| This body of work is new and comprises theoretical analysis, numerical simulation and experimental investigations of the vircator, a tunable, compact, simply constructed, high power reflexing electron device that is used as a microwave source. A 1D theoretical model is formulated that is based on a time-varying, nonlinear potential to better understand the sensitivities of individual electron trajectories to macroscopic parameters of the system. Self-consistent 2D (space) and 3D(velocity) numerical PIC simulations are used to find common dynamical behavior that links the individual test-particle trajectories examined in the 1D nonlinear model with the experimental measurements of the final state of the electron trajectories.;The experiment designed and built for these studies is unique and contrasts with other virtual cathode electron beam sources because it operates at lower voltage, and longer pulse width than previous forms of the device. The anode-cathode (AK) gap spacing is externally tunable expediting parametric studies. In addition, it is repetitively pulsed, which permits discrimination between single-shot anomalies and repeatable variation that often occurs in pulsed power experimental devices. These attributes distinguish this experimental apparatus from previous experimental platforms. This novel, repetitively pulsed device has particular relevance to the industrial and medical applications of rf.;The approach followed in the present studies is (1) to investigate the dynamic behavior of individual test-particles in a simplified 1D, nonlinear, time-varying, potential; (2) utilize a 2D, self-consistent, electromagnetic PIC code to generate single particle trajectories subject to the collective effects of the electron beam; and (3) measure macroscopic quantities of voltage, current, electromagnetic fields, and electron flux in an experimental platform. Results of the 3D experiments are compared to predictions of the 1D and 2D models.;Electron flux calculated in the 1D nonlinear potential model and the 2D PIC simulations agree well with experimental measurements. Classes of orbit trajectories integrated in the 1D model are common to those calculated in the 2D PIC code. Comparison of frequency content calculated and measured also show favorable overlap. Experimental observations are supported by analytic and numerical modelling.;Experimental pulsewidth limitations do not permit a data stream (time-series) sufficient to make a clear estimate of the fractal dimension of the system, though the sensitivity of particle trajectory to initial conditions is clear. The nonlinear equation contains parameter regimes that generate chaotic solutions, however the parameter regime in which the device operates is not chaotic, merely sensitive to initial conditions.;Random initial conditions that occur as a result of beam thermalization and nonuniform electron-emission at the surface of the cathode are present in the experiment. These characteristics alone do not explain the experimentally observed fluctuations in power and frequency. The time-varying nonlinear potential exhibits classes of particle trajectories that follow trends in the experimental results, fluctuations in particle asymptotic states, and particle motion sensitive to the shape of the virtual cathode. (Abstract shortened by UMI.)... |
