Title: Our approach: Why PIC
1-0-
NSTX Research Forum, Dec. 5-7, 2006
XGC simulation of NSTX edge plasmas
C.S. Changa, S. Kua and the CPES teamb aCourant
Institute of Mathematical Sciences, NYU bUS DOE
Fusion Simulation Prototype Center for Plasma
Edge Simulation
2XGC code
-1-
- First principles simulation of the entire edge,
including the self-consistent interaction of the
turbulence-neoclassical-neutral dynamics. - Full-f ion, electron, neutrals
- Realistic B geometry with X-point wall (EFIT
g_eqdsk) - Allow plasma variation along the field line
- Heat flux from core and particle source from
neutral ionization (Monte Carlo) - 2006-2007 capabilities
- Macroscopic 2D plasma and potential/rotation
profiles in the edge after averaging out the
fluctuations (2006) - Electrostatic turbulence spectrum is under
analysis - Classical collision effects will be installed for
NSTX
3-3-
Averaging over the fluctuations
- Solutions over slowly varying initial
- profile
- Complete toroidal and Local
- poloidal averaging (lt10 cm)
- to allow plasma variation along B
- The first kinetic neoclassical
- simulation in the edge
4-4-
Gyrokinetic Particle Simulation of
Particle motions In the edge
5-5-
Strongly sheared neoclassical V and VExB in
the entire H-mode edge
(no neutrals, turbulence suppressed)
DIII-D, outside midplane
6Discussions
-8-
- XGC-0 and XGC-1 can analyze (and predict) edge
physics from first-principles kinetics - Neoclassical and neutral physics now
- Electrostatic turbulence next year (followed by
EM) - By validating the code using the experimentally
measurable quantities, we can predict the
un-measurable quantities and understand the
underlying physics.
7-2-
Experimental evidences of anisotropic
non-Maxwellian edge ions in DIII-D (K. Burrell,
APS 2003)
XGC-0 prediction of fi at outside-midplane
scrape-off
8ExB shearing becomes weaker toward L-mode
pedestal slope
-6-
?o
?n3 cm
?n1 cm
9XGC-0 simulation of pedestal buildup by neutral
ionization (B0 2.1T, Ti500 eV) 164K particles
on 1024 processors
Plasma density
VExB