The standard paraxial ray equation method for the design of electrodes for an electrostatically focused gun is extended to include relativistic effects and the effects of the beam's azimuthal magnetic field. Solutions for parallel and converging beams are obtained and the predicted currents are compared against those measured on the High Brightness Test Stand. 4 refs., 2 figs.

Several techniques using small radius collimating pipes with and without axial magnetic fields to measure the brightness of an extracted 1 - 2 kA, 1 - 1.5 MeV electron beam will be described. The output beam of the High Brightness Test Stand as measured by one of these techniques is in excess of 2 x 10/sup 5/ amp/cm/sup 2//steradian. 5 refs., 4 figs.

The total emission rate is (4.5 +- 0.4) 10/sup 7/ n/s, and the average neutron energy is (1.64 +- 0.07) MeV. The factor for converting from neutron fluence to dose equivalent for this spectrum is (3.10 +- 0.24) 10/sup -5/ mRem/n-cm/sup -2/. The factor for converting from neutron fluence to tissue absorbed dose is (3.18 +- 0.26) 10/sup -6/ mRad/n-cm/sup -2/.

Pump neutral beams, which are directed into the loss cone of the TMX-U plugs, are normally used to pump ions from the thermal barriers. Because these neutral beams introduce cold gas that reduces pumping efficiency, and require a straight line entrance and exit from the plug, alternate methods are being investigated to provide barrier pumping. To maintain the thermal barrier, either of two classes of particles can be pumped. First, the collisionally trapped ions can be pumped directly. In this case, the most promising selection criterion is the azimuthal drift frequency. Second, the excess sloshing-ion density can be removed, allowing the use of increased sloshing-beam density to pump the trapped ions. The selection mechanism in this case is the Doppler-shifted ion-cyclotron resonance of the high-energy sloshing-ions (3 keV less than or equal to U/sub parallel/ less than or equal to 10 keV).

Accelerator injector designs have been evaluated using two computer codes. The first code self consistently follows relativistic particles in two dimensions. Fields are obtained in the Darwin model which includes inductive effects. This code is used to study cathode emission and acceleration to full injector voltage. The second code transports a fixed segment of a beam along the remainder of the beam line. Using these two codes the effects of electrode configuration on emittance, beam quality and beam transport have been studied.