A technique is presented for measuring the parallel energy distribution of magnetically confined electrons in a cylindrically symmetric pure electron plasma. In essence, the technique measures how many electrons are energetic enough to escape past applied confinement potentials. The technique does not require any secondary magnetic fields. Simplified variations of the technique are also presented which can be used at the expense of some loss of information. These techniques have been successfully used in three experimental contexts.

Radial transport produced by static nonaxisymmetric fields is thought to limit the confinement of non-neutral plasmas and experiments with applied asymmetries have verified that such fields do produce transport. A theoretical model of such transport is presented which is appropriate for long, thin plasmas. The theory allows for asymmetries with nonzero frequency and includes the linear collective response to applied wall voltages. For the regime where the effective collision frequency is large, the asymmetry-induced radial particle flux is derived from the drift kinetic/Poisson equations including collisions. For low collision frequencies a heuristic derivation is given. In both regimes the resulting transport is dominated by particles that move in resonance with the asymmetry. Possible applications of the theory to several experiments are discussed.

Experimental evidence is presented showing that collective effects may enhance the radial transport produced by an asymmetric perturbing field applied to a rotating nonneutral plasma column. The field asymmetry drives an asymmetric plasma mode and the mode field produces additional transport. Of particular interest for confinement systems is the existence of a zero-frequency plasma mode which can be driven by static field asymmetries. Applications to tandem mirrors are discussed.PACS numbers: 52.25.Fi, 52.55.Mg

A key prediction of the theory of asymmetry-induced transport is that the particle flux will be dominated by particles that move in resonance with the asymmetry. For the case of a time-varying asymmetry, the resonance condition is ω−lωR−nπv/L=0, where v is the axial velocity, L is the plasma length, ωR is the E×B rotation frequency, and ω, l, and n are the asymmetry frequency, azimuthal wavenumber, and axial wavenumber, respectively. Data are presented from experiments on a low density trap in which ω, ωR, and n are varied and the resulting radial particle flux is measured. The experiments show a resonance in the flux similar to that predicted by theory. The peak frequency of this resonance increases with ωR and varies with n, in qualitative agreement with theory, but quantitative comparisons between experiment and theory show serious discrepancies.

Thomson radar returns from the ionosphere subjected to a step-function excitation of electromagnetic waves show the immediate rise of electrostatic waves near the resonant height oo p ^ co 0 . The observed secular temporal variation E <* t agrees with a theorybased on direct conversion of electromagnetic waves to electrostatic waves by preexisting density inhomogeneities. The rise time is at least an order of magnitude smaller than previously reported. The frequency spectrum shows this conversion is characterized by frequencies at the pump frequency co 0 .PACS 52.25.Ps, 52.35.Fp, 52.35.Hr There have been several ionosphere-modification experiments 1 in which an electrostatic wave spectrum consistent with parametric decay instabilities 2 has been observed. This process has been well documented in the laboratory 3 and consists of the decay of the incident electromagnetic (EM) wave of frequency co 0 and wave number 5 0 into electrostatic (ES) modes. The decay modes have short wavelength (compared to 27r/|5 0 l) and correspond to a Langmuir (o; z ,E z ) and an ion acoustic (wi^i) wave. The decay process has the following signatures: (1) matching conditions (x) 0 = oo l + a? i? ^0 = ^! +5^; (2) exponential growth in time, i.e., exp(yt); and (3) a power threshold must be exceeded, i.e., it depends nonlinearly on EM amplitude. In this Letter we report the observation of a new process which also gives rise to the excitation of ES modes in the ionosphere. The underlying physics differs from the parametric process and its existence has important consequences for the interpretation and planning of future ionosphere-modification experiments. In addition, the new process reported can be used as a diagnostic tool that samples the properties of preexisting density inhomogeneities in the ionosphere, as well as the local amplitude of the EM wave.The essential physics behind the process investigated, which we refer to as "direct conversion," consists of the resonant pumping of a Langmuir wave by the electric field of the EM wave. The resonant excitation occurs below the ES cutoff, i.e., at a height such that QJ P 0 ; and (4) the amplitude scales linearly with the EM field amplitude and does not exhibit a threshold. These properties are si...

A non-neutral plasma trap has been constructed in which the radial electric field of a non-neutral plasma column is simulated by a biased wire stretched along the axis of the device. The confinement time of test electrons in this device is found to be comparable in magnitude and scaling with that found in pure electron plasma experiments, in spite of the fact that the test electron density is 10 2 times smaller than in a typical pure electron plasma. The confinement time is only weakly dependent on the central wire bias. These results may provide useful input to theoretical efforts to explain transport in these traps.

No abstract

Initial experiments on asymmetry-induced transport in the Occidental non-neutral plasma trap found the radial particle flux at small radii to be proportional to φa2, where φa is the applied asymmetry amplitude. Other researchers, however, using the global expansion rate as a measure of the transport, have observed a φa1 scaling when the rigidity (the ratio of the axial bounce frequency to the azimuthal rotation frequency) is in the range of 1–10. In an effort to resolve this discrepancy, measurements have been extended to different radii and asymmetry frequencies. Although the results to date are generally in agreement with those previously reported (φa2 scaling at low asymmetry amplitudes falling off to a weaker scaling at higher amplitudes), some cases have been observed where the low amplitude scaling is closer to φa1. Both the φa2 and φa1 cases, however, have rigidities less than 10. Instead, the φa1 cases are characterized by an induced flux that is comparable in magnitude but opposite in sign to the background flux. This suggests that the mixing of applied and background asymmetries plays an important role in determining the amplitude scaling of this transport.