In the tandem mirror GAMMA 10, confinement potentials are generated by ECR heating at both ends of the device. The generated potential is measured with Gold Neutral Beam Probes (GNBP) using the gold neutral particle as the primary beam installed in the central cell and the barrier cell. We measured the radial and two-dimensional profiles of the potential by the GNBPs[1] [2], and also measure fluctuations with low and high frequencies [3]. However, the beam probe methods always have the problem of the path-integrated signals (Path Integral Effects) [4]. The estimation of the effects is important for the measurement using beam probes.
An improved electrostatic analyzer for the GNBP system was newly installed in the central cell. This analyzer has two Micro Channel Plates (MCP) detectors; one is for the measurement of the potential, the other is for the measurement of the beam deflection which appears along the z axis by the change of the magnetic field. Because of using the neutral particle as the primary beam, only the secondary beam is influenced by the change of the magnetic field. This is an important point different from the Heavy Ion Beam Probe (HIBP) method. Therefore, the Path Integral Effects for the measurement of the magnetic field fluctuation is revealed only in the secondary beam. We simulated the effects for the magnetic field fluctuation measurement.
At first, the beam velocities of the radial and azimuthal components are examined. When we considered the beam whose injection angle was swept by the electrostatic deflector, the velocity of the beam passing the plasma center was dominant in radial, on the other hand, the velocity of the beam passing the plasma edge was dominant in azimuthal. This means path-integrated ΔB_θ and ΔB_r are measured respectively. But, when we considered the energy-swept beam, the radial velocity of the beam was dominant within the almost energies.
Next, we simulated the Path Integral Effects in the plasma with magnetic field fluctuation. In this simulation, we assumed the fluctuation as standing waves, and estimated frequency and intensity dependency of the Path Integral Effects. We found ΔB_z component could be neglected when the deflection was along the z axis. The beam passing the plasma center was deflected by 2 or 0.2 [mm] when the peak of ΔB_θ or ΔB_r were assumed about 0.01[T] respectively. Therefore the beam dominant in the radial velocity is almost influenced by ΔB_θ. Estimating dependency of the energy-swept beam with some intensity profile, we obtained the Path Integral Effects of the magnetic field fluctuation measurement using a GNBP by beam simulation.

References