The next step on the development path to a safe and economically attractive source of fusion energy
is construction of the ITER device. The power to drive the ITER plasma to the burning regime will be
supplied by external heating systems in addition to the induced plasma current that also creates the
discharge magnetic equilibrium. The external systems are either driven by radio frequency (RF) power or
by the injection of energetic ions from neutral beam injection (NBI). The RF sources include waves in
the ion cyclotron range of frequencies (ICRF), the lower-hybrid range of frequencies (LHRF), and the
electron cyclotron range with frequencies from less than 100 MHz, to a few GHz, to over 100 GHz.
These external sources of power will be used for heating of the bulk plasma as well as localized control of
the current density and pressure profiles in order to access regimes of improved plasma energy
confinement. In this proposal we plan to further develop, verify, and validate a sophisticated simulation
capability for ICRF and LHRF power sources taking advantage of continuing advances in massively
parallel computing architectures. This capability will be required for the planned Fusion Simulation
Project (FSP) and the proposed project will attempt to meet FSP needs while attaining our proposed
research goals. This capability will integrate the RF physics of the core and edge plasma.
We have developed algorithms that self-consistently couple the core full wave spectral solvers with
quasilinear diffusion codes in the ICRF regime and have validated this model against experiment using
synthetic diagnostics that detect energetic ions. We have begun to include finite orbit width effects in
these combined full-wave Fokker-Planck codes and those simulations now routinely include three—
dimensional (3D) electric field effects. We have also developed this capability in the LHRF regime and
have now started to validate that model using synthetic diagnostics for hard x-ray emission. An
impedance matrix approach has been devised that will form the interface between our existing core
spectral wave solvers and the detailed 3D edge RF simulation modules that we propose to develop to
describe the edge. We have succeeded in extending the solution domain for one of the core spectral
solvers past the last closed flux surface to the vacuum vessel wall and have begun to explore finite
element method approaches for the regions between the antenna and the last closed flux surface that
include models for nonlinear sheath boundary conditions or adaptive meshes that allow concentration of
grid points in regions where higher spatial resolution is likely to be required, such as sharp edges near
structures where high intensity fields may develop.
Though significant progress has been made in understanding wave-plasma interactions in the good
confinement regions of tokamaks and Spherical Torus (ST) devices, recent comparisons of our core ICRF
codes against experiments in which high harmonic fast waves interact with fast NB ions indicate that
additional nonthermal ion effects must be included in order to accurately model the experimental
observations. These RF-induced finite orbit effects on the evolution of the fast particle distributions in
both configuration and velocity space need to be resolved before self-consistent simulations of fast
particle instabilities in burning plasma devices such as ITER can be utilized with any degree of
confidence. Furthermore, the simulation codes are not yet able to predict the absolute amount of power
that can be coupled into the plasma from a given launcher for various waves. Experimental observations
indicate that a number of different linear and nonlinear interactions contribute to parasitic RF power loss
in this region, including:
- Localized “hot spot” formation on the divertor plates, interactions with edge localized modes
(ELMs) and excitation of parametric decay instabilities (PDI) in NSTX.
- Edge interactions of LH waves in Alcator C-Mod and concomitant loss of core suprathermal
- Power coupling difficulties during H-modes on DIII-D.
- Tile erosion and possible sheath effects on Alcator C-Mod.
Because of the widely varying scale lengths that are needed to represent the fields in this region, it is
only with the advent of tera-scale to peta-scale computing platforms that it has become feasible to attempt
to model this region in conjunction with solutions for the wave fields within the last closed flux surface.
We propose to utilize this computing capability in order to develop an integrated edge to core description
for ICRF and LHRF waves in tokamak plasmas. The problem will be separated into four major areas:
- Coupled core-to-edge simulations that lead to an increased understanding of parasitic losses in the
boundary plasma between the RF antenna and the core plasma.
- Simulations of core interactions of RF power with energetic electrons and ions to understand how
these species affect power flow in the confined plasma.
- RF effects on fast-particle driven instabilities to understand if these interactions increase
(decrease) the instability drive that can lead to reduced fusion power.
- Improved algorithms to achieve the needed physics, resolution, and/or statistics to address these
issues and to efficiently utilize new computer architectures.
These simulation tools will be validated against experiment through on-going collaborations with
experiments on the DIII-D, NSTX, and Alcator C-Mod devices. The most complete simulation models
we develop will be compared against reduced descriptions in order to delineate regimes of applicability of
the simpler models so that these codes can be used in whole device simulations such as those
contemplated for FSP. We shall also collaborate with applied mathematicians to implement algorithmic
improvements in our codes, especially in cases where the most complete and largest scale simulation is
required. These activities will therefore involve active collaborations among computational physicists,
theorists, applied mathematicians, and experimentalists.