JOINT THEORY/PST SEMINAR

Date/Time: April 24th (Thursday) 10:45am

Location: Theory Seminar Room (T169)

Speaker: Dr. Yu. Petrov, Prairie View A&M, Prairie View, TX

Title: Status of Rotamak program at Prairie View.

Abstract:

Two Rotamak devices, where the steady plasma current is driven by applied Rotating Magnetic Field (RMF), are being operated at Prairie View A&M University: one with a spherical and another with a cylindrical chamber. We discuss latest experiments on plasma reshaping in cylindrical Rotamak. During 40-ms plasma discharge, a pulse current is briefly fed to a magnetic coil located at the midplane (middle coil). The internal magnetic field is scanned with pick-up coils in almost all cross-section of plasma. The use of the middle coil allows switching from D-shape to doublet shape plasma and back. We also compare main results of experiments in spherical and cylindrical chambers, in particular, the structure of self-generated toroidal magnetic field. The evidence is presented that in Rotamak-ST case (with external toroidal magnetic field) the mechanism of RMF penetration is different from that in Rotamak-FRC (no external tor. field); it is more efficient due to whistler wave excitation.

Date/Time: April 17th (Thursday) 10:45am

Location: Theory Seminar Room (T169)

Speaker: Allen Boozer, Columbia University

Title: Perturbed Equilibria.

Abstract:

The response of plasmas to small external magnetic perturbations is an
important issue for both the tokamak and the stellarator programs.
Tokamak applications include (1) Control of magnetic field errors
(errors ten thousand times smaller than the main field can cause
disruptions). (2) Control of edge localized modes (ELMs) and resistive
wall modes (RWMs). (3) Determination of the level and the effects of
the toroidal torques produced by toroidal asymmetries. Stellarator
applications include: (1) Information useful for plasma and coil
design. (2) Specification of allowable construction tolerances. (3)
Assessment of intrinsic islands and magnetic surface quality. Ideal
MHD stability codes, such as DCON and CAS3D, give the fundamental
information, but appropriate postprocessors for these codes are
required for the applications. Jong-Kyu Park has written and
implemented IPEC, which is the required postprocessor for the DCON
code, and Carolin Nuehrenberg has appropriately modified the CAS3D
code. The methods and the results will be discussed with an emphasis
on the areas in which the fundamental understanding has been changed by
these new computational capabilities.

Date/Time: April 11th (Friday) 10:45am

Location: Theory Seminar Room (T169)

Speaker: Elena Belova, Princeton Plasma Physics Laboratory

Title: Numerical simulations of NBI driven GAE modes in NSTX.

Abstract:

Hybrid 3D code HYM is used to investigate beam ion effects on MHD modes in a NSTX, aiming at simulations of NSTX shots where chirping frequency GAE/CAE modes have been observed. The thermal plasma is modeled using the MHD equations, and full-orbit delta-f kinetic description is used for the beam ions. The simulations show that for large neutral beam injection velocities and strong anisotropy in the pitch-angle distribution, many Alfven modes are excited. Unstable GAEs modes for 2<n<7 and weakly unstable CAE for n>7 are observed. Scaling of the growth rate of GAE mode with beam ion density is stronger than linear due to significant modification of plasma equilibrium profiles. Profile modification is due to self-consistent beam ion effects, and it has indirect effect on the stability. It was demonstrated that phase velocity of the unstable GAE mode has opposite sign compared to the beam injection velocity, and the resonant particles satisfy Doppler-shifted cyclotron resonant conditions. Dependence of the growth rate on dissipation parameters is studied. Nonlinear simulations show that the GAE instability saturates at low amplitude.

Date/Time: April 10th (Thursday) 10:45am

Location: Theory Seminar Room (T169)

Speaker: E. A. Startsev, Princeton Plasma Physics Laboratory

Title: Dynamics of electromagnetic two-stream interaction processes during longitudinal and transverse compression of an intense ion beam pulse propagating through background plasma.

Abstract:

To achieve maximum energy density charged particle beam must be compressed radially and longitudinally while its space-charge is neutralized by background plasma. The beam propagating in plasma is subject to electrostatic two-stream instability and electromagnetic Weibel instability. The electrostatic two-stream instability may lead to longitudinal bunching of the beam pulse and eventual longitudinal beam heating. Consequently, this could degrade longitudinal compression of the beam. Similarly, the electromagnetic Weibel instability may cause transverse filamentation of the beam, which may degrade transverse compression. To achieve stronger transverse focusing, it has been proposed to pass the beam through a strong solenoidal magnetic field. The solenoidal magnetic field can extend long distance away from the solenoid into the neutralizing plasma where the beam is compressed longitudinally. In this paper, we review how transverse and longitudinal compression changes the dynamics of two-stream and Weibel instabilities. We also discuss how these instabilities are modified by the solenoidal magnetic field.

Date/Time: April 4th (Friday) 10:45am

Location: Theory Seminar Room (T169)

Speaker: Christine Nguyen, C.E.A. Cadarache, France

Title: Low frequency modes triggered by fast particles in Tore-Supra.

Abstract:

The recent observation of MHD modes destabilized by fast supra-thermal particles in Tore-Supra has triggered a theoretical and experimental program designed to model those modes and understand their interaction with fast particle populations. For the success of burning plasmas, this study is of major importance to understand and control the confinement of alpha particles, whose resonant interaction with MHD modes may be detrimental for plasma yields. Two particular types of instabilities are currently investigated in Tore-Supra, following the experimental observations of low-frequency modes: the Beta Alfvén Eigenmode (BAE) and the electron fishbone.
Using a variational formalism coupled to a Fourier mode decomposition, and a gyrokinetic-MHD model for the plasma, the BAE dispersion has been derived and found consistent with F. Zonca et al. earlier derivation in the ballooning representation. In particular, this derivation allows to identify Tore-Supra observed modes as BAEs despite of the ambiguity linked to the BAE/GAM degeneracy, and to calculate a threshold for BAE destabilization. This calculated threshold has been compared with experimental observations, using in particular the Monte-Carlo code PION to access the fast ion distribution function. This comparison validates the consideration of ion Landau damping of the mode sidebands as a main damping process.
Experiments designed to study electron fishbone modes have been conducted, and display a puzzling behavior that provides some insight in the interaction of those modes with the supra-thermal electron distribution. Some modes, observed with the ECE diagnostic are consistent with the traditional characteristics of precessional fishbone modes, and appear to be characterized by frequency jumps associated with a fast electron phase space redistribution. This redistribution of fast electrons is likely to lead to a modification of the q profile, which may have an impact on transport.

Speaker: Prof. Adam Burrows, Princeton University Astrophysics Dept.

Time: March 20, 10:45am

Place: Theory Conference Room

Title:

Abstract:

To address the theoretical supernova explosion problem
with physical fidelity requires the development and use of
sophisticated numerical radiation/hydrodynamic codes that
simulate the multi-dimensional flow in a variety of Mach-number
regimes. Though the latest simulations incorporate rotation,
multi-group radiative transfer, and magnetic fields, they
are not yet general-relativistic, do not solve the Boltzmann
equation in its full multi-D context, and are not fully 3D in space.
One must eventually do the calculations in six-dimensional phase
space (plus time), and such seven-dimensional calculations are
currently beyond reach. Nevertheless, there has been much recent
progress and this progress has been informed by numerical
experiments that will only get better in the next 3-5 years.
In this talk, I will discuss the latest physical ideas in the
theory of the mechanism of core-collapse supernovae and the
variety of results that have emerged from recent massive
computations. Moreover, I will motivate what more may need to be
done to solve in credible fashion the enigma of stellar death and
supernova explosion.

Date/Time: March 13th (Thursday) 10:45am

Location: Theory Seminar Room (T169)

Speaker: Gennady Shvets, University of Texas at Austin

Title: Filamentation of high-current beams in plasmas: physics and applications to Fast Ignition.

Abstract:

The filamentation (a.k.a Weibel) instability of high-current beams propagating in background plasmas is one of the most basic and long-studied collective plasma processes. The dynamics and energetics of its nonlinear saturation is important for both laboratory and astrophysical plasmas. The WI is likely to play an important role in the Fast Ignitor scenario because it may result in the collective energy loss of a relativistic electron beam in both coronal and core plasma regions. Collisionless WI has been suggested as an important mechanism for relativistic collisionless shock formation in gamma ray bursts. In this talk I will describe the theoretical framework predicting the long-term evolution, structure, and coalescence energetics of beam/return current filaments during the Weibel instability of an electron beam in a collisionless plasma is developed. I will emphasize the strongly nonlinear stage of the instability, during which the beam density of filaments is compressed to the background plasma density, and the ambient plasma is fully evacuated. Our analytic and numerical results demonstrate that the beam filaments can carry super-Alfvenic currents by assuming current and density profiles similar to the Hammer-Rostoker equilibrium. This has profound

implications for the long-term evolution of the magnetic field and beam current and explains the long-standing puzzle: why magnetic field energy initially increases, but eventually decreases during the collisionless Weibel instability.

In collaboration with O. Polomarov, A. Sefkow, and I. Kaganovich. Supported by the US DOE grant DE-FG02-05ER54840.

Speaker: Dr. Jeronimo Garcia

Time: March 6 (2:00 PM)

Place: PPPL Display Wall Room

Title: ITER steady-state analysis with the CRONOS suite of codes

Abstract:

Integrated simulations with the CRONOS suite of codes, developed at CEA-Cadarache, are used to study the physics involved in the Internal Transport Barrier (ITB) sustainment and to identify the main obstacles for the establishment of a steady-state scenario in ITER. It is shown that any current drive inside the ITB leads to a progressive shrinking and disappearence of the barrier (known as the current misalignment effect) which means that Neutral Beam Current Drive, which is naturally localized in the central part of the plasma, proves to be of little use in these scenarios. In contrast, a pure Radio Frequency scenario is proposed showing that it provides a solution of principle to the current alignment problem. The main feature of this scenario is that there is a strong minimum negative magnetic shear to steadily sustain the ITB for 3000s, below which low performance inductive scenarios are recovered. The actual design of the ECRH/ECCD system in ITER can provide such a negative magnetic shear, leading to a clear dependence of the temperature gradient (with a well defined threshold) on the Pech/ parameter. The threshold obtained can be characterized as a second order phase transition as it has been done previously for the ITB formation of other completely different fusion devices as, e.g. the Large Helical Device (LHD) The extension of these scenarios to the future fusion demonstration commercial reactor (usually called DEMO) will be considered.

Speaker: Dr. Kathy Yelick, NERSC Director

Title: The Future of NERSC

Date: February 25 (Monday) at 3:00 PM

Place: PPPL Display Wall Room

Abstract:

NERSC provides production computing to the DOE Office of Science community
with over 3000 users working on 300 distinct projects. The NERSC systems
include the Cray XT4 system, know as Franklin, which has over 100 Teraflops
of peak performance and a sustained application performance of 19 Teraflops.
In addition, NERSC provides archival and online storage, cluster computing,
and application consulting services. I will highlight some of the science activities performed at NERSC and describe two new strategic directions for the next 10 years. The first significantly expands the support for analysis of large scientific data sets through a combination of hardware, software, and service activities. The second is to re-think high end computing system design from a power-efficiency perspective, which I believe is necessary to reach the target of Exascale computing. I will talk about initial efforts in these two areas, both using an application-driven approach to finding a general solution to these problems.

Date/Time: February 14th (Thursday) 10:45am

Location: Theory Seminar Room (T169)

Speaker: S. C. Jardin, Princeton Plasma Physics Laboratory

Title: The M3D-C1 Approach to Calculating Two-Fluid Equilibrium, Stability, and Magnetic Reconnection in Magnetized Plasmas

Abstract:

The M3D code [1] has proven itself to be an invaluable tool for the simulation and understanding of global nonlinear phenomena in magnetic fusion confinement devices. However, the structure of M3D is not optimal for computing in regimes where two-fluid (2F) effects dominate, or for times that are very long compared to the Alfven transit time. We have built upon many of the favorable features of the M3D approach to construct the M3D-C1 code [2], which is based on high-order, compact conformal finite elements with C1 continuity on an unstructured adaptive grid. The efficient split-implicit time advance is shown to be closely related to the ideal MHD energy principle, and allows time steps several orders of magnitude in excess of the Courant condition based on the Alfven or whistler waves. The full model consists of 8 3D scalar variables. Nontrivial, energy conserving, subsets of the full equations exist including 2-variable 3D reduced MHD which is a toroidal generalization of [3] and a 4-variable 3D reduced model which is a toroidal generalization of [4]. The structure of the code makes linear calculations exceptionally efficient. Illustrative results in 2F toroidal equilibrium, 3D linear stability and 2F magnetic reconnection are given. Future capabilities including a surrounding resistive wall and a scalable full 3D nonlinear time evolution are discussed.

Speaker: Dr. Scott Klasky

Date and Time: February, 12 (11:00AM)

Place: Display Wall Room (PPPL)

Title: End-to-end computing for petascale simulations.

Abstract: ORNL has embraced leadership-class computing, and has quickly
become one of the leading institutions for the DOE and NSF for large
scale computations. One of the most challenging problems associated with
running on the large computers is dealing with the huge amounts of data
that is generated. Researchers are quickly becoming overwhelmed with the
daunting task, not only of running their simulations on 100K processors,
but also of efficiently extracting and transporting the many TB's of
data generated by the simulations, and analyzing this data, and share it
with their colleagues in a timely manner. The impact of these challenges
and the overall time to solution is only growing as computers are
getting faster. In order to help address these challenges we have been
developing a suite of software solutions, which are gaining acceptance
by the largest codes that are part of the DOE open science.

Our suite of software solutions includes new API's (ADIOS) that allow
for both MPI-IO and asynchronous I/O through Remote Direct Memory Access
(RDMA), workflow automation using the Kepler workflow package, fast wide
area data transfers, and dashboards that combine data management,
provenance management, and data analysis for monitoring complex
simulations. In this talk I introduce these solutions and will show how
this technology is being used in several fusion SciDAC projects.

Date/Time: February 7th (Thursday) 10:45am

John A. Krommes
Princeton University, Plasma Physics Laboratory

Location: Theory Seminar Room (T169)

The remarkable similarity between the
scaling of kurtosis with squared skewness
for TORPEX density fluctuations
and sea-surface temperature fluctuations:
Baby steps toward a theory

The striking similarity between certain higher-order statistics of drift-interchange plasma turbulence in the TORPEX device [B. Labit et al., Phys. Rev. Lett. 98, 255002 (2007)] and sea-surface temperature fluctuations [P. Sura and P. D. Sardeshmukh, J. Phys. Oceanogr. (2007), in press] (SS) is described. A successful nonlinear Langevin theory due to SS is reviewed, then generalized to include linear wave propagation; it is shown to make a reasonable prediction for the shape of the kurtosis versus skewness curve for TORPEX. The relevance of these somewhat naive calculations will be discussed, and future research directions will be indicated.

Frank Jenko (from Garching) will give a seminar on

"Nonlinear Gyrokinetic Simulations for the W7-X and NCSX Stellarators"

this Thursday, Jan. 31, at 10 a.m. (earlier than the usual theory
seminar time), in the Theory Seminar Room.

Speaker: Prof. Jinchao Xu
Pennsylvania State University

Date and time: Feb. 1, 2008

Place: Display Wall Room

Title: Nonconforming Finite Elements for High Order
PDE Systems and Relevant Algebraic Solvers

Abstract:

In this talk, I will first discuss a few issues on finite element
methods for high order partial differential equations including
application of conforming elements (such as $C^1$ elements for fourth
order problems), possible danger in reducing a high order PDE into a
system of lower order PDEs, and the design of minimal order
nonconforming elements for 2m-th order PDEs in $R^n$ (for any $n \ge m
$). I will also talk about a few recent results on
optimal and practical solvers for Maxwell equations and Navier-Stokes
equations.

Speaker: Dr. Mark Adams
Columbia University/PPPL

Place: Display Wall Room

Time: January, 25 10:30 AM

Title: Toward optimal multigrid algebraic solvers in magnetohydrodynamics
simulations of fusion plasmas

Abstract:

Magnetohydrodynamic simulations of tokamak fusion plasmas exhibit a
large separation of temporal scales. To overcome the temporal
stiffness associated with the fast compressive and Alfven waves in
single-fluid resistive MHD, we consider the development of optimal
implicit algorithms. We strive to achieve "textbook" multigrid
efficiency in which the set of nonlinear equations is solved to
discretization accuracy at each time step, with a cost equivalent
to a few (less than 10) residual calculations (or work units).
We present results from a few canonical MHD problems: magnetic
reconnection in 2D and in the presence of a strong guide field.

Date/Time: January 24th (Thursday) 10:45am

Location: Theory Seminar Room (T169)

Speaker: Hong Qin, Princeton Plasma Physics Laboratory

Title: Variational Symplectic Integrator for the Guiding Center Motion of Charged Particles for Long Time Simulations in General Magnetic Fields

Abstract:

A variational symplectic integrator for the guiding center motion of charged particles in general magnetic fields is developed for long time simulation studies of magnetized plasmas. Instead of discretizing the differential equations of the guiding center motion, the action of the guiding center motion is discretized and minimized to obtain the iteration rules for advancing the dynamics. The variational symplectic integrator conserves exactly a discrete Lagrangian symplectic structure, and has better numerical properties over long integration time, compared with standard integrators, such as the standard and variable time-step 4th order Runge-Kutta methods. Standard integrators only guarantee the error to be small in each time-step. The errors at different time-steps often accumulate coherently, and result in a large error over a large number of time-steps. The symplectic integrator conserves the symplectic structure exactly, and guarantees that the energy error is bounded by a small number for all the time-steps. Numerical examples with more than 25 million time-steps are given to demonstrate the superiority of the variational symplectic integrator. This significant improvement in long term simulation capability of gyrokinetics is a direct, otherwise-impossible result of the geometric formulation of the gyrokinetic theory using the modern language of differential geometry [H. Qin, et al, Physics of Plasmas 14, 056110 (2007)].

Date/Time: January 10th (Thursday) 10:45am

Location: Theory Seminar Room (T169)

Speaker: Leonid E. Zakharov, Princeton Plasma Physics Laboratory

Title: The kink mode during the disruptions in tokamaks

Abstract:

The talk explains the locked $m/n=1/1$ kink mode during the vertical
disruption event when the plasma has an electrical contact with the
plasma facing conducting surfaces. It is shown that the kink
perturbation can be in equilibrium state even with a stable safety
factor $q > 1$, if the halo currents, excited by the kink mode, can
flow through the conducting structure. This suggests a new explanation
of the toroidal asymmetry in magnetic measurements and so-called
sideway forces on the in-vessel components
during the disruption event.