FIG. 2.1. Experimental Facility.

Recent Accomplishments

• Finished construction of Test Cell
• Design and fabrication of a wall-embedded probe to measure current flux loss directly to the conducting surface within the cusp region
• Obtained data for electron flux profiles upstream of magnetic cusp as varying background pressures
• Data taken and roughly analyzed for configuration with Helmholtz Coil and single magnetic cusp

Description

The proposed experiment will investigate the physical processes of plasma confinement within a magnetic cusp. These processes will be methodically examined by utilizing a simple and well-defined experimental domain or "test cell" that can be readily compared with results from the computational efforts. Using this test cell, we will start with simple, low energy electron plasma and then methodically increase the complexity of the plasma to include the effects of all plasma species: electrons, neutral gas, and ions. Throughout these plasma regimes the focus will be to investigate the confinement behavior by examining features such as the near-wall loss area for a magnetic cusp. Simple magnetic fields (using just one or two magnets) will be used to provide a fundamental understanding that can be extended computationally and analytically to multi-magnet plasma confinement geometries and micro plasma discharges.

Setup

FIG. 2.2. "Test Cell" for measuring charged particle fluxes and neutral particle density in the presence of a permanent magnet cusp.

FIG. 2.3. EGA-1012 Electron Gun Assembly.

FIG. 2.4. Wall-embedded probe mounted at downstream end of the test cell.

FIG. 2.5. Prototype miniature discharge under operating conditions with a tungsten filament. This discharge is also being used with hollow cathodes.

The vacuum chamber to be used for this effort, as shown in Fig. 2.1, has an internal length of 1.5 m and a diameter of 0.6 m. This chamber is cryogenically pumped by two 10" CTI Cryo-Torr pumps with a combine pumping speed of 6000 l/s and has a base pressure of 5 x 10-8 Torr. Internal to the chamber is a three-axis computer-controlled translation stage with a multi-function mount for diagnostics.

The experimental apparatus shown in Fig. 2.2 is a well-defined "test cell" that contains precision-machined surfaces that measure the flux of charged particles to the boundaries of the experimental domain. A constant axial field can be superimposed using a Helmholtz coil to guide the electron particles to the cusp so that they do not terminate prematurely at the sidewalls. The permanent magnet is placed just behind a "cusp surface" plate that serves as the anode surface of the test cell discharge. Surface measurement probes are integrated with the back wall to obtain an accurate measurement of the loss area. The entire assembly is mounted onto the vacuum chamber wall inline with the electron gun for proper alignment. The recently introduced EGA-1012 electron flood gun shown in Fig. 2.3 provides ideal conditions for this effort. The gun can provide high currents (up to 2 mA) and low energies electrons (5 - 1000 eV) with minimal thermal spread (0.4 eV).

The initial surface measurement probe (wall-probe) shown in Fig. 2.4 has a collection diameter of 0.24 mm. The orifice contains a knife edge to reduce any shape factor of a finite thickness aperture. Instead of the Samarium Cobalt magnets commonly used in plasma discharge, weaker Alnico magnets are used for the initial phase of the effort studying the interactions between electrons and neutral particles. Stronger magnets will later be used once development of the MEMs surface probe is completed.

In addition, a prototype 3 cm miniature discharge as shown in Fig. 2.5 was built to isolate the plasma discharge chamber from a normal ion thruster configuration. This allows the isolation of ring cusp magnet confinement that can be examined by precision diagnostic. This scale experiment is being used to obtain preliminary results for confinement behavior in a small scale discharge to help reveal the important macroscopic design considerations when reducing the discharge volume.

Preliminary Results

FIG. 2.6. Current flux profiles to wall probe at various axial positions and test cell pressures. The flux measured is reduced with increasing pressure, indicating the impedance of parallel particle motion due to elastic collisions.

FIG. 2.7. Surface profile of wall-probe current at increasing axial distance.

FIG. 2.8. Loss area simulation using particle tracker. (left) Injection of particle offset by 1 mm radial distance. (right) Helmholtz Coil offset by 1 degree.

FIG. 2.9. Plasma density and potential for miniature discharge.

Figure 2.6 is a comparison of the cusp profile at increasing axial distance from the magnet and at various test cell pressures. The results are expected due to the aspect ratio of the test cell domain. Although approximately 60% of the electrons undergo at least one elastic collision, a majority of them occurs well before the electrons are actually confined to the cusp field. The scattered electrons are then more likely to be reflected by the transverse magnetic fields toward the sidewalls where they are lost. Therefore, the results show a greater change in the magnitude of the flux profile rather than the shape. To mitigate this effect, a Helmholtz coil was to be superimposed to confine the scattered electrons

Figure 2.7 shows current density surface scan of a single magnet with an axial Helmholtz coil field. Prior computational simulation showed that the Helmholtz field would guide the electrons into the cusp and greatly improve the density inside the cusp. However, the simulated cases were of an ideal condition of perfect alignment that is nearly impossible to imitate with the experimental setup. Figure 2-8 demonstrates how the loss area can drastically by altered by even the slightest of misalignments. Without the Helmholtz coils, electrons only penetrate to certain field lines which they then follow to the cusp at the center of the magnet. However, the Helmholtz field confines the electrons from the beginning and leads the guiding centers of the electrons off axis in a manner that is difficult to predict.

Figure 2.9 shows results from Langmuir traces inside the prototype miniature discharge that is discussed above. These figures show a weak (left) and strong (right) magnetic field. Results do indeed show an inversion of the plasma potential due to the pervasive magnetic field in the small domain. The weaker field shows better plasma uniformity suggesting a trade-off between plasma containment (strength of magnetic field) and desirable plasma performance (uniformity), thus leading to a optimization challenge for developing a high performance miniature discharge. This emphasizes the importance of understanding the single magnet plasma loss region in miniature discharges so that these results can be extend to multi-magnet confinement optimization.

Future Work

Various experimental measurements will continue to be taken using the current test cell to obtain more intricate data sets and allow for validation of the collision models for the computational effort. Designs are also being considered to allow for higher particle densities of all species by either modifying the current test cell or developing a completely new one. A miniature hollow cathode will replace the currently used electron gun, allowing for higher discharge currents and test cell pressure on par with common plasma discharges. This is required to generate adequate population of ions to create intermediately ionized plasma and allow for analysis of multispecies behavior in the presence of a magnetic cusp very near an anode wall.