Author: Richard W King
Requested Type: Poster
Submitted: 2018-03-01 17:37:08
Georgia Institute of Technology
North Ave NW
Atlanta, GA 30332
We have developed a 2D multi-fluid neoclassical transport solver to predict rotation. We use Fourier techniques to capture flux surface variations and an arbitrary collisionality generalization of the Braginskii equations to model viscosities [1-2].
In previous works, a circular flux surface implementation of our extended viscosity predicted core rotation velocities within an order of magnitude [3-4]. The use of a Miller coordinate system fitted to GS flux surfaces markedly improved that fluid velocity calculation . That led to the creation of an iterative code for a first order poloidal Fourier decomposition, GTROTA . Later work resulted in an FS aligned local orthogonal frame for improvement of the poloidal asymmetry calculation .
In this work, we extend the applicability of this model to the plasma edge region, where the effects of asymmetries are more pronounced. To do this we have three main improvements: 1) we use a higher order Fourier expansion in the poloidal direction, 2) we allow asymmetries to be arbitrarily large, thereby retaining both the derivatives and the nonlinear terms, and 3) we take advantage of time dependence to obtain a realistic steady state.
The resulting nonlinear system of equations for carbon and deuterium rotation velocities is solved numerically.
1) Phys. Fluids 28 (1985) 2800. 2) Nucl. Fusion 25 (1985) 463. 3) Phys. Fluids B 5 (1993) 1828. 4) Phys. Plasmas 13 (2006) 062508. 5) Nucl. Fusion 53 (2013) 043011. 6) Comp. Phys. Comm 184(2013) 2571-2587. 7) Phys. Plasma 23 (2016) 052505.