A Scheme for E×B-Drift Injection into a Tabletop Stellerator
POSTER
Abstract
The EPOS (Electrons and Positrons in an Optimized Stellerator) Project aims to confine a pair plasma magnetically. To this end, trajectory simulations were used to design a system of electrodes that will allow for e+ injection. Low-energy positrons from an external source are guided up to the device on “stray” field lines and encounter electrostatic potentials that transport them into the confinement region. We give an overview of the E×B drift injection plates' preliminary design.
Plates' dimensions and biases were scanned using ELWIS, an in-house particle simulation software. A sensitivity analysis was conducted using initial velocity distributions as measured for prior e+ injection experiments into dipole traps.
The configuration consists of two rectangular plates (16.5 cm x 9 cm, 1-mm thick) biased at ±550 V, with 2-cm spacing, coupled with two rebound plates (15.5 cm x 2 cm, 1-mm thick), both at +10V. This results in successful bunch injection for a beam with E∥ = 7.3eV, (σ = 0.02eV), E⊥ = 0.11eV (σ = 0.01eV), a 1-mm radius, and a FWHM of 0.4 mm. Elastic scattering is avoided for a vacuum pressure of 10e−9 mbar. The range of allowable perpendicular energy spread is σ⊥ ∈ [0, 0.05] eV to prevent mirroring at the exit of the injection plates and enable successful e+ injection past the plasma’s last closed flux surface.
A more avant-garde design was also explored, in which the shape of two non-planar plates was optimized to produce near-perpendicular E and B fields using cubic Bézier surfaces. The plate geometry was optimized by minimizing the sum of the local electric and magnetic field alignment, quantified via the dot product E·B evaluated at each measurement point. This yielded the control points defining the two spline surfaces.
These results will be used for a preliminary e+ injection system and will inform future work on E × B plates’ optimization.
Plates' dimensions and biases were scanned using ELWIS, an in-house particle simulation software. A sensitivity analysis was conducted using initial velocity distributions as measured for prior e+ injection experiments into dipole traps.
The configuration consists of two rectangular plates (16.5 cm x 9 cm, 1-mm thick) biased at ±550 V, with 2-cm spacing, coupled with two rebound plates (15.5 cm x 2 cm, 1-mm thick), both at +10V. This results in successful bunch injection for a beam with E∥ = 7.3eV, (σ = 0.02eV), E⊥ = 0.11eV (σ = 0.01eV), a 1-mm radius, and a FWHM of 0.4 mm. Elastic scattering is avoided for a vacuum pressure of 10e−9 mbar. The range of allowable perpendicular energy spread is σ⊥ ∈ [0, 0.05] eV to prevent mirroring at the exit of the injection plates and enable successful e+ injection past the plasma’s last closed flux surface.
A more avant-garde design was also explored, in which the shape of two non-planar plates was optimized to produce near-perpendicular E and B fields using cubic Bézier surfaces. The plate geometry was optimized by minimizing the sum of the local electric and magnetic field alignment, quantified via the dot product E·B evaluated at each measurement point. This yielded the control points defining the two spline surfaces.
These results will be used for a preliminary e+ injection system and will inform future work on E × B plates’ optimization.
Presenters
-
Grégoire-Hubert Ducas
Authors
-
Grégoire-Hubert Ducas
-
Paul Huslage
Max-Planck Institute for Plasma Physics, Max Planck Institute for Plasma Physics
-
Pedro F Gil
Max Planck Institute for Plasma Physics
-
Stefan Nissl
Max Planck Institute for Plasma Physics
-
Adam Deller
Max Planck Institute of Plasma Physics, Max Planck Institute for Plasma Physics
-
Tristan Schuler
Max Planck Institute for Plasma Physics
-
Jason Smoniewski
Max Planck Institute for Plasma Physics
-
Eve V Stenson
Max Planck Institute for Plasma Physics