Theory Department Highlights

Jin Chen has extended the M3D scaling studies up to 5120 nodes
(10,240 processors) on the Jaguar Cray XT3 computer at ORNL.
Particularly impressive were the results obtained with weak toroidal
scaling when using only one of the two processor cores on each node.
She obtained in excess of 80% overall efficiency when going from 64
to 5120 nodes. This result was obtained using the "Hypre" algebraic
multigrid solver within PETSc for solving the compute intensive
linear equations in M3D each timestep that arise from the elliptic
equations. More details of the scaling results can be found athttp://w3.pppl.gov/md3/scale.html.

On December 14, 2006 Dr. A. Reiman organized and hosted a National Stellarator Theory Teleconference as a follow-up to the NCSX Research Forum. The teleconference was organized to bring theorists into the planning process for the NCSX research program and provide a forum for the discussion of the theory community's plans. There were ten presentations, representing five different institutions.
Presentations were given by Drs. D. Spong, (ORNL), J. Hanson,
(Auburn University) ; C. Hegna and K. Shaing, (Wisconsin); H.
Strauss, (NYU); and from the PPPL theory group, A. Reiman, H.
Mynick, G. Fu, N. Gorelenkov, and D. Monticello. The presentation
materials are available online at “http://ncsx.pppl.gov//Physics/NSTT/
Theory_Index.html.”

Drs. Herb Berk and Boris Breizman of IFS-Texas visited Drs. Guoyong
Fu and Nikolai Gorelenkov to discus integrated simulations of
energetic particles in burning plasmas and to plan for collaboration
in this topical area.

On December 12, 2006, Dr. Boris Breizman of IFS gave a theory seminar titled, “Fast Particle Interaction With Waves In Fusion Plasmas.” He discussed the challenge of avoiding strong collective losses of alpha particles and beam ions in future burning plasma experiments and the compelling need to quantitatively interpret the large amount of experimental data from JET, TFTR, JT-60U, DIII-D, and other machines. He then discussed how the present theory responds to the stated challenges and what theoretical and computational advances are required to address the outstanding problems.

PPPL participated in a conference call for the Center for Simulation
of Wave Interaction with MHD (CSWIM) on Thursday and was able to
report significant progress on two of the major components that we
are responsible for: (1) the Plasma State (PS) and (2) the
Equilibrium and Profile Advance (EPA). The PS component is the basis
for the communication mechanism that all CSWIM software components
use to exchange information. A prototype version has now been
released and has been successfully used to provide input to and
receive output from both the RF component (which presently invokes
AORSA) and the EPA component. It has now been demonstrated that the
EPA can produce all equilibrium and profile information needed to
initialize the PS component, and on subsequent calls will advance the
equilibrium and temperature, current, density, and flow profiles one
macro time-step using the energy, current drive, particle, and
momentum sources contained in the PS component. The TRANSP code is
also in the process of being modified to have the option of being
invoked as a “server,” receiving all of its input through a series of
plasma states from the PS component (rather than from tokamak
experiment MDS+ or Ufile databases) and writing all of its output
back into the PS component. This will be used to provide accurate
neutral beam and alpha heating sources.

Dr. Gorelenkov was appointed as a deputy leader of the NSTX MHD
experimental task force. The NSTX research team and theory division
efforts are working together to meet the DOE Joule milestone on the
observation and identification of Alfven instabilities in NSTX
experiments. Two prior meetings of the Energetic Particle Science
Focus Group enabled a coordinated set of experimental proposals aimed at studies of energetic particle instabilities and in particular
multi-mode induced transport of fast ions, which is of direct
relevance to ITER.

Dr. A. Reiman gave a presentation on “Theory and Modeling” at the
NCSX Research Forum on 12/7. The talk gave an overview of theory
projects and plans, both domestic and international, of potential
relevance to NCSX.

Tom Jenkins, a graduate student at the department of Astrophysical Sciences, Princeton University gave a Theory Seminar entitled, “Fluctuations and discrete particle noise in gyrokinetic simulations of drift waves.” He discussed recent research about the relevance of the fluctuation-dissipation theorem (FDT) to marginally stable nonequilibrium states of the plasma, with particular focus being given to the contribution of weakly damped normal modes to the fluctuation spectrum. It is found that the fluctuation energy carried in these normal modes exhibits the proper scaling with particle count (as predicted by the FDT in thermal equilibrium) even in the presence of drift waves which grow linearly and attain a nonlinearly saturated steady state. This favorable scaling is preserved, and the saturation amplitude of the drift wave unaffected, for parameter regimes in which the normal modes become strongly damped and introduce a broad spectrum of discreteness-induced background noise in frequency space. He also discussed the relationship of the present work to the more general issue of discrete particle noise in particle-in-cell simulations. This work was supervised by Dr. Wei-Li Lee.

A National Stellarator Theory Teleconference was hosted by Dr. A. Reiman on November 30. Dr. T. Kaiser from the Lawrence Livermore National Laboratory gave a presentation on "Heat Load on Divertors in NCSX". He described field line tracing, with a diffusion coefficient to model the diffusion of particles across the field lines. The magnetic field in the edge region for these calculations was obtained from the MFBE/VMEC code and from the PIES code with the Drevlak postprocessor. The non-uniform heat deposition calculated suggests modifications in the design of the divertor plates.

Neil Pomphrey presented a talk titled “Magnetic Flux Loop Design for
NCSX” at the 2nd Joint Meeting of US-Japan Workshop and Kyoto
University 21st Century COE Symposium on NEW APPROACHES IN PLASMA CONFINEMENT EXPERIMENTS IN HELICAL SYSTEMS at Auburn University, Nov. 13-15.

The Fusion Energy Science area had a strong presence at the
Supercomputing 2006 Conference (SC'06) held in Tampa, Florida, Nov.
11-17. An interview with PPPL's Chief Scientist William Tang was
featured in the official SC'06 promotional video, where he discussed
the impact of high performance computing and associated scientific
discovery on fusion research and the ITER project. This video was
prominently displayed during the course of the conference and can be
found at: rtsp://vstream.acm.org/supercomp/SC06.mp4).

The simulation was carried out by Stephane Ethier and the
visualization by ORNL Scott Klasky. A data workflow was used to
transfer the 1.6 Tbytes of data required for the visualization from
NERSC to ORNL while the simulation was running and with no impact on the calculation.

Dr. Wei-li Lee of the Theory Department presented the SciDAC GPS
Center’s research on advanced gyrokinetic particle simulations
(“Kinetic Simulation of Magnetic Fusion Plasmas”) as a featured
speaker in the Masterwork Session of SC'06 -- a session showcasing
state-of-the-art scientific research enabled by high performance
computing. Tang, Lee, and Dr. Stephane Ethier of the CPPG gave
invited presentations at the ORNL booth about recent research
advances enabled by their utilization of the massively parallel
leadership class computers at the National Center for Computational
Science (NCCS). FES-relevant talks were also presented by
researchers from ORNL and General Atomics.

Phil Colella (LBNL) visited R. Samtaney and S. Jardin to discuss the
SciDAC APDEC/CEMM collaboration on Adaptive Mesh Refinement
techniques applied to fusion MHD problems. While at Princeton, Dr.
Colella presented a well attended PICSciE/PICASso lecture on main
campus on “Design of High Resolution and Adaptive Methods for Partial
Differential Equations” and visited with several other faculty and
students from the Astrophysics Department.

Improvements have been made to TRANSP's ability to receive MHD
equilibrium data from external sources. Free boundary equilibrium
data, with accurate fields beyond the plasma boundary, can now be
provided. Also, direct representation of plasma surfaces (with no
Fourier transformation) can also be input, yielding better fidelity
to the original, externally calculated MHD equilibrium solution. The
improvements will be available for production use shortly.

It is observed that in the National Spherical Torus Experiment (NSTX)
the energy loss rate in the ion channel is often close to
neoclassical levels. Global gyrokinetic simulations have been carried
out to investigate both turbulent and neoclassical transport
properties for NSTX discharges. This study contributes several
interesting footnotes to the observation that the ion transport is at
neoclassical levels in NSTX plasmas. Nonlinear turbulence
simulations using the General Geometry GTC code show that ion
temperature gradient, (ITG), driven turbulence has significant
fluctuation amplitude, but drives insignificant ion energy transport
in NSTX (about neoclassical level, sometimes even below neoclassical
level). This distinguished feature is particularly in contrast to
anomalous transport in other machines, such as D3D, where ITG
turbulence is shown to drive large transport (10 x neoclassical
level), even though the mean turbulence fluctuation levels for the
two machines are actually comparable. It is also found that self-
consistent equilibrium EXB flows, which are determined by
neoclassical dynamics and calculated by GTC-Neo, a global PIC
simulation code, can strongly stabilize ITG modes. Ion-ion collisions
are shown to enhance ITG driven thermal transport, but not
significantly. In some NSTX H-mode discharges, ITG is shown to be
stable even without equilibrium EXB flow. This work was carried out
by W.X. Wang collaborating with T. S. Hahm, S. M. Kaye, W.W. Lee, J. Manickam, G. Rewoldt and W. Tang.

Four members of the SWIM project (Simulation of Wave Interaction with Magnetohydrodynamics...a prototype for the Fusion Simulation
Project), D. Batchelor, L. Berry, W. Elwasif (ORNL) and Prof. R.
Bramley (Indiana U.) visited PPPL for a 2-day “coding camp.”
Progress was made in interfacing several RF components to the “Plasma State” component, and in developing the driver routines for advancing the profiles and equilibrium and computing the relevant RF and NB sources.

Researchers from the Theory department contributed to eighty-four
presentations at the 48th Annual Meeting of the APS Division of
Plasma Physics, October 30 - November 3 in Philadelphia. There were 5 invited talks, 24 contributed first author papers and 55 co-authored
papers.

The following scientists gave invited talks: Hong Qin on “Geometric
Gyrokinetic Theory for Edge Plasmas,” Igor Kaganovich on “Kinetic
effects in Hall thruster discharge,” Neil Pomphrey on “New Design
Methods for Magnetic Flux Loop Arrays in the NCSX Experiment,” Edward Startsev on “Collective Temperature Anisotropy Instabilities in
Intense Charged Particle Beams,” and Leonid Zakharov on “Ignited
Spherical Tokamaks as a Reactor Development Facility.”

On October 31, Joshua Breslau presented the invited talk “3D Modeling
of the Sawtooth Instability in a Small Tokamak.” The talk described
progress in the numerical simulation of sawtoothing discharges in the
CDX-U tokamak using the M3D and NIMROD nonlinear extended MHD codes of the CEMM SciDAC Center.

The first-author presentations were: ‘XTranspin, a Visual Data Input
Utility for TRANSP,’ by C. Ludescher-Furth et al.; ‘Status of TRANSP
Parallel Services,’ by K. Indireshkumar et al.; ‘New MHD Equilibrium
Solver Options in TRANSP’, R. Andre et al.; ‘Nonlinear Perturbative
Simulation Studies of Collective Processes in 3D Finite-Length Charge
Bunches at High Space-Charge Intensities,’ Hong Qin et al.; ‘Effects
of finite pulse length, magnetic field, and gas ionization on ion
beam pulse neutralization by background plasma,’ Igor D. Kaganovich
et al.; ‘Collective Interaction Processes in Intense Heavy Ion Beam-
Plasma Systems,’ Ronald Davidson et al.; ‘Simulation of Finite-Beta
Effects in Gyrokinetic Plasmas,’ Edward Startsev et al.;‘Motion of a
Localized Region of High-Density Plasma across Magnetic Flux
Surfaces,’ Ravi Samtaney et al.; ‘Application of GEM Code for
Experimentally-Realistic Tokamak Cases,’ G. Rewoldt et al.; ‘An
Implicit Method for Magnetic Fusion MHD Calculations using Adaptive,
High-Order, High-Continuity Finite Elements,’ S.C. Jardin et al.;
‘Two-fluid simulations of counter-helicity spheromak merging and
oblate FRC stability in MRX and SSX experiments,’ E.V. Belova et al.;
‘New beam driven modes below the geodesic acoustic frequency in
NSTX,’ Nikolai Gorelenkov et al.; ‘Theory and Simulation of Toroidal
Momentum Transport,’ T.S. Hahm et al.; ‘Implementation of Large Scale E×B Shear Flow in the GS2 Gyrokinetic Turbulence Code’, G.W. Hammett; ‘TRANSP and PTRANSP at PPPL: Status and Plans’, D. Mccune et al.; ‘The Influence of Noise on Turbulent Transport’, J.A. Krommes; ‘Simulation Studies of Electron Temperature Gradient Turbulence in NSTX Plasmas’, W.X. Wang et al.; ‘Isotope Effects due to Ion Temperature Gradient Drift Instabilities in Global Gyrokinetic
Particle Simulation,’ W.W. Lee et al.; ‘Equilibrium Flux Surface
Calculations for the W7AS and NCSX Stellarators,’ A. Reiman et al.;
‘Full-wave Simulations of ICRF heating in toroidal plasma with non-
Maxwellian distribution functions,’ E.J. Valeo et al.; ‘Magnetic
Alignment of NCSX,’ M.C. Zarnstorff et al.; ‘High frequency
gyrokinetic (HFGK) particle simulation of ion heating in the
cyclotron frequency range,’ R.A. Kolesnikov et al.; ‘Validation of
DEGAS 2 with NSTX Gas Puff Imaging Experiments,’ D.P. Stotler et al.
and ‘Comparisons of transport in quasi-axisymmetric stellarators,’
H.E. Mynick.

CPPG GROUP :
J. Chen “Scaling properties of the M3D code from CDX to ITER,” S.
Jardin “M3D-C1 update, GEM benchmark, and moving to 3D,” and R.
Samtaney “AMR update and plans” made presentations and participated in the SciDAC CEMM meeting on Oct 29th in Philadelphia. These and all other presentations have been posted on the site: http://w3.pppl.gov/CEMM.

R. Andre, S. Ethier, E. Feibush, K. Indireshkumar, S. Jardin, C.
Ludescher-Furth, D. McCune, and R. Samtaney attended the DPP/APS
meeting in Philadelphia. Together, they were co-authors on 24 oral
or poster presentations including 8 posters on the TRANSP/pTRANSP code.

R. Samtaney participated in the SciDAC CPES meeting on Nov 2nd and
made a presentation “Code coupling in CPES.”

Eliot Feibush attended the IEEE Visualization conference in Baltimore
and presented the poster, “ElVis: A Portal for Scientific Graphics,”
co-authored by Douglas McCune (PPPL) and Scott Klasky (ORNL).

Drs. N. Gorelenkov and J. Manickam participated in the 8th ITPA MHD
Topical Group Meeting in Chengdu, China, Oct. 23-25 2006

Stephane Ethier gave a 2-part hands-on tutorial entitled
“Introduction to Parallel Programming with MPI” at Princeton
University as part of the Computer Science Department Program in
Integrative Information, Computer and Application Sciences (PICASso). The course was attended by approximately thirty mainly Princeton University graduate students from several science departments.

Stephane Ethier and Ravi Samtaney attended the Fall Creek Falls
Conference organized by the ORNL National Center for Computational
Sciences. The focus of the conference was on petascale computing and the science that can be done at this scale. Dr. Ethier gave an
invited talk entitled “Gyrokinetic Particle Simulation of Fusion
Plasma,” which represented the work of both the SciDAC GPSC and CPES projects.

Igor Kaganovich attended the 2006 Gaseous Electronics Conference
(http://meetings.aps.org/Meeting/GEC06/Content/538) in Columbus,
Ohio, and presented three contributed papers on the kinetic modeling
of low-pressure discharges. In two presentations, the structures of
the capacitive sheath and skin layer were discussed. A thorough self-
consistent kinetic treatment of the problem reveals the existence of
an additional transition layer associated with finite electron
thermal effects. This transition layer is usually omitted in textbook
descriptions of low-pressure discharges. The third presentation
discussed conditions under which the electron velocity distribution
functions in low-pressure discharges can become non-isotropic and non- Maxwellian.

The numerical code Cbebm, created last year by Leonid E. Zakharov for 3-D heat transport simulations in sandwich-like materials (including
the case of a liquid lithium layer), has been complimented with a
numerical model of lithium evaporators (LITER-series) and the 3-D
geometry of the plasma facing components in both the NSTX and L-245 test chamber. The model assumes a collisionless lithium vapor with re evaporation of the Li molecules after their collision with the wall.
It uses an integral equation rather than the Monte-Carlo technique
for fast calculations of the evaporation diagrams. For the 2006 NSTX
experimental campaign, the code has generated the Li deposition data, which can serve as a reference distribution in the post-experiment material surface analysis.

Daren Stotler presented an invited keynote talk entitled “Atomic
Physics in ITER - The Foundation for the Next Step to Fusion Power”
at the 5th International Conference on Atomic and Molecular Data and
Their Application, October 15-19, Meudon, France. Dr. Stotler also
visited the Laboratoire d'Ingenierie des Materiaux et des Hautes
Pressions (Laboratory for the Engineering of Materials and of High
ressures) where he presented a seminar on “Plasma-Material
Interactions in a Monte Carlo Neutral Transport Code - What's Not in
the Model.”

At the IAEA Fusion Energy Conference in Oct. 16 – 21, Chengdu, China. Theory Department members presented six papers as lead authors. Three of the papers were presented orally as rappoteured papers. The titles and lead authors are: Dr Belova on'Effects of energetic beam ions on stability properties of FRCs'; Dr. Gorelenkov on, ‘Phase space gradient driven discrete CAEs in tokamaks, simulations and observations;’ Dr. Reiman on, ‘Localized breaking of flux surfaces and the equilibrium beta limit in the W7AS stellarator;’ Dr Hahm rappoteured the two gyrokinetic talks, Dr Wang's paper on
‘Gyrokinetic Studies of nonlocal properties of turbulence driven and
neoclassical transport;’ and Dr Lee's paper on ‘Long time simulations
of microturbulence in fusion plasmas;’ Dr Fu's paper on ‘Nonlinear
simulation of fishbone instabilities and sawteeth in tokamaks and
STs’ was rapporteured.

R. Samtaney and S. Jardin attended the SciDAC APDEC Center for
Enabling Technologies for the center's kick-off meeting on October
18,19 in Berkeley CA. The fusion CEMM is one of three application
projects that are being supported by the APDAC center. Samtaney
presented a talk entitled “Magnetic Fusion Application in APDEC.” The
talk described the current status of the adaptive MHD code used for
pellet injection, and future requirements for fusion MHD applications
to be investigated with the APDEC Center. This was followed by
extensive discussions of how the center would meet the needs of the
CEMM applications.

A prototype Plasma State Component has been released to the SciDAC SWIM Fusion Simulation Project. Built over the recently upgraded XPLASMA NTCC module, this tool contains sophisticated capabilities for extraction of information from tokamak MHD equilibria as well as interpolation or regridding of profiles of up to 3 dimensions. The contents of the Plasma State is controlled through a specification
file which is input to a Python script that generates all associated
code-- allowing for easy modification of the State definition, a key
requirement for research applications.

Jay Johnson attended the Applied Information Systems Research Program Principal Investigators meeting in College Park, MD and presented recent research on new methods to detect high-order dependencies in geospace data sets using an information-theoretical approach. In particular, it was shown that high speed solar wind velocity streams lead to a highly nonlinear response in magnetospheric dynamics, and a physical process was identified that could lead to such a response.

Professor Michael Berry from the Molecular Biology Department,
Princeton University, presented a CPPG seminar on “How Do Populations of Neurons Encode Visual Stimuli?” He reported on laboratory studies of neural computation and coding in the retina. Nearby ganglion cells are found to have spatial receptive fields that overlap significantly, leading to correlated firing and redundancy in the
visual information that the cells encode. These correlations can be
explained by a model that includes only pair-wise correlations,
identical to the Ising model from statistical physics. This predicts
that larger populations may exhibit a form of freezing transition
that allows for robust error correction.

In a recent calculation, detailed properties of the multispecies Weibel instability have been investigated for an intense ion beam propagating through background plasma. Assuming that the plasma electrons provide complete charge and current neutralization, detailed linear stability properties have been calculated within the framework of a macroscopic cold-fluid model for a wide range of system parameters. As a general remark, it is found that the exponentiation length for the multispecies Weibel instability scales approximately inversely as the beam ion plasma frequency, and is relatively long even for moderately dense ion beams. For example, for singly-ionized Aluminum beam ions propagating through dense background Argon plasma, the e-folding length is 12.7 m for beam density =10 10 cm - 3, and the e-folding length is 1.27 m for beam density =10 12 cm - 3. Details of these investigations are summarized in “Multspecies Weibel Instability for Intense Charged Particle Beam Propagation Through Background Plasma”, R. C. Davidson, M. Dorf, I. D. Kaganovich, H. Qin, A. B. Sefkow and E. A. Startsev, D. R. Welch, D. D. Rose and S. M. Lund, Nuclear Instruments and Methods in Physics Research A, submitted for publication (2006), which can also be accessed under Publications at http://nonneutral.pppl.gov .

An article detailing the construction of cantori for Hamiltonian flows (equivalently toroidal magnetic fields), by S.R. Hudson, has been accepted for publication in Physical Review E. Cantori are the invariant irrational sets remaining after the destruction of KAM surfaces and can create strong, but partial, barriers to transport in chaotic regions. Cantori are approximated by high-order periodic orbits. Field line tracing methods for locating periodic orbits perform poorly in chaotic regions, so Lagrangian variational principles are used. The method is robust to chaos, converges quadratically, and the computational cost scales linearly with the periodicity length of the orbit. Periodic-orbits have been constructed, with periodicities in the tens of thousands, that closely approximate cantori. Future work on this topic will determine what effect cantori have on numerical simulations (eg. by the M3D code) of heat transport in chaotic fields. A copy of the article may be found on http://w3.pppl.gov/~shudson/PAPERS/PUBLISHED/H_06.cantori.pdf

Stephane Ethier recently succeeded in running the GTC code on 10,386 processing cores on the ORNL Cray XT3 (Jaguar). He was able to advance 5.4 billion particles per step per second, a 13% percent
improvement over his previous record of 4.8 billion particles per
second on the Japanese Earth Simulator. This achievement was
highlighted in the recent HPCwire:http://www.hpcwire.com/hpc/943289.html

Fiscal year 2006 saw a large increase in PPPL Fusion Grid TRANSP run
production, compared to the preceding year. In FY 2006 records show
2742 runs on 9 current or proposed tokamak experiments; in FY 2005
there were 1401 runs. The US tokamak experimental projects making
use of the service were C-Mod, DIII-D, and NSTX. The international
projects making use of the service were ITER, Asdex-U, HL2A, JET,
JT-60, and MAST. US project run counts were 829 and 2317 in FY-2005 and FY-2006 respectively; international project run counts were 572 and 425 respectively.

The production version of the new Extended MHD code M3D-C1 has now installed the CPPG developed ElVis visualization and job monitoring
capability. ElVis displays extensive detailed graphics of 1D and 2D
data as a job is running on a remote computer. It can be invoked
from any Internet browser, and has capabilities for multiple
collaborative viewers. The same software can then be used at job
completion for exploring the data and the production of publication
quality PostScript files. This software can be downloaded for use in
other applications from http://w3.pppl.gov/elvis

Dr. Viktor Decyk of UCLA, Dr. Scott Klasky of ORNL and Prof. Zhihong
Lin of UC-Irvine visited their SciDAC GPS Center colleagues at PPPL
this week. The purpose of the visit was to coordinate the
collaborations on the GTC code and prepare for the upcoming IAEA and APS conferences.

Fang Liu from Indiana University visited PPPL this week and worked
with Jin Chen to port M3D. During the visit she made successful runs
on the 128-node Linux cluster, Odin, at Indiana University.

Viktor Decyk from UCLA and Scott Klasky from NCCS ORNL visited PPPL
to discuss with the GPS group and work with Stephane Ethier on GTC-
related projects. Drs. Decyk and Ethier completed their conversion of
the GTC code to a fully object-oriented structure that will
facilitate team coding. In order to maintain GTC's high performance,
the core routines remain essentially unchanged. Dr Ethier also worked
with Dr Klasky on testing new information-rich output files for GTC
diagnostics. The new data structure will greatly simplify the
management of diagnostics and its related tools.

A National Stellarator Theory Teleconference was hosted on June 29 by Dr. A. Reiman. Professor Thomas Pedersen of Columbia University gave a presentation on “The Columbia Non-neutral Torus: Status, Plans, and Opportunities for Theory Research”. The Columbia Non-neutral Torus (CNT) is a stellarator that is presently being used to confine electron plasmas. There are also plans to use it in the future to study partially neutralized plasmas, and to study the possibility of
producing electron-positron plasmas. The physics in this device
differs from that in conventional stellarators because of the strong
electrostatic field. Some theoretical work has been done on the
equilibrium, macroscopic stability, and neoclassical transport.
Intriguing theoretical questions remain: What is the physics of the
observed ion instability in CNT? Why is confinement so sensitive to
neutral pressure in CNT? Why does the neutral driven transport scale
as B^(-1.5)? Can we understand the profiles of density and
temperature in CNT? Can a CNT-like configuration be fusion-relevant?

Dr. Jay Johnson attended the International Symposium on Space Climate in Sinaia, Romania and presented an invited talk entitled, “The
Variability of Magnetospheric Dynamics over the Solar Cycle.”

A parallel computational service consisting of a serial client
(TRANSP) and a parallel server (NUBEAM) has been successfully tested
and implemented. The tokamak transport analysis code TRANSP has a
large number of modules, which have traditionally run in a serial
mode. However, accurately modeling plasma processes involving fast
ion behavior requires computational power not available on a single
processor machine. To meet this need, we have parallelized the Monte Carlo fast ion module NUBEAM and are now running it as a separate “parallel service.” The serial client (TRANSP) runs on the Sunfire machines while the parallel service is run on the Gridn cluster. The service has been tested up to 8 processors. In the near future we
will deploy the parallel service on the more powerful Kestrel cluster
as part of the PPPL compute server. It is expected that the parallel
NUBEAM will enable routine TRANSP computations using a million or
more Monte Carlo particles which will result in greatly increased
accuracy and much smoother fast ion distribution functions.

J. Chen, S. Jardin, and D. McCune, together with M. Chance (PPPL
Theory) attended the all-hands SWIM workshop (Simulation of Wave
Interaction with MHD) at Indiana University Sept 25-27. Status
reports were presented in all Physics, Math, and Computer Science
(CS) areas. The PPPL developed Plasma State module XPLASMA2 is
serving as the central communication component which the RF,
Stability, and Transport components use to exchange data. Several
near term physics applications on CMOD, DIII-D, and ITER were
identified to help prioritize the software development activity. A
complete posting of presentations and a summary of the discussions is
available at the project web site: http://www.cswim.org.

Dr. T.S. Hahm attended IEA large Tokamak Workshop on Edge Transport in Fusion Plasmas which was held on Sept 11-13, 2006 in Krakow, Poland, and presented a talk entitled “Role of turbulence spreading in edge-core coupling.”
This presentation was partially based on recent GTC simulations in
shaped toroidal geometry performed by Drs. W. Wang et al.

Daren Stotler gave an invited plenary talk entitled “Atomic Physics
in ITER - The Foundation for the Next Step to Fusion Power” at the
2006 Research Meeting of the Atomic, Molecular and Optical Sciences
(AMOS) Program in Warrenton, Virginia (September 10-13, 2006). This
annual meeting provides an opportunity for principal investigators in
the Office of Basic Energy Sciences AMOS Program to exchange
information.

Stotler's talk described the key roles played by atomic physics in
ITER and fusion devices in general. Although much of the emphasis was on edge and divertor plasmas, three core plasma diagnostics with
strong atomic physics connections were also discussed.

Version 2 of XPLASMA has been released this week. XPLASMA is an NTCC software module that stores a time slice tokamak “plasma state,” defined as an MHD equilibrium plus a collection of named profiles defined over flux, cylindrical, or velocity space coordinates.
Examples of the software capabilities are: interpolation of
profiles, coordinate transformations, and flux surface averages.
XPLASMA datasets are stored to, or retrieved from NetCDF files by
means of simple calls, allowing communication of state between
cooperating distributed processes. Version 2 allows multiple
distinct XPLASMA instances in a single executable program, enabling
time interpolation between two successive plasma states. The new
XPLASMA was developed to be used as the plasma state component in the SWIM Fusion Simulation Project's Integrated Plasma Simulator system.

Steve Krenzel, a Drexel Co-op student supervised by Eliot Feibush,
presented his work on scientific graphics for the ElVis software at a
CPPG seminar. He also demonstrated the integrated software development environment being used on the project

The collective processes in three-dimensional, finite-length charge
bunches at high space-charge intensities are described self-
consistently by the nonlinear Vlasov-Maxwell equations. The nonlinear
delta-f method, a particle simulation method for solving the
nonlinear Vlasov-Maxwell equations, is being used by Hong Qin et al.,
to study collective effects in high-intensity bunched ion beams of
interest for heavy ion fusion and high energy density physics
applications. For bunched beams, the equilibrium and collective
excitation properties are qualitatively different from those for
coasting beams. Due to the coupling between the transverse and
longitudinal dynamics induced by the 2D nonlinear space-charge field,
there exists no exact kinetic equilibrium that has anisotropic
temperature in the transverse and longitudinal directions. Even in a
thermal equilibrium distribution with uniform, isotropic temperature,
the particles' trajectories on constant energy surfaces are non-
integrable, which implies that it is not possible to perform an
integration along unperturbed orbits to analytically calculate the
linear eigenmodes. An approximate self-consistent kinetic equilibrium
has been established for bunched beams with anisotropic temperature.
Then, the collective excitations about the equilibrium are
systematically investigated using the delta-f method implemented in
the Beam Equilibrium Stability and Transport (BEST) code.
Dr. Lionello Marrelli from RFX, Padova Italy, is visiting PPPL to
collaborate with Dr. Roscoe White on high energy particle transport
in reversed field pinches and tokamak devices.

It was formally announced by DOE that the winner of the SciDAC2
Fusion competition is the proposal of the consortium, of which CPPG
is a participant, on “Framework Application for Core-Edge Transport
Simulations” (J. Cary, TechX lead PI). More details on the project
can be found at http://www.scidac.gov/fusion/fusion.html.
Dr. Andy Bauer, from Rensselaer Polytechnic Institute, spent the week
at PPPL interacting on the M3D-C1 project, and he presented a CPPG
seminar on “Developments to Improve Extended MHD Simulations in M3D- C1.” The talk covered two areas in which the RPI team has made improvements to the M3D-C1 extended MHD code. The first is on adding adaptive mesh technology to M3D-C1 and the second is on multilevel stabilization. More information on the talk can be found at http://w3.pppl.gov/cppg/seminars The poster, ElVis: A Portal for Scientific Graphics, by Eliot Feibush and Doug McCune (PPL) and Scott Klasky (ORNL) has been accepted for presentation at the IEEE Visualization 2006 conference. The poster describes a multi-tier architecture for adding browser based interactive graphics to the TRANSP plotting programs while minimizing the changes to the existing Fortran code.

Dr. N. N.Gorelenkov visited JET at Culham laboratory as a part of an
ongoing collaboration between PPPL and JET EFDA. Applying the
nonperturbative code NOVA-KN, with compressibility included, to JET
low density plasmas helped to show that the common understanding of high frequency fishbone-like oscillations (f=50-80kHz) as
conventional precessional frequency fishbones, is not correct.
Numerically it was shown that the fishbones are stabilized by finite
orbit width effects of fast ions. A new interpretation is offered,
based on the excitation of beta induced Alfven Eigenmodes. This was
supported by frequency comparisons and by magnetic signal analysis,
which also revealed that the instabilities are excited first as
steady modes with several toroidal mode numbers. In addition the
analysis revealed a new class of plasma eigenmodes with frequency
below the GAM frequency, consistent with theoretical prediction. Such
modes are similar to the Alfven cascades and exhibit low mode
frequency chirping on equilibrium time scales, as q drops below a
rational value. This new class of modes is unique in the frequency
range 0-30kHz and could be used as an MHD spectroscopy tool to
reconstruct the central value of the safety factor. This result
potentially offers the opportunity for central q or qmin measurements
in burning plasma experiments, such as ITER.

Alan Chin successfully completed his summer internship at PPPL this
week. Alan worked on the DG code, a graphical interfaced tool for
assembling geometries for use with edge simulation codes such as B2-
EIRENE and DEGAS 2. He completed his primary task, which was to
resolve incompatibilities between DG and the Lesstif and OpenMotif
libraries used to draw the graphical elements on the user's screen,
early enough that he was able to then add an auto-save feature to DG.

Gwang-Son Choe attended the third annual meeting of the Asia
Oceania Geosciences Society (AOGS) in Singapore on July 10-14, the
thirty-sixth Committee on Space Research (COSPAR) Scientific Assembly in Beijing, China, on July 16-23, and the 2006 Western Pacific Geophysics Meeting (WPGM) in Beijing on July 24-27. Dr. Choe was invited to give an oral presentation on "Stepwise Evolution of
Coronal Magnetic Field Leading to Large-Scale Solar Eruption" at the
AOGS Meeting, and also gave an invited presentation on "Energetics
and Dynamics of Solar Eruption with Magnetic Field Opening" at the
Western Pacific Geophysics Meeting. At the COSPAR Scientific Assembly in Beijing, he gave an oral presentation entitled "A Theory of the Roll Effect in Prominence Eruption." Dr. Choe was also a co-author of three other papers presented at these meetings.

Two-dimensional and three-dimensional simulations of counter-
helicity spheromak merging in the Magnetic Reconnection Experiment
(MRX) and the Swarthmore Spheromak Experiment (SSX) have been carried out by Elena Belova using the Hall-magnetohydrodynamic (MHD) version of the HYM code. The observable changes in global profiles caused by local Hall effects are seen even in the collisional regime, when the reconnection rate is comparable to the MHD reconnection rate. The dependence of the reconnection rate on plasma resistivity has been studied, and it is found to be similar to the scaling reported for the coalescence of magnetic islands. Three-dimensional simulations have also been performed to investigate the stability properties of oblate field-reversed configurations (FRCs). It is found that strong equilibrium field shaping can stabilize the n = 1 tilt mode in oblate FRC configurations, and improve the stability of the n > 1 MHD modes. The numerically-calculated structure of the perturbed magnetic field has been used to identify the most unstable modes in MRX experiments. Numerical simulations have also been performed to identify the effects of the central conductor and the residual small toroidal field on the stability properties of MRX-FRC plasmas.

A SWIM project remote meeting was held on 8/24/06. There was a
discussion of the present state of the framework services. There is
also now a 44 page Integrated Plasma Simulator (IPS) Design
Description document for the use of the different component
developers. The status of the RF component and of the Plasma State
component was discussed. Plans were discussed for the SWIM workshop to be held at the Univ. of Indiana on Sept 25-27, 2006. Progress is being documented on the SWIM web site: www.cswim.org.

Physics student Michael Richman worked with Stephane Ethier and Neil
Pomphrey on the optimization of the Extender code as part of the
National Undergraduate Fellowship (NUF) Program in Plasma Physics and Fusion Energy Sciences.

See .http://science-education.pppl.gov/Nuf/) Extender is a parallel
code for calculating the magnetic field in a stellarator outside of
the plasma. The code was optimized for speed, and augmented with
tools to maintain a specialized netCDF database. Two parallel
algorithms were examined and benchmarked. Large speedup factors
(10-15x) were achieved by representing the plasma surface with a
spline rather than Fourier series. Development of a Fortran 90
module makes it now easy to store Extender output in a manageable
database. New or updated data can be added to existing databases,
while a generalized PBS job script handles the generation of a
database from scratch.

Dr. Wolf-Christian Mueller from the Max Planck Institute for Plasma
Physics gave a recent seminar entitled "The Interplay of Kinetic and
Magnetic Energy in Magnetohyrdodynamic (MHD) Turbulence". In this
research, spectral direct numerical simulations of incompressible MHD
turbulence at a resolution of up to 1024 X 1024 X 1024 collocation
points for a statistically isotropic system as well as for a
configuration with an imposed strong mean magnetic field show
characteristically distinct behavior. The spectra of residual energy
and total energy are observed in the simulations to scale self-
similarly in the inertial range. A model of dynamic equilibrium
between kinetic and magnetic energy, based on the corresponding
evolution equations of the eddy-damped quasi-normal Markovian closure approximation, are found to explain the findings. The interplay of the turbulent dynamo and Alfven effects is also confirmed by the
simulations.

The values of ion-atom ionization and stripping cross sections are
frequently needed for many applications that utilize the propagation
of fast ions through matter. When experimental data and theoretical
calculations are not available, approximate formulas are frequently
used. A recent paper by Igor Kaganovich, Edward Startsev and Ronald
Davidson [New Journal of Physics, in press (September, 2006)]
summarizes the most important theoretical results and approaches to
cross section calculations in order to place the discussion in
historical perspective, and offers a concise introduction to this
topic. In addition, based on available experimental data and
theoretical predictions, a new fit formula for ionization cross
sections is proposed. The range of validity and accuracy of several
frequently used approximations (classical trajectory, the Born
approximation, and so forth) are discussed using, as examples, the
ionization cross sections of hydrogen and helium atoms by various
fully-stripped ions. Finally, a formulary of analytical
approximations for cross sections is presented.

Eliot Feibush presented a poster at the ACM SIGGRAPH conference on
computer graphics and interactive techniques held in Boston, MA. The
refereed poster, Multi-Tier Graphical Web Service for Simulating
Reflectometry in Plasma, was co-authored by Gerrit Kramer, Ernest
Valeo, Raffi Nazikian, and Douglas McCune.

Tarun Pondicherry completed his summer internship by presenting a
CPPG seminar on Developments for Simulation and Analysis Programs.
He was sponsored by the Science Education program and mentored by
Eliot Feibush. Tarun added capabilities to the ElVis graphics software for displaying the results of MSE simulations and generalized monitoring of programs that write data to netCDF files over time.

S. Jardin met with D. Pearlstein and L. Lodestro (LLNL) concerning
the details of implementing the TEQ equilibrium code as an option in
the free boundary pTRANSP program and the SWIM framework. A special iteration required for the simultaneous implicit treatment of the
Ohm's law and the TEQ solution of the equilibrium equation with
adiabatic constraints was discussed, along with a plan for
implementing it.

Leonid E. Zakharov gave a talk entitled “Getting Serious about
Fusion” to the UKAEA Fusion Theory Colloquium. The talk presented
his perspective on the optimal path for developing a fusion power plant.

D. McCune, together with A. Kritz (Lehigh U.) presented an update on
PTRANSP status and plans at DOE Headquarters on Wednesday, with audio participation by members of the project collaboration team from General Atomics, Lehigh University, LLNL, and PPPL. A significant
project milestone was reported, namely that PTRANSP is now capable of evolving a free boundary equilibrium combined with the full high
fidelity TRANSP alpha-particle, neutral beam, and RF heating and
current drive sources. This represents a unique capability in the
worldwide fusion program. A simulation was presented in which the
first 250 seconds of a projected ITER Hybrid scenario was modeled by
this technique, using the GLF23 thermal conductivity model. There
were significant changes in the predicted alpha-particle deposition
profile and neutral-beam driven current profiles from earlier
simulations in which more approximate heating and current-drive
models were utilized.

The VACUUM component of the VACUUM-DCON code used for modeling the feedback of the RWM has been modified to enable up-down asymmetric tokamak studies. It was checked to ensure that the degeneracies and orthnormality of the eigensystem used in our (Normal Mode Analysis) method were compatible with those of the DCON code and also with the circuit component of the closed loop feedback system. The code accurately reproduces the results of previous up-down symmetric cases. It's now being thoroughly checked to ensure that the sensors and feedback coils are calibrated correctly to enable physical comparisons, in real units, with experimental results. The previously diffuse, (in the poloidal coordinate), model of the scalar potential representation of the feedback coils has been replaced with a more accurate Heaviside model, and the magnetic fields and fluxes induced in the sensor coils are in excellent agreement with independent calculations using a direct Biot-Savart integration.
This, in addition to the already incorporated ancillary spin-offs
like the Mirnov loops and surface eddy currents modeling, will be of
significant utility to the feedback applications.

The code, using the DIII-D simulation as the test bed, is now in
public use. The NSTX configuration is under preparation as we
prepare to initiate feedback simulations on that device.

This is work being done as a collaboration with Ming Chu at General
atomics and with Michio Okabayashi and Hiro Takahashi,
experimentalists at PPPL.

The Finite Larmor Radius, Ion Cyclotron Radio Frequency wave propagation code TORIC has been generalized to allow the prescription of arbitrary distribution functions of the form, f(v_parallel, v_perp), at each point on a poloidal cross-section, (theta, psi), in the computation of the hot plasma dielectric tensor. Previously, the function was restricted to a Maxwellian shape, constant on a flux surface. This advance will enable more realistic computations of heating and current drive in toroidal plasmas, to be performed, since the injection of energy via beams and RF often leads to significant modifications of the distribution function. E. Valeo and C. Phillips performed this work as part of PPPL's contribution to the RF SciDAC project, with support from the PPPL Advanced Scientific Computing initiative.

Dr. Nuno Loureiro from the Center for Multiscale Plasma Dynamics, University of Maryland, gave a CPPG seminar on 7/28/06 on “An iterative semi-implicit scheme for KAW-mediated
magnetic reconnection.” In the talk, he demonstrated a semi-analytic semi-implicit method, which allowed him to increase the timestep by a factor of 100 for kinetic Alfven wave (KAW) mediated reconnection problems. Dr. Roman Samulyak from the Computational Science Center at Brookhaven National Laboratory visited on 7/26/06 to discuss collaboration on pellet fueling calculations with R. Samtaney. While here, he presented a CPPG seminar on “Axisymmetric MHD Simulation of Pellet Ablation”

About eighty plasma physicists from the United States, Europe, Russia and Japan met in San Malo, France on July 9 - 14 to participate in the 16th International Symposium on Heavy Ion Inertial Fusion. PPPL research in this important area was well represented by seven invited leadauthor papers, including: “Multspecies Weibel Instability for Intense Charged Particle Beam Propagation Through Background Plasma,” by Ron Davidson et al.; “Nonlinear Delta-f Particle Simulations of Collective Effects in High-Intensity Bunched Beams,” by H. Qin et al.; “Dynamic Stabilization of the Two-Stream Instability During Longitudinal Compression of Intense Charged Particle Beam Propagation Through Background Plasma,” by Ed Startsev et al.; “Effects of Finite Pulse Length, Electron Temperature, Magnetic Field, and Gas Ionization on Ion Beam Pulse Neutralization by Background Plasma,” by Igor Kaganovich et al.; “Optimized Simultaneous Transverse and Longitudinal Focusing of Intense Ion Beam Pulses for Warm Dense Matter Applications,” by Adam Sefkow et al.; “Conditions for Minimization of Halo Particle Production During Transverse Compression of Intense Ion Charge Bunches in the Paul Trap Simulator Experiment (PTSX),” by Erik Gilson et al.; and “Halogens for Negative Ion Beams and Ion-Ion Plasmas,” by L. R. Grisham et al.; and two contributed papers entitled: “Ferroelectric Plasma Source for Heavy Ion Beam Charge Neutralization,” by Philip Efthimion, Erik Gilson et al.; and “Robust Method for Evaluating Ionization, Charge Exchange and Stripping Cross Sections in Atom-Ion Collisions,” by Igor Kaganovich et al. Manuscripts describing this research have been submitted for publication in a special peer-reviewed issue of Nuclear Instruments and Methods in Physics Research - Part A.

In response to a proposal developed by Dr. Stephane Ethier of the CPPG and Chief Scientist Bill Tang, IBM selected the Gyrokinetic Toroidal Code (GTC) to carry out not only continuing
scaling studies but also some dedicated physics production runs at the IBM T. J. Watson Research Center. Dr. Ethier successfully ran the largest concurrency GTC simulation to date: 32,768 processors on the Blue Gene L computer enabled by the winning PPPL-PSACI proposal for the joint IBM-ANL “BGW (Blue-Gene-Watson) Day” for which researchers are awarded time on the 20-rack BGL system at the IBM Watson Center. Although GTC had previously been ported to the single-rack (2048 processors) BGL computer at ANL for initial scaling studies, very valuable BGL-specific optimizations were subsequently added to the code with the help of IBM BGL expert Bob Walkup as a result of a separate IBM/Princeton University collaboration. Overall, the benchmark tests now show perfect scalability of GTC on the full machine and also demonstrated very efficient use of the second processor core on each node. A physics production run simulation was carried out on the largest BGW partition available, totaling 32,768 processors. This calculation focused on the challenging problem of assessing the influence of collisions on nonlineary well-evolved plasma microturbulence driven by ion temperature gradients (ITG) dynamics using a phase space resolution at the unprecedentedly high value of 1024 particles per cell. The IBM hosts were very pleased that this ambitious run was successfully completed in a little over 3 hours -- a record time for such a large simulation. Analysis of the data from these simulations is in progress.

On July 14, 2006 Dr J. Manickam presented the Theory Department's response to the Recommendations following the Theory Review at the DOE Headquarters in Germantown MD. He indicated that the Theory Department appreciated the positive review of its Five Year Plan; that the Review committee made valuable suggestions, which have been recognized as adding value to the plan; and that the Department intends to address the recommendations. He presented plans to address each of the recommendations and identified preliminary efforts, presently underway. Participants in the discussion included, Drs. C. Bolton, M. Crisp, R. Dagazian, S. Eckstrand, J. Mandrekas. D. Markevich, A. Rosenberg, and F. Thio from the Office of Fusion Energy Science and R. Hawryluk from PPPL.

S. Jardin and R. Samtaney attended the Annual SIAM Meeting in Boston and made presentations in the “Advances in Computational Magnetohydrodynamics” session on “An implicit Method for Magnetic Fusion MHD Calculations Using Adaptive, High-Order, High-Continuity Finite Elements” and “Adaptive Mesh Simulations of Pellet Injection in Tokamaks,” respectively. Copies of the talks can be found at the site http://w3.pppl.gov/CEMM at the bottom of the “talks” page.

Chief Scientist Bill Tang chaired the annual Scientific Discovery through Advanced Computing (SciDAC) 2006 Conference in Denver, Colorado from June 25-29, 2006. This conference showcased the exciting SciDAC-sponsored computational science that has been accomplished during the past year across many scientific domains, with an emphasis on science at scale. The Fusion Energy Science (FES) area was well represented by invited oral presentations from Dr. Wei-li Lee from PPPL on “Gyrokinetic Particle Simulation of Fusion Plasmas: Path to Petascale Computing,” and Dr. Dylan Brennan from General Atomics/MIT on “Computing Nonlinear Magnetohydrodynamic Instabilities in Fusion Plasmas.” Invited poster presentations from FES included contributions from Drs. Josh Breslau of PPPL on “Massively Parallel Modeling of the Sawtooth Instability in Tokamaks,” C. S. Chang of NYU/Courant on Gyrokinetic Particle Simulation of Neoclassical Transport in the Edge Region of Tokamak Plasmas,” Fred Jaeger of ORNL on “Quasilinear Evolution of Non-thermal Ion Distributions in ICRF Heating,” Yang Chen of U. Colorado on “Gyrokinetic Particle Simulation of Electromagnetic Effects on Turbulent Transport in Fusion Plasmas,” and David Schissell of GA on “Collaborative Technologies for Distributed Science: Fusion Energy and High-Energy Physics.” In addition the FES applications were prominently featured in other invited oral presentations from the Computer Science and Applied Mathematics areas (e.g., Drs. Timothy Tautges of Sandia and Jeff Hittinger of LLNL). Exciting computational science that has been accomplished outside of the SciDAC Program both nationally and internationally was also a key component of the conference agenda. This helped to foster communication between SciDAC computational scientists and those funded by other agencies. New challenges and opportunities in interdisciplinary computational science were emphasized in the Advanced Architectures Panel and by Dr. Victor Reis, Senior Advisor to the Secretary of Energy, who gave a featured presentation on “Simulation, Computation, and the Global Nuclear Energy Partnership.” Overall the conference provided an excellent opportunity to highlight the rising importance of computational science in the scientific enterprise and to motivate future investment in this area. More comprehensive information about the SciDAC 2006 Conference may be obtained at: http://www.scidac.org/Conference2006/.

Dr. A. Reiman visited the Max-Planck Institute in Greifswald, Germany, from June 28 through July 5. He presented a seminar on “Flux Surface Calculations for the W7AS and NCSX Stellarators.” During the week, there were extensive discussions of the W7AS calculations, and of plans for further calculations. There were also discussions concerning applications of the PIES code to the W7X stellarator that the Greifswald theory group is pursuing. In addition, a new project has been initiated for benchmarking the PIES and HINT codes against each other, with participants from Princeton, Germany, and Japan. The German group has volunteered to do much of the code development that will be needed to construct an interface between the two codes. Dr. Gwang-Son Choe attended the 37th Meeting of the American Astronomical Society Solar Physics Division held in Durham, NH on June 26-30, where he presented a paper entitled,“Generation and Expulsion of Erupting Flux Rope by Multi-step Merging Processes.” In this paper, he showed how an eruptive flux rope is formed from a collection of small scale flux tubes and how mutual and self magnetic helicities are transformed during this evolution. He was also a co-author of two other papers: one on the magnetic breakout in coronal mass ejections and the other on the helicity accumulation in solar flares.

A prototype pTRANSP client-server configuration has been successfully tested using ITER simulation data. In this configuration, TRANSP acts as a server, computing RF, neutral beam and fusion product heating and current drive at the request of the free boundary predictive code client (TSC). The TRANSP server simultaneously carries out an analysis of the predictive code results, using the same methods as currently employed for tokamak experimental data analysis. A standard TRANSP archive output is produced. Equilibrium and profiles are communicated using the NTCC XPLASMA module; synchronization is done by ascii signaling files.

S. Jardin and R. Samtaney each presented a CPPG seminar on “An Implicit Method for Magnetic Fusion MHD Calculations using Adaptive, High-Order, High-Continuity Finite Elements,” and“Adaptive Mesh Simulations of Pellet Injection in Tokamaks,” respectively. These talks will be presented again next week at the SIAM Annual meeting in Boston.

Dr. Wei-li Lee presented an invited talk on "Gyrokinetic Particle Simulation of Fusion Plasmas: path to petascale computing," at the SciDAC 2006 Conference in Denver (6/25-29/06). The talk described the achievements using gyrokinetic particle simulation for studying turbulent transport on the state-of-the-art computers since the early nineties. These include the identification of ion temperature gradient (ITG) drift instabilities as a primary cause for the observed turbulence spectra in the tokamak experiments, the importance of the nonlinearly-generated zonal flows for the global ITG turbulence, the transition from Bohm to GyroBohm scaling as a function of the plasma size and the role of parallel acceleration of the particles in steady state turbulence. Moreover, with the availability of terascale computing capability recently, particle number convergence studies unprecedented in particle simulation history have also been carried out. The present activities. in terms of developing more realistic physics models and efficient numerical
algorithms to meet the new challenges of petascale computing before the end of the decade, wasalso discussed.

Neil Pomphrey attended the 33rd EPS Conference on Plasma Physics in Rome, Italy, where he presented a poster entitled, "Magnetic Flux Loop Design for NCSX".

Dr. Gwang-Son Choe developed a novel method of estimating magnetic reconnection rates in solar flares from the velocity and area of the moving H alpha blue-wing emission features and the corresponding X-ray flux. Profs. J. Lee and D. E. Gary, his collaborators at NJIT, applied this method to their observations and found that the time variation of the magnetic reconnection rate thus derived matches well the time profiles of the observed hard X-ray and microwave fluxes in a solar flare. A co-authored paper entitled “Magnetic Energy Release During the 2002 September 9 Solar Flare” has been accepted for publication in Astrophysical Journal.

R. Andre, S. Ethier, D. McCune, as well as L. Randerson and K. Ying
[PPPL Computer Division] attended a five-hour training on the
Totalview debugger-- uniquely capable licensed software used to debug serial and parallel applications written in C, C++, Fortran and/or a mixture of these compiled languages. Totalview is available on many
PPPL linux cluster computers. The training was presented by Peter
Thompson, senior customer support engineer for Etnus Corporation,
provider of the Totalview software.

R. Samtaney and K. Indireshkumar represented PPPL at the
Computational Science Graduate Fellowship (CSGF) Conference in
Washington DC from June 20 - 22. They presented a poster
highlighting the computational plasma physics research at PPPL. One
of the objectives of their attendance was to attract CSGF fellows to
spend a summer at PPPL working with a staff member on a computational project.

Dr. Scott Kruger, Tech-X Co., visited for three days to initiate work
on the linear stability component of for the SWIM project. The
objective is to provide a convenient and automated way of interfacing
the various linear MHD codes, including the NOVA-K code, to the
evolving Fusion Simulation Project (FSP) Integrated Plasma Simulator
(ISP). Also discussed was progress and plans on the NIMROD/M3D CDX-U nonlinear comparison thrust.

A paper entitled, “Finite element implementation of Braginskii's gyroviscous stress with application to the gravitational instability,” by graduate student N. M. Ferraro and Dr. S. C. Jardin was accepted for publication by Physics of Plasmas. In this article, a general coordinate-independent expression for Braginskii's form of the ion gyroviscosity in the two-dimensional potential field representation is presented, and is implemented in a full two-dimensional, two-fluid extended magnetohydrodynamic (MHD) numerical model. The expression for the gyroviscous force requires no field to be differentiated more than twice, and thus is appropriate for finite elements with first derivatives continuous across element boundaries (C1 finite elements). From the extended MHD model, which includes the full gyroviscous stress, are derived linear dispersion relations of a homogeneous equilibrium and of an inverted-density profile in the presence of gravity. The treatment of the gravitational instability presented here extends previous work on the subject. Linear and nonlinear simulations of the gravitational instability are presented. Simulations are shown to agree closely with the derived dispersion relations in the linear regime. The “gyroviscous cancellation” effect is demonstrated, and some limitations of the ~v-star approximation are discussed.

Dr. H. Mynick presented a talk, “Perturbing QA stellarators: What helps and what hurts confinement” at a National Stellarator Theory Teleconference on Tuesday, June 13, hosted by Dr. A. Reiman. The work, done in collaboration with Drs. A. Boozer and L.-P. Ku, distills features of the magnetic geometry in quasi-axisymmetric stellarators such as NCSX, which are harmful or beneficial for confinement. Loss of stellarator symmetry does not in general degrade neoclassical confinement, and in some cases it may improve it.

Dr. Hong Qin attended the 39th ICFA Advanced Beam Dynamics Workshop on “High Intensity High Brightness Hadron Beams, HB2006” in Tsukuba City, Japan, from May 29 to June 2, 2006. Dr. Qin presented an invited talked titled “ Progress in Accelerator R&D for High Energy Density Physics and Warm Dense Matter Applications” on behalf of the U.S. Heavy Ion Fusion Science Virtual National Laboratory. The current research objectives of the U.S. Heavy Ion Fusion Science Virtual National Laboratory include: achieving Warm Dense Matter (WDM) conditions on near-term experiments and addressing the top-level scientific question for High Energy Density Physics and Heavy Ion Fusion Science: “How can heavy ion beams be compressed to the high intensities required for creating high energy density matter and fusion ignition conditions?” The accelerator R&D effort is focused on the Neutralized Drift compression Experiment (NDCX), experimental and WARP 3D simulation studies of electron cloud effects in high brightness beam transport, and advanced theory and simulation. NDCX has achieved a longitudinal compression factor of 60 by imposing a linear head-to-tail velocity tilt to the beam, and letting the beam subsequently drift through a background plasma, which neutralizes the otherwise intense space-charge force. Comprehensive 3D simulations of the neutralized drift compression dynamics using the LSP code have been carried out, and the agreement with experiments is excellent. A three-dimensional kinetic model for longitudinal compression in charge-neutralizing background plasma has been developed. It has been shown that the Vlasov equation possesses a class of exact solutions describing both transverse and longitudinal compression of the beam pulse. Electron-ion two-stream instabilities and the temperature-anisotropy instability have been simulated using a low-noise delta-f method by the Beam Equilibrium Stability and Transport (BEST) code, which was recently optimized for massively parallel computers and applied to studies of the collective effects in 3D bunched beams.

Prateek Sharma, a Graduate Student in the Astrophysics Department of Princeton University presented the CPPG seminar on “Anisotropic conduction with large temperature gradients”. In his talk he critically examined various finite difference methods used for numerical implementation of anisotropic thermal conduction along the magnetic field direction.

The annual NERSC Users Group (NUG) meeting was held at PPPL, hosted by NUG Vice-Chair, Stephane Ethier. The 2-day meeting included a discussion of the NERSC-V procurement and an all day session on programming the new Bassi machine. Application talks were given by H. Qin (PPPL Theory) and S. Ethier.

Igor Kaganovich participated in the RHIC “Electron Cooling and High-brightness beams” Workshop at BNL, http://www.bnl.gov/cad/ecooling/. In the electron cooler, “low-temperature” electron bunches co-propagate with relativistic 100 GeV/u, completely stripped Au (gold) ions for about 80 meters and cool “hot” ions. The goal of electron cooling is to reduce the emittances of the ion beams and to increase RHIC luminosity by about ten fold. Igor Kaganovich gave a presentation entitled “Theory of recombination and experimental results” on critical issue in the cooling section that determines the ion beam losses.

Wei-li Lee was the speaker for the PPPL colloquium this week. The title of the talk was “Kinetic Simulations of Fusion Plasmas and the SciDAC Challenge,” and the focus was on the use of particle codes in simulating gyrokinetic physics in magnetic fusion plasmas on modern-day massively parallel computers. This work, headed by PPPL in collaboration with seven other universities and laboratories, is in response to the DoE initiative of Scientific Discovery through Advanced Computing (SciDAC), which calls for an increase in the capabilities of scientific codes in keeping up with the corresponding increases in the power of supercomputing. The use of the global Gyrokinetic Turbulence Code (GTC) for turbulence studies in tokamaks as well as its use in the future for the realistic simulations of ITER-type burning plasmas was discussed.

The poster, Multi-Tier Graphical Web Service for Simulating Reflectometry in Plasma, by Eliot Feibush, Gerrit Kramer, Ernest Valeo, Raffi Nazikian, and Douglas McCune, has been accepted for presentation at the ACM SIGGRAPH 2006 conference. The poster describes adding web-based graphical input and visualization to the 2-D Full Wave Reflectometer simulation while making minimal changes to the original code. A multi-tier system was developed to optimize computing and display resources.

Doug McCune made a remote presentation to the General Atomics Integrated Modeling group on the progress of NUBEAM parallelization. It was shown that NUBEAM scales linearly up to at least 16 processors. A MPI Parallel NUBEAM server is presently under development.

Dr. Viktor Decyk from UCLA visited PPPL to work with Stephane Ethier on object-oriented modifications to the GTC code. The new code structure will facilitate team development and allow simulations with multiple ion species. This work is part of the SciDAC GPS Project

Roscoe White and co-authors, Ya.I. Kolesnichenko, V.V. Lutsenko, V.S. Marchenko, et al. submitted a paper entitled 'Magnetohydrodynamic activity and energetic ions in fusion plasmas', to Plasma Physics and Controlled Fusion. The new results are in relation to issues of lack of axial symmetry in stellarators and effects of finite orbit width. They show that in stellarators, high frequency Alfven eigenmodes can be localized on the inner circumference of the torus and that finite orbit width of fast ions can lead to additional resonances. They address these features in the context of the W7-AS experiment.

Wei-li Lee, representing the SciDAC Center for Gyrokinetic Particle Simulation, reported to the annual PSACI PAC review panel on the latest progress of the Center. The focus of the presentation was on the issues of long time simulations and numerical convergence studies of microtubulence as well as on the performance of PIC codes on modern MPP platforms.

S. Jardin made a presentation to the PSACI PAC on behalf of the SciDAC Center for Extended MHD Modeling. Items highlighted were: progress in fluid closures, numerical methods and numerical benchmarks, and updates on the major physics studies in the areas of sawtooth simulations, ELM modeling, energetic particle modes, and pellet fueling simulations.

A. Kritz and G. Bateman [Lehigh University] visited to discuss progress made and future plans of the Lehigh group on the predictive TRANSP project, PTRANSP. Items discussed were new capabilities added in the areas of computational steering (the ability to change input parameters at restart time) and the incorporation of an edge pedestal model into a predictive simulation.

On May 19th Jerome Lewandowski presented an invited talk at the 5th
International Conference on Scientific Computing held in Banff,
Canada (May 18-May 22). The talk, entitled “Marker method for the
solution of nonlinear partial differential equations”, is an
extension of a new method used for the implementation of low-noise
collision operators in particle-in-cell simulations. The new method,
unlike the conventional approach for the implementation of
collisional effects, does not rely on operator splitting or the
generation of random numbers.

Robert Budny, Guoyong Fu, Taik-Soo Hahm, Janardhan Manickam, Hong Qin, Bill Tang, and Weixing Wang, from PPPL, participated in the
Workshop on ITER Simulations held in Beijing, China, May 15-19,
2006. This workshop was sponsored by the ITER-CN-PT, the two fusion laboratories in China (ASIPP and SWIP), and the Peking University and was attended by a total of 80 peoples including about 20 from U.S., Japan and Europe. The purpose of the workshop was to provide a survey of the status of fusion simulation and modeling among ITER partners; to discuss long-term strategy for ITER simulation in China, and to initiate international collaborations. A whitepaper was written on the status of fusion simulation and modeling and will be published in a fusion journal. Based on the discussions at this workshop, a working group was formed to develop a proposal for China ITER Simulation Initiative (CISI) by the end of this year.

A series of two-dimensional and three-dimensional high-resolution
simulations of counter-helicity spheromak merging have been performed by Elena Belova using the two-fluid version of the HYM code. These simulations show significant differences in the radial current, toroidal magnetic field and radial flow profiles compared to the
resistive MHD simulations. New signatures of Hall-reconnection,
characteristic of counter-helicity merging in the toroidal geometry,
have been found in the simulations. These include, in particular, the
radial shift of the reconnection X-point, 'V'-shaped radial current
contours, and nearly unidirectional radial ion flows, all dependent
on the polarity of the initial toroidal field. The observable changes
in global profiles caused by the local Hall effects, are seen even in
the collisional regime, when the reconnection rate is comparable to
the MHD reconnection rate. These effects are specific to the counter-
helicity magnetic field geometry, where the reconnection plane is
tilted relative to the poloidal plane. The observed features are
driven by the electron flows (local to the X-point) and are related
to the generation of a quadrupole magnetic field in Hall-MHD
reconnection. The simulation results are found to be consistent with
measurements in the Magnetic Reconnection Experiment (MRX). The
dependence of the reconnection rate on plasma resistivity has also
been studied, and the scaling dependence is found to be similar to
the scaling reported for the magnetic island coalescence problem.

The values of ion-atom ionization and stripping cross sections are
frequently needed for many applications that utilize the propagation
of fast ions through matter. When experimental data and theoretical
calculations are not available, approximate formulas are frequently
used. A recent paper by Igor Kaganovich, Edward Startsev and Ron
Davidson [New Journal of Physics, in press (2006)] summarizes the
most important theoretical results and approaches to cross section
calculations in order to place the discussion in historical
perspective and offer a concise introduction to this topic. In
addition, based on experimental data and theoretical predictions, a
new fit formula for ionization cross sections is proposed. The range
of validity and accuracy of several frequently used approximations
(classical trajectory, the Born approximation, and so forth) are
discussed using, as examples, the ionization cross sections of
hydrogen and helium atoms by various fully stripped ions. Finally, a
formulary of analytical approximations for cross sections is presented.

A National Stellarator Theory Teleconference was hosted by Dr. A.
Reiman on Thursday, May 11. The presentation by Professor S.
Knowlton of Auburn University described the status, plans and
opportunities for the recently completed CTH device at Auburn
University. CTH is able to produce a substantial externally generated
rotational transform, as well as a substantial rotational transform
generated by an ohmic current. The device will offer an opportunity
to study key physics issues related to island formation,
magnetohydrodynamic stability, and disruptions. The experiment will
also serve as an important test bed for the development of new 3D
equilibrium reconstruction capabilities.

Prof. Kumar Bobba from the Department of Mechanical Engineering,
University of Massachusetts, Amherst, presented a CPPG Seminar on
“Computational Distributed Control of Complex Flows.” He described
the computational challenges and an approach for feedback control of
multi-scale fluid flows using a spatially distributed network of
sensors, actuators and controllers.

Prof. David Keyes from Columbia University and two of his students
visited to discuss two projects they are involved in as part of the
Fusion Simulation Project. One is to develop an improved plasma
equilibrium code using PETSc and a Newton iteration to converge the
current nonlinearity. The other involves using dynamical grid
equations to evolve a mesh in an implicit MHD calculation to keep it
field-aligned. Presentations were also made by PPPL staff on ongoing
equilibrium code and implicit MHD code development.

To achieve the high focal spot intensities necessary for high energy density physics and heavy ion fusion applications, the ion beam pulse must be compressed longitudinally by factors of ten to one hundred before it is focused onto the target. The longitudinal compression is achieved by imposing an initial velocity profile tilt on the drifting beam in vacuum. To achieve maximum longitudinal compression, the space charge of the beam is neutralized by propagation through a dense neutralizing background plasma. If the space charge is fully neutralized by the plasma, the final compression is limited only by the initial longitudinal temperature of the beam ions and possible collective processes (such as the two-stream instability) which prevent full neutralization. The beam's longitudinal thermal spread which could stabilize the instability also inhibits full longitudinal compression. In a recent calculation by Ed Startsev and Ron Davidson, use has been made of a macroscopic fluid model to investigate both analytically and numerically the electrostatic two-stream instability for a cold, longitudinally-compressing charged particle beam propagating through a background plasma. It is found that the longitudinal beam compression alone strongly modifies the space-time development of the two-stream instability. In particular, it is found that the dynamic compression leads to a significant reduction in the growth rate of the two-stream instability compared to the case without an initial velocity tilt.

Bruce Scott, Max-Planck-Institut for Plasmaphysik, presented a theory seminar entitled ‘Energetics in full-f and delta-f gyrokinetic equations.’ The derivation of the delta-f Vlasov equation was presented starting with the general gyrokinetic field theory. The total-f version with Hamiltonian structure serves as the intermediary. The main delta-f approximation corresponds to a linearization of the field polarization and the parallel bracket. An energy conserving-form is found by applying the delta-f ordering to the spatial derivatives, and the conserved quantity is the well-known delta-f free energy, related to the entropy. The conservation properties are compared to those following from the application of the Noether theorem to the total-f model. Although a different energy is conserved, the field/particle transfer terms are the same

S. Jardin participated in a two day on-site review of the NSF funded “Center for Integrated Space Weather Modeling,” a 10 Institution consortium centered at Boston University.

Drs. Harry Mynick, Allen Boozer, and Long-Poe Ku have submitted a
paper entitled, “Improving confinement in quasi-axisymmetric
stellarators”, to the Physics of Plasmas. From analysis of the
confinement, for a set of quasiaxisymmetric (QA) stellarator designs,
they have found some simple rules to identify features of QAs, which
are harmful or beneficial for confinement. This provided physical
understanding of a loss channel present in some NCSX configurations,
and absent in others.

A Workshop on Long Time Simulations of Kinetic Plasmas sponsored by
the SciDAC GPS Center took place on April 21, 2006 in conjunction
with the Spring APS and Sherwood Meeting in Dallas, TX. There were
about 40 participants from the US, (PPPL, LLNL, LANL, Columbia,
Colorado, UC-Irvine, ORNL, NYU and DoE) and abroad, (France, Korea,
Australia). Twenty talks were given representing the MFE and IFE
communities as well as other related disciplines with focus on
theory, simulation and experiments related to the long time behavior
of simulation plasmas. The PPPL Theory participants were: S. Ethier,
G. Hammett, J. Krommes, T. Jenkins, R. Kolesnikov, W. Lee, J.
Manickam, H. Qin, G. Rewoldt, W. Wang, L. Zakharov). For more Detail
of the workshop, please visit: http://w3.pppl.gov/theory/GPSC.html.

Stephane Ethier attended this workshop made the presentation entitled "Convergence Studies of ITG using GTC". GTC simulations with a large number of particles where performed on the CRAY X1E at ORNL and showed that ITG results are well converged. The highly scalable
particle distribution algorithm implemented in GTC which allowed
such simulations also made possible the largest GTC run so far: a 28
billion particle simulation of ETG turbulence run on 4800 processors
of the CRAY XT3 leadership computer at ORNL's National Center for
Computational Sciences (NCCS).

Wei-li Lee gave an invited talk at the 2006 APS/Sherwood meeting on
“Long Time Simulations of Microturbulence.”

On April 27, Jerome Lewandowski (PPPL) presented a Theory Seminar on the topic of particle-in-cell simulations of electromagnetic
microturbulence using a fully kinetic approach for both the ions and
the electrons. The basic method is based on the fact, that for the
low-frequency modes of interest, most electrons respond adiabatically
to the waves: it is sufficient, therefore, to kinetically resolve
only the nonadiabatic response of the electrons. The model was shown to work well in a simple (quasi one-dimensional) geometry.

Emily Belli successfully defended her Ph.D. Her dissertation was on
“Studies of Numerical Algorithms for Gyrokinetics and the Effects of
Shaping on Plasma Turbulence,” and her advisor was Greg Hammett.

Prof. Bob Dewar of the Australian National University has been
visiting the Theory Department. He has been collaborating with
Stuart Hudson on the topic of constructing an MHD code suitable for
stellarator configurations. The motivation for the code is that
pressure gradients are incompatible with rational surfaces, which
experience shows leads to convergence problems for codes which assume smooth pressure profiles. Thus an equilibrium model is constructed where pressure gradients only exist at a finite set of irrational surfaces, between which the field is in a relaxed Beltrami state (with grad p = 0). A finite-element construction of the equilibrium has commenced.

Ravi Samtaney and S. Jardin attended the APS/Sherwood Fusion Theory Conference in Dallas, April 22-25, and presented papers entitled "Simulations of Edge Localized Modes and Pellet Injection" and
“Progress and plans for the PTRANSP project. They also attended the
CEMM satellite meeting on April 25th and presented talks on the
status of the AMR MHD code and on the M3D-C1 code, featuring new
benchmark results on the GEM reconnection problem and the
gravitational instability. The complete set of CEMM presentations is
available on the site: http://w3.pppl.gov/CEMM on the “workshops” page.

Doug McCune attended the 10th Meeting of the ITPA Confinement
Database & Modeling Topical Group held at PPPL on April 24-27 and
presented a description of the PTRANSP project at the International
Modeling session. He and S. Jardin also participated discussion on
Collaborative Opportunities in International Modeling.

An upgrade to the XPLASMA NTCC module has enabled support for a
robust representation of NSTX equilibria in Boozer coordinates,
despite the extreme non-linearity of the relationship between flux
coordinates and real space coordinates inherent in this form,
particularly in compact aspect ratio tokamaks. Progress has also
been made in a fortran-95 re-implementation of XPLASMA that will
enable "multiple instantiation", meaning (for example) that Boozer
and non-Boozer XPLASMA representations of the same equilibrium could coexist in a single executable program.

Dr. Daniel Reynolds (Dept. of Math., UC San Diego) was hosted by CPPG from April 17-19; and Dr. Carol Woodward (CASC, LLNL) visited PPPl on April 19. This is part of an ongoing collaboration between the CEMM,TOPS and APDEC SciDAC Centers.

This week a fully implicit scheme based on the Jacobian Free Newton-Krylov approach utilizing the SUNDIALS package was implemented in magnetic flux coordinates for resistive MHD; and tested for a pellet injection application without the use of preconditioners. The next steps will be to implement physics based preconditioners for this scheme.

Daren Stotler and Taik Soo Hahm attended the 2006 Transport Task Force Meeting in Myrtle Beach, South Carolina. Both participated in the Edge Coordinating Committee Workshop, (ECC), on Edge Plasmas in ITER and Burning Plasma Devices that preceded the TTF meeting. Hahm gave a talk entitled "Turbulence Spreading through a Transport Barrier", which was heavily based on Weixing Wang's GTC simulation in a shaped torus. This simulation not only offers a possible explanation for the existence of residual fluctuations and anomalous transport in the region where's no linear instability drive, but also clearly resolves some confusion regarding zonal flow effects on turbulence spreading. Stotler's presentation described a format for "reference problems" that can be used to test (verify) simulation codes; the ECC is assembling a database of these "reference problems" relevant to the plasma edge. Stotler's talk resulted in TTF discussions of and a commitment to assemble an analogous database of core plasma- relevant reference problems.

On April 13th, Dr Scott Tremaine, from the Department of Astrophysical Sciences, presented a Theory Seminar on the topic of the long-term stability of planetary systems. Using accurate N-body integrations, Dr. Tremaine showed that it is possible to follow the motion of planets reliably for billions of years, providing information on the evolution and current properties of planetary systems.

A general interface for constructing Poincare plots for stellarator
magnetic fields has been constructed by S.Hudson, N.Pomphrey and
D.Monticello. This interface takes the magnetic field produced by a
coil set, eg. the NCSX coil set, and the magnetic field produced by a
plasma, as calculated by an equilibrium code. The construction of the
Poincare plot is extremely fast (because the field is saved on a grid
and interpolated using 3d splines from the NTCC library http://
w3.pppl.gov/ntcc/PSPLINE) and allows the `point-and-click' feature of
interactively selecting field lines. Additional capabilities,
including determination of the magnetic axis, locating periodic
orbits and islands, constructing flux-surfaces, determining the
location of the last flux-surface, calculating Lyapunov exponents,
etc. are presently being added. The interface is presently being used
to study the footprint of particles on the first-wall. Contact Stuart
Hudson shudson@pppl.gov for more information.

The initial implementation of the higher order Lagrange and Lumped
finite elements in M3D suffered from an increased bandwidth in the
matrices, leading to substantially longer running times. J. Chen was
recently able to reorder these matrices using the RCMK technique
which resulted in a speedup in the solver time by almost a factor of
10. With this technique, the elliptic solvers in M3D using the
higher order elements are now of comparable speed to those using the
linear elements. The most recent version of the TORIC ICRF full wave solver has been installed in TRANSP, and its coupling has been adjusted per the advice of the TORIC authors-- with some improvements in the stability of TORIC operation expected as a result. TRANSP has been modified with an option to run as the client of a TORIC server-- a separate program-- enabling the option of accessing an MPI-parallelized version of TORIC in the near future. Prof. Randall Bramley (Indiana U.), Co-PI of the SWIM Fusion Simulation Project, visited PPPL the week of April 3. Among items discussed and accomplished were: servers for managing parallel jobs within a batch managed system; Issues related about using the I.U. cluster for TRANSP/TORIC remote jobs; Issues related to using ElVis
in SWIM; Linear solver technology and problems, and; a joint
discussion with Tech-X about joint projects with SWIM involving
Xplasma and associated interfaces.

A paper titled, "Effects of Pressure Gradient on Existence of Alfvén
Cascade Modes in Reversed Shear Tokamak Plasmas" by G. Y. Fu (PPPL) and H. L. Berk (IFS) has been accepted for publication in Physics of Plasmas. In the paper, it is shown analytically that pressure
gradient effects are favorable to the existence of Alfvén Cascade
modes localized near the minimum-q surface in a tokamak plasma with
reversed shear. This settles a recent debate on the sign of the
pressure gradient term in the mode existence criterion. What is
crucial for obtaining the improved existence criterion of these modes
is the averaged normal curvature. Near the rational surface, there is
one to one correspondence between Mercier stability and the cascade
mode existence, specifically, when the averaged curvature is
favorable to Mercier modes, it is also favorable to the existence of
the cascade modes.

S. Jardin, together with G. Fu, G. Hammett, and J. Breslau (PPPL
Theory) attended and gave presentations at a DOE sponsored "Closures Workshop" held at ORNL 22-24 March 2006. The workshop had approximately 35 participants. Besides presentations on different
approaches for obtaining a closed set of fluid equations for fusion
plasmas, there were several focused discussion sessions on (1)
computational benchmarks for extended MHD, (2) feasibility of a
kinetic model for global MHD, and (3) research required for RF and
NBI interaction with MHD. Prof. Callen led a summary session and
discussion the final day on "Discussion of MHD Research Strategy for
next 10 years". The presentations and summary sessions are posted on the site: http://w3.pppl.gov/CEMM/workshops.html

The OFES 2006 Theory 2nd quarter ELM High-Level Milestone has been
successfully completed, and a report has been posted on the site:
http://w3.pppl.gov/CEMM/milestones.html

G. Rewoldt presented a Theory Department micro-seminar on March 23, 2006 entitled “Comparison of Microinstability Properties for
Stellarator Magnetic Geometries.” The microinstability properties of
nine distinct magnetic geometries corresponding to different
operating and planned stellarators with differing symmetry properties
are compared. Specifically, the kinetic stability properties (linear
growth rates and real frequencies) of toroidal microinstabilities
(driven by ion temperature gradients and trapped-electron dynamics)
are compared, as parameters are varied. The familiar ballooning
representation is used to enable efficient treatment of the spatial
variations along the equilibrium magnetic field lines. These studies
provide useful insights for understanding the differences in the
relative strengths of the instabilities caused by the differing
localizations of good and bad magnetic curvature and of the presence
of trapped particles. The associated differences in growth rates due
to magnetic geometry are large for small values of the temperature
gradient parameter eta (the local ratio of the relative temperature
gradient to the relative density gradient), whereas for large values
of eta, the mode is strongly unstable for all of the different
magnetic geometries.

On March 21, Arnold Kritz, Glen Bateman, and Federico Halpern of the
Lehigh University Physics Department visited PPPL for collaboration
on PTRANSP. The Lehigh team is planning to add predictive features
to the code: pedestal boundary condition, transport effects of
neoclassical tearing modes, and the Porcelli sawtooth trigger model.
The system for shared TRANSP code development, and the testing of the system using tokamak experimental data, was demonstrated on the PPPL Display Wall. For this session, the Multi-Cursor Window Manager developed by the Princeton University Computer Science Department was used to allow multiple parallel interactive sessions to run simultaneously on the large shared display, a technique which shows promise for collaborative group learning and sharing of technical
information. The software was developed in the SciDAC Fusion
Collaboratory project and is also being tested in experimental
control rooms with large shared displays.

Dr. Giovanni Lapenta, from the Theoretical Division of Los Alamos
National Laboratory presented the CPPG seminar on “Kinetic approach
to microscopic-macroscopic coupling in fusion plasmas.”

Dr. Roscoe White has submitted a paper, entitled, “Phase integral
Methods” to the Royal Society, London. In this paper he offers a
generalization of a technique for solving differential equations, a
method more commonly known as the WKB approach. The phase integral method was previously described by Heading, through a set of rules for the continuation of solutions through the complex plane,
involving Stokes constants. These constants were derived analytically
for isolated singularities. White extends this to the case with
multiple singularities and shows that the Stokes constants are given
by the Euler gamma function with an argument depending on the
singularity separation.

The annual US/Japan JIFT Workshop, on “Issues in the theoretical
analysis of three dimensional configurations” was held at Princeton
Plasma Physics Laboratory on March 14-16. The meeting was attended by a party of five from Japan and several participants from around the US, in addition to various staff members of PPPL. Presentations at the meeting included those given by Drs. G. Rewoldt, H. Mynick, S. Hudson, D. Monticello and A. Reiman. Electronic copies of the presentations are available online at http://w3.pppl.gov/~shudson/ JIFT/2006/JIFT2006.html. The next JIFT workshop will be held next year in Japan.

Dr. Sanjukta Bhowmick from the Applied Physics and Applied
Mathematics Department, Columbia University, visited and presented a
CPPG seminar on “Selecting and Combining Solvers---A Multimethod
Approach to Solving Linear Systems.” The talk discussed how machine
learning techniques can be used to select an optimal solver among the
many available for large sparse linear systems of equations.

Both the pTRANSP and Fusion Simulation Project (FSP) are making use
of a “plasma state” component that is based on the NTCC module
Xplasma. In order to facilitate those projects, we’ve placed a
ssample Xplasma file -- a timeslice from DIII-D 104276 at 4.71s -- in
a package at: ftp://ftp.pppl.gov:/pub/dmccune/fsp/fsp.tar.gz. This
also includes the driver for a program that reads the data and shows
examples of accessing the data. The program, xplasma_fsp_test, has
also been added to the NTCC xplasma library module. The Xplasma file
contains basic plasma state information-- temperatures, densities,
equilibrium, and, heating and current drive profiles. Work is
presently proceeding on a modernization of Xplasma as required to
meet FSP requirements. This will include a fortran-90 style
interface (only a fortran-77 style interface is available at present,
as is shown in the test program). The fortran-77 interface is the
one documented at http://w3.pppl.gov/ntcc/Xplasma.

A review paper entitled “Foundations of Nonlinear Gyrokinetic Theory”
co-authored by A. J. Brizard (St. Michael's College) and T.S. Hahm
(PPPL) has been submitted to Review of Modern Physics for
publication. In this paper, the foundations of nonlinear gyrokinetic
theory are reviewed with an emphasis on the rigorous applications of
Lagrangian and Hamiltonian methods used in the variational derivation
of nonlinear gyrokinetic Vlasov-Maxwell's equations. The physical
motivations and applications of the nonlinear gyrokinetic equations,
which describe the turbulent evolution of low-frequency electromagnetic fluctuations in a nonuniform magnetized plasmas with arbitrary magnetic geometry, are also discussed. It contains the analytic theoretical progress made by many members of PPPL Theory Department since the early 80's when nonlinear gyrokinetic theory was invented.

Stephane Ethier, Eliot Feibush, W. W. Lee, Ravi Samtaney, Daren
Stotler, Taik Soo Hahm, and Weixing Wang attended a meeting of the
Center for Plasma Edge Simulation (CPES) at Rutgers University on
March 6 and 7. Lee provided an overview of gyrokinetic simulations.

Yaroslav Kolesnichenko and Yurii Yakovenko form the Nuclear Research
Institute in Kiev, Ukraine, are spending a month with Roscoe White,
Eric Fredrickson and Nikolai Gorelenkov, collaborating on the
interaction of high-energy particles with MHD modes in NSTX.

Leonid E. Zakharov gave a talk “Getting serious about fusion: ignited
spherical tokamaks and thei