations made to investigate the occurrence of ''stochasticity,'' and is organized in a manner similar to that adopted for presentation at a 1974 accelerator conference.

A laboratory-scale experimental device having an outside diameter of 1.2 m has been designed to test the idea of electrostatic end plugging of an open-ended magnetic-field configuration. The configuration is a toroidal quadrupole having four very thin (less than 1-mm-thick) line cusps produced by four circular copper coils. Iron is used to concentrate the magnetic flux density to 2.0 T; without the use of iron, the power consumption, which is about 1 MW, would be about 25 times higher. The use of iron also produces a precisely known magnetic field and allows good access for diagnostics and pumping. Iron is also used for both the flux return path and the vacuum chamber. A hollow anode with an adjustable (nominally 1-mm-wide) gap is biased from 10 to 20 kV. Plasma densities of about 10/sup 13/ cm/sup -3/ and temperatures of about 1 keV might be produced by an electron beam and by electron cyclotron resonance heating. Higher-order multipoles (hexapoles and octopoles) also are described.

Two novel techniques which produce a uniform frozen layer of deuterium-tritium on the inside surface of a glass microsphere are presented.

The Tandem Mirror Experiment (TMX) system, part of the Lawrence Livermore Laboratory magnetic mirror program incorporates in its design various types of coils or magnets. This paper describes the physical construction of each coil within the system as well as the structural design required for their support and installation.

The Andersson-Braun remmeter was modified to improve the directional dependence of the instrument to thermal and fast neutrons and increase its sensitivity to thermal neutrons. The modifications consist of partially rounding the end of the instrument, moving the BF/sub 3/ tube forward by /sup 1///sub 2/ in., increasing the size of the holes in the boron-loaded polyethylene sleeve, replacing one of the boron-loaded end plugs with a polyethylene plug, changing the location of the other end plug, and adding a small disk of cadmium over the hole where the BF/sub 3/ tube enters the moderator. The cost of making these modifications to the Lawrence Livermore Laboratory Andersson-Braun remmeters is approximately $50.00 each. These modifications to the instrument increase the thermal neutron sensitivity by a factor of approximately 3 when exposed at the side and a factor of approximately 9 at the end opposite the instrument packet. The high thermal sensitivity at the instrument packet end (a factor of approximately 17) is eliminated, and the directional dependence to both thermal and fast neutrons is reduced to +-10% (except at the instrument packet end). The sensitivity of the instrument to fast neutrons is increased by approximately 15%. The sensitivity of the modified …

The geometry and size of the superconducting coils for the Mirror Fusion Test Facility (MFTF) proposed by Lawrence Livermore Laboratory (LLL) impose certain constraints on the Nb-Ti superconductor. The most promising fabrication process is a wrap-around technique in which a superconducting core is ''wrapped'' in stabilizing copper that contains built-in cooling channels. Insulation between pancake coils and turns is provided by perforated sheets and buttons of epoxy-impregnated fiberglass. Preliminary heat-transfer tests conducted on short samples of single conductor and on a nine-conductor bundle are reported and related to the heat generated in ''normal'' conductors. Investigation of joining techniques, necessary because of the length of conductor needed for the MFTF magnet (about 21 km per coil), show that cold-welded butt joints best meet all requirements. In a test coil now being built, approximately 2 km of prototype MFTF conductor will provide a self-field of about 4 T. Supplementary coils will boost the field to about 6.7 T. The test coils will be used to study cryostatic stability, the propagation and recovery of normal zones, and diagnostic techniques.

The parametric design study of the tandem mirror reactor (TMR) is described. The results of this study illustrate the variation of reactor characteristics with changes in the independent design parameters, reveal the set of design parameters which minimizes the cost of the reactor, and show the sensitivity of the optimized design to physics and technological uncertainties. The total direct capital cost of an optimized 1000 MWe TMR is estimated to be $1300/kWe. The direct capital cost of a 2000 MWe plant is less than $1000/kWe.

None

None

Neutral-beam injection is a primary means of building high-energy plasmas in present mirror fusion machines. The injectors need periodic maintenance while a machine is operating, so isolation of them by valving is a desirable chracteristic. Because their energy densities have practical limits, the beams require large cross-sections. Thus, valves having apertures of 10 ft/sup 2/ would be common. Traditional cam-seated vacuum gate valves have been built this large, but they are bulky, costly, and heavy. Improvement is possible by use of an inflatable-bellows assembly for the valve gate. This paper describes the design of such a valve. The design allows a clear aperture of 20 by 36 in. with a valve body that is only 5-/sup 1///sub 4/ in. thick.