H (sub) alpha & Neutral Density Scaling in the Maryland Centrifugal eXperiment

Abstract

The Maryland Centrifugal eXperiment (MCX) is a hydrogen plasma confinement experiment with a rotating mirror magnetic configuration. This experiment was designed to test the concepts of centrifugal confinement and velocity shear stabilization which may allow scaleability to a fusion reactor. These two concepts, however, rely on supersonic plasma fluid velocities, which, apart from possible plasma instabilities, could be greatly reduced by fluid drag with neutral hydrogen, leading to decreased confinement. Resonant charge exchange between a hydrogen ion and a hydrogen atom is believed to be the dominant drag mechanism on the rotating plasma. Neutral hydrogen emission lines (particularly the Balmer-alpha line, H (sub) alpha) are therefore of primary interest in diagnosing how neutral hydrogen affects plasma confinement. For this purpose, a multi-chord H (sub) alpha emission detector (multi-chord HED) was designed and constructed by the author in order to measure emissivity profiles. These profiles, together with an atomic collisional-radiative model, provide estimates of neutral hydrogen density and local charge-exchange times. Varied experimental parameters were applied to MCX discharges and the resulting variations in neutral density are compared to theoretical scaling laws. The charge-exchange times are compared to the measured momentum confinement time. We find that the inner and outer-most flux surfaces are not distinctly identified by the emissivity profile and the emissivity is dominant at the vacuum chamber wall. We also find that, while the overall emissivity profile does not match theoretical prediction, neutral density scaling is approximately described by the models. In addition, charge-exchange times are found to be much smaller than the momentum confinement time as well as to scale differently than the momentum confinement time.

This dissertation includes a detailed description of the multi-chord HED system and its calibration, both spectrally and absolutely. We also present models based on neutral and plasma interaction which provide the scaling laws used to compare to experimental results.