Abstract
The end effects, or sheaths, of the plasma column of the Q-1 device are analyzed. The analysis utilizes an adaptation of the methods that have been used by Auer and McIntyre to predict the potential distributions in thermionic converters. Poisson's equation is solved under the assumptions of no collisions, no trapping, and a Maxwellian distribution of particles at the emitting boundaries and in the plasma. The model used requires the presence of a potential well in front of each plate. Although simpler than the actual situation with collisions and trapping, the model nevertheless seems to predict, qualitatively, the observed asymmetries in the volt-ampere characteristics, the observed density variations with current flow, and other observed variations with operating conditions. The analysis also indicates a possible explanation for certain of the observed instabilities.