2010 Theory Department Highlights

2005 Highlights

2006 Highlights

2007 Highlights

2008 Highlights

2009 Highlights

MARCH 5, 2010

THEORY GROUP

D. McCune organized a meeting of interested people at PPPL on March 2 to solicit feedback on the Fusion Simulation Program (FSP) “Whole Device Modeling" Science Driver and roadmap.  The meeting was attended by about 15 people, including leaders from both the theoretical and experimental community.  There were brief presentations by McCune and by A. Reiman, intermingled with about 90 minutes of discussion.  The relationship between the FSP and the base theory program was one focus of discussion, as was the relationship between the FSP and the SCIDAC projects supporting component development. There appeared to be wide recognition of the desirability of whole device modeling, going well beyond current (e.g. TRANSP-like) capabilities. One concern expressed at the meeting was whether FSP whole device modeling capabilities will be adequately supported for production use. Another focus of discussion was the possibility of replacing the 2D Grad-Shafranov solver in 1.5 D transport codes with a 3D equilibrium code. It was pointed out that this was envisioned from the outset by the developers of 1.5 D transport codes in the 1970’s, and that they laid much of the theoretical foundation needed for this.
 
At a Boundary Plasma Science Focus Group meeting on March 3, D. Stotler presented the roadmap for the Fusion Simulation Program (FSP) “Integrated Boundary Layer (SOL), Divertor, Plasma Wall Interactions Science Driver", and he led a discussion of this subject.  One focus of the discussion was the relative priority of items on the roadmap. Another focus was the limited scope of work contemplated on plasma-material interactions, and the need for more research on that subject.
 
On March 5, there was a CPPG Workshop/Tutorial on the Allinea DDT parallel debugger.  The presenters were David Maples – V.P. Americas, and Dr. David Lecomber – CTO and Head of Development – Allinea Software.
 
 
COMPUTATIONAL PLASMA PHYSICS GROUP

Dr. Francesca Poli, from the University of Warwick, UK, completed a two week visit at PPPL to work with Stephane Ethier on the coupling of a synthetic diagnostic of high-k coherent wave scattering with results of large-scale ETG simulations carried out with PPPL's GTS code. The synthetic diagnostic is being developed to reproduce as closely as possible the conditions of the actual high-k tangential scattering diagnostic on NSTX. Several NSTX shots have been identified for direct comparison between actual and simulated measurements.
 
Stephane Ethier traveled to Berkeley, CA, to serve on the NERSC Policy Board. The Board advises NERSC on strategic planning and directions for the Supercomputing Center in the coming years. Ethier also met with collaborators of the LBNL Future Technologies Group regarding the optimization of the gyrokinetic PIC algorithm on new computing technologies, such as many-core processors and GPUs.
 
Mr. David Maples (onsite) and Dr. David Lecomber (in UK) of Allinea Software gave a presentation and demo on the capabilities of the DDT debugging software. The DDT software is capable of helping users debug programs written in fortran, C and C++ on massively parallel computers. The software is currently available on the NERSC, ORNL, and Argonne National Lab computers and it has been used for debugging a parallel application on the ORNL Jaguar machine using 220000 cores.

February 26, 2010

COMPUTATIONAL PLASMA PHYSICS GROUP

We have succeeded for the first time in running the M3D massively parallel extended MHD code on over 10,000 processors at the National Energy Research Scientific Computing Center (NERSC).  Favorable scaling was obtained by implementing the Algebraic Multigrid preconditioned GMRES linear solver and RCM matrix reordering techniques. We have performed fully 3D weak scaling studies and found that the wall clock time increases only by 60% as we increase the problem size by 5 in going from 2448 to 12288 processors

January 29,2010

THEORY GROUP

In the DOE-SC announcement this week of the INCITE 2010 awards, scientists from the PPPL Theory Department are the lead PI's for two of the projects in the Fusion Energy Sciences (FES) area and others are co-investigators on two other FES projects that are led by PI's from collaborating institutions. The INCITE (Innovative and Novel Computational Impact on Theory and Experiment) program promotes cutting-edge research that can only be conducted with state-of-the-art super-computers. PPPL recipients include: Weixing Wang as PI of the new project on "Investigation of Multi-Scale Transport Physics of Fusion Experiments Using Global Gyrokinetic Turbulence Simulations" with Stephane Ethier and Wei-li Lee as co-Investigators; William Tang as PI of the renewed project on " High Resolution Simulations of Global Microturbulence" with Stephane Ethier as co-Investigator; Greg Hammett and David Mikkelsen as co-Investigators on the renewed project led by Bill Nevins of LLNL on "Validation of Plasma Microturbulence Simulations for Finite-Beta Fusion Experiments;" and Stephane Ethier as co-Investigator on the renewed project led by Pat Diamond of UCSD on "Verification and Validation of Petascale Simulations of Turbulent Transport in Fusion Plasmas." Detailed descriptions of these projects are located on the INCITE 2010 Awards Announcement web-site:  http://www.er.doe.gov/ascr/incite/index.html .
 
 
COMPUTATIONAL PLASMA PHYSICS GROUP

In a successful modernization effort, the source code server for the PPPL TRANSP codebase was upgraded from cvs to svn. Using the public domain cvs2svn software, the code’s entire cvs collaborative development history, going back to February 1993, was preserved, with 13,180 discreet code modification commit actions recorded. Individual TRANSP developers are able to transition from cvs to svn with a modest effort. The svn service offers a number of improvements in convenience and functionality. For example, certain shareable source components, such as Plasma State, can be imported from TRANSP into other projects using “svn external”. Access requests are handled on a case by case basis. For details see the TRANSP website https://w3.pppl.gov/TRANSP <https://w3.pppl.gov/TRANSP> , follow the “Source Code” link.
 
Dr. Samuel Lazerson, from the University of Alaska, presented a CPPG seminar on “Numerical Techniques in Dusty Plasmas”.  He described The Dust Electron Neutral Ion Self-consistent Integration Scheme (DENISIS) code, a multi-fluid code capable of performing simulations of large-scale dusty plasma phenomena and treating their partially ionized nature. He described the comprehensive validation and testing methodology used for these codes, including linear wave mode propagation, tests of source terms, and convergence testing.

January 22, 2010

THEORY GROUP

An article entitled, "Transport of Parallel Momentum by Toroidal Ion Temperature Gradient Instability near Marginality" by E.S. Yoon and T.S. Hahm has been accepted for publication in the special issue of Nuclear Fusion for the International Workshop for H-mode Physics and Transport Barriers. The abstract reads as follows: The turbulent angular momentum flux carried by ions resonant with toroidal ion temperature gradient (ITG) instability is calculated via quasilinear calculation using the phase-space conserving gyrokinetic equation in the laboratory frame. The results near ITG marginality indicate that the inward turbulent equipartition (TEP) momentum pinch [Hahm T.S. et. al., 2007 Phys. Plasmas 14 072302] remains as the most robust part of the pinch. In addition, the ion temperature gradient driven momentum flux is inward for typical parameters, while the density gradient driven momentum flux is outward as in the previous kinetic result in slab geometry [Diamond P.H. et. al., 2008 Phys. Plasmas 15 012303].
 
On January 18-20, Leonid E. Zakharov attended the NIFS-CRC International Symposium on Plasma Surface Interactions, CRC-NIFS, Toki, Gifu, Japan and gave a talk, "LiWall Fusion - the new concept of magnetic fusion".
 
 
COMPUTATIONAL PLASMA PHYSICS GROUP

The ray tracing code GENRAY has been installed in TRANSP. GENRAY is a code developed and maintained by Dr. Robert Harvey and others at CompX in San Diego. TRANSP can now perform ECRH heating and current drive calculations with GENRAY. The TRANSP documentation page at https://w3.pppl.gov/~pshare/help/transp.htm contains instructions for incorporating the GENRAY capability in a TRANSP run.
 

January 15, 2010

THEORY GROUP

Daren Stotler and Ravi Samtaney participated in the Proto-FSP Frameworks Workshop, January 11-13, at Oak Ridge National Laboratory on behalf of the Center for Plasma Edge Simulation. Each day of the workshop focused on the simulation framework being used by one of the three SciDAC prototype Fusion Simulation Program projects, including introductory material, demonstrations, and hands-on exercises.
 
Joshua Breslau attended Department of Energy and White House ceremonies in Washington on January 13 to receive a 2009 Presidential Early Career Award for Scientists and Engineers (PECASE).  The citation reads, "For playing an essential role in the development of the massively parallel fusion magnetohydrodynamics code, M3D, and for original and unique applications of this code to the nonlinear dynamics of tokamak instabilities."
 
 
COMPUTATIONAL PLASMA PHYSICS GROUP

S. Jardin, D. McCune, and R. Samtaney, together with D. Stotler (PPPL Theory) attended the three-day Workshop of the Three Proto-FSP projects at ORNL. The three projects, CPES: Center for Plasma Edge Simulation, CSWIM: The Center for Simulation of Wave Interaction with MHD, and FACETS: Framework Application for Core-Edge Transport Simulation, each spent one day giving a tutorial on how to use the software and frameworks developed within their projects. It was a hands-on workshop where the attendees gained actual experience in using each of the different frameworks to solve a fusion simulation problem. Each of the three projects has at least one member of the PPPL Computational Plasma Physics group as a participant.

January 8, 2010

An article entitled, "Nonlinear Simulation of Toroidal Alfv\'en Eigenmode with Source and Sink" by J. Lang, G.-Y. Fu, and Y. Chen (University of Colorado at Boulder) has submitted to Phys. Plasmas. The abstract reads as follows: Kinetic/MHD hybrid simulations are carried out to investigate the nonlinear dynamics of energetic particle-driven Toroidal Alfv\'en eigenmode with collision and source/sink. For cases well above marginal stability, the mode saturation is approximately steady state with finite collision frequency. The calculated scaling of saturation level with collision frequency agrees well with analytic theory. For cases near marginal stability at low collision rates, the mode saturation exhibits pulsation behavior with frequency chirps up and down.

CPPG Group

The NTCC now includes a "data component", based on the Plasma State software, which allows SciDAC Fusion Simulation Program (FSP) prototype codes to access output data of TRANSP simulations. This can be a means for FSP prototypes such as SWIM, FACETS, and CPES to access experimental data such as temperature and density profiles, as well as heating and current drive sources and fast ion pressures obtained from the experimental analysis.

Dr. Alexander Pletzer, from TechX, Boulder, CO, visited this week and presented a CPPG seminar titled: "Chompst: Marrying the best of Chombo and PETSc for solving "hard" Multi-grid problems". Chompst is a Small Business Innovation & Research (SBIR) project funded by DoE to address the need for exposing the large collection of PETSc iterative solvers from within the adaptive mesh refinement (AMR) tool Chombo. Standard multi-grid is known to be ill-suited for some elliptic operators, notably some Helmholtz operators, and thermal diffusion problems where the heat conductivity is highly anisotropic and/or exhibits strong spatial dependency. In his talk, he showed how the ability to go beyond multi-grid relaxation schemes by exposing PETSc's rich set of iterative solvers will allow Chombo users to solve some of these hard problems.