Finite Larmor radius effects on E x B weak turbulence transport

• Published in: Journal of plasma physics Vol. 84; no. 3
• Main Authors: Kryukov, N, Martinell, J J, del-Castillo-Negrete, D
• Format: Journal Article
• Language: English
• Published: Cambridge Cambridge University Press 01.06.2018
• Subjects: 70 PLASMA PHYSICS AND FUSION TECHNOLOGY; Amplitudes; Computational fluid dynamics; Diffusion coefficient; Diffusion rate; Distribution functions; Economic models; Larmor radius; Magnetic fields; Mathematical models; Normal distribution; Particle trajectories; plasma confinement; plasma nonlinear phenomena; Plasma physics; Shear flow; Simulation; Transport; Turbulence; Wave propagation; Mexico
• ISSN: 0022-3778; 1469-7807
• DOI: 10.1017/S0022377818000351

Transport of test particles in two-dimensional weak turbulence with waves propagating along the poloidal direction is studied using a reduced model. Finite Larmor radius (FLR) effects are included by gyroaveraging over one particle orbit. For low wave amplitudes the motion is mostly regular with particles trapped in the potential wells. As the amplitude increases the trajectories become chaotic and the Larmor radius modifies the orbits. For a thermal distribution of Finite Larmor radii the particle distribution function (PDF) is Gaussian for small [formula omitted: see PDF] (thermal gyroradius) but becomes non-Gaussian for large [formula omitted: see PDF] . However, the time scaling of transport is diffusive, as characterized by a linear dependence of the variance of the PDF with time. An explanation for this behaviour is presented that provides an expression for an effective diffusion coefficient and reproduces the numerical results for large wave amplitudes which implies generalized chaos. When a shear flow is added in the direction of wave propagation, a modified model is obtained that produces free-streaming particle trajectories in addition to trapped ones; these contribute to ballistic transport for low wave amplitude but produce super-ballistic transport in the chaotic regime. As in the previous case, the PDF is Gaussian for low [formula omitted: see PDF] becoming non-Gaussian as it increases. The perpendicular transport presents the same behaviour as in the case with no flow but the diffusion is faster in the presence of the flow.