Ion kinetic transport in the presence of collisions and electric field in TJ-II ECRH plasmas

Abstract

The ion collisional transport is estimated for the TJ-II device without the assumption of radially narrow particle trajectories, which is usually made in neoclassical transport studies. One million particles are followed in a realistic TJ-II magnetic configuration, taking into account collisions and the electric field. Global features of transport, not present in the customary neoclassical models, appear. The main results reached by considering the actual particle geometries are the monotonic increasing of heat and particle fluxes with minor radius, the non-diffusive character of transport, the appearance of asymmetries on the magnetic surfaces and the non-Maxwellian character of the distribution function.

Ion kinetic transport in TJ-II

“…I the typical size of the radial excursions with the relevant length scales of our plasma (which are calculated, included colhsions, as shown in the following section). The profiles, corresponding to a typical ECRH plasma, are taken from [9], and are qualitatively similar to those shown in Fig. 2.…”

“…Nevertheless, up-down asymmetries due mainly to the effect of the i?x5 drifts are observed experimentally [12], as well as toroidal asymmetries, due to the effect of trapped particles [9].…”

“…The electric field also contribute to the ion heating by communicating energy to the ions that drift outwards. The delay between the change in the ion temperature and that of the electron temperature is about 2 -3 ms, of the order of the time scale that the ions take to react to changes in the electric field [9]. The electron-ion collision time is 10^^ s -10^^ s. Since both characteristic times are much longer than the ion-ion collision time (10^'* -10^^ s), a linear Fokker-Planck calculation cannot properly reproduce any ion temperature increase caused by the electrons or the electric field.…”

“…The rest of coefficients follow from the usual guiding center approximation. Their explicit form is shown in [9] and [13].…”

“…The latter assumption is required by the collision operator The former is taken for computational reasons: a three-dimensional precise measurement of the ion density and temperature is out of reach for this NL calculation. Nevertheless, although non exact, both simplifications are reasonable: non-zero but small asymmetries and non-Maxwelhanities were measured in [9]. Therefore, in our scheme, /^^^'^ = /J^^\ means that we measure the radial density and average kinetic energy profiles of the plasma from the function y^l'j and construct, following Eq.…”