NERSC Initiative for Scientific Exploration (NISE) 2011 Awards

Simulating ETG Plasma Turbulence with Impurities

David Mikkelsen, Princeton Plasma Physics Lab

Associated NERSC Project: Experimental Tests of Gyrokinetic Simulations of Microturbulence )m64)

Understanding plasma turbulence, particularly electron plasma turbulence, is a key step in the development of efficient nuclear fusion reactors. Nuclear fusion promises to be an environmentally friendly and plentiful energy source, but the performance of reactors is diminished as the plasma they contain becomes turbulent. One possible method to control plasma turbulence is through the magnetic configuration. Reversing its shear is known to experimentally reduce turbulence and trigger high plasma performance. If we can gain a theoretical understanding of this mechanism, we are one step closer to a new energy source.

Recent simulations of plasma turbulence in the National Spherical Torus Experiment demonstrate that magnetic shear can raise the nonlinear threshold for electron temperature gradient (ETG) driven turbulence. This effect may explain the observation of electron internal transport barriers in these plasmas, a regime of enhanced thermal confinement known to be triggered experimentally by reversing the magnetic shear in the plasma.

However, these simulations, while done with realistic majority ion plasma densities, do not take into account the relatively lower density of impurity ions. These could be important because the major impurity in NSTX is carbon, which, thanks to its high charge, significantly raises the effective Z value of the plasma. In the ‘adiabatic ion’ approximation raising Zeff is known to linearly reduce the drive of ETG turbulence, so including carbon should alter the linear threshold for plasma turbulent transport. The present nonlinear simulations include a kinetic treatment of ions because this is known to be important to achieve realistic turbulence saturation levels; the net impact of adding an impurity ion to the dynamics is unknown.

Previously, there have been no nonlinear simulations of ETG turbulence that include multiple kinetic ion species, as simulations that accurately account for even a single kinetic ion species are already expensive (~ 100,000 cpu hours). We need to add impurities in order to obtain realistic simulations of the effects of magnetic shear on nonlinear simulations of ETG turbulence in NSTX. To do so requires scanning electron temperature gradients with strongly reversed shear using electrons, deuterium and carbon at realistic density levels.