K. H. Burrell

General Atomics, San Diego, California, United States

Are you K. H. Burrell?

Claim your profile

Publications (652)1024.84 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: The geodesic acoustic mode (GAM), a coherent form of the zonal flow, plays a critical role in turbulence regulation and cross-magnetic-field transport. In the DIII-D tokamak, unique information on multi-field characteristics and radial structure of eigenmode GAMs has been measured. Two simultaneous and distinct, radially overlapping eigenmode GAMs (i.e., constant frequency vs. radius) have been observed in the poloidal E×B flow in L-mode plasmas. As the plasma transitions from an L-mode to an Ohmic regime, one of these eigenmode GAMs becomes a continuum GAM (frequency responds to local parameters), while the second decays below the noise level. The eigenmode GAMs occupy a radial range of ρ = 0.6-0.8 and 0.75-0.95, respectively. In addition, oscillations at the GAM frequency are observed for the first time in multiple plasma parameters, including ne, Te, and Bθ. The magnitude of T~e/Te at the GAM frequency (the magnitude is similar to that of n~e/ne) and measured ne-Te cross-phase (~140° at the GAM frequency) together indicate that the GAM pressure perturbation is not determined solely by n~e. The magnetic GAM behavior, a feature only rarely reported, is significantly stronger (×18) on the high-field side of the tokamak, suggesting an anti-ballooning nature. Finally, the GAM is also observed to directly modify intermediate-wavenumber n~e levels (kρs ~ 1.1). The simultaneous temperature, density, flow fluctuations, density-temperature cross-phase, and magnetic behavior present a new perspective on the underlying physics of the GAM.
    No preview · Article · Sep 2013 · Physics of Plasmas
  • [Show abstract] [Hide abstract]
    ABSTRACT: Charge-exchange spectroscopy is a powerful diagnostic tool for determining ion temperatures, densities and rotational velocities in tokamak plasmas. This technique depends on detailed understanding of the atomic physics processes that affect the measured apparent velocities with respect to the true ion rotational velocities. These atomic effects are mainly due to energy dependence of the charge-exchange cross-sections, and in the case of poloidal velocities, due to gyro-motion of the ion during the finite lifetime of the excited states. Accurate lifetimes are necessary for correct interpretation of measured poloidal velocities, specially for high density plasma regimes on machines such as ITER, where l-mixing effects must be taken into account. In this work, a full nl-resolved atomic collisional radiative model coupled with a full kinetic calculation that includes the effects of electric and magnetic fields on the ion gyro-motion is presented for the first time. The model directly calculates from atomic physics first principles the excited state lifetimes that are necessary to evaluate the gyro-orbit effects. It is shown that even for low density plasmas where l-mixing effects are unimportant and coronal conditions can be assumed, the nl-resolved model is necessary for an accurate description of the gyro-motion effects to determine poloidal velocities. This solution shows good agreement when compared to three QH-mode shots on DIII-D, which contain a wide range of toroidal velocities and high ion temperatures where greater atomic corrections are needed. The velocities obtained from the model are compared to experimental velocities determined from co- and counter-injection of neutral beams on DIII-D.
    No preview · Article · Aug 2013 · Nuclear Fusion
  • [Show abstract] [Hide abstract]
    ABSTRACT: A series of carefully designed validation experiments conducted on DIII-D to rigorously test gyrofluid and gyrokinetic predictions of transport and turbulence stiffness in both the ion and electron channels have provided an improved assessment of the experimental fidelity of those models over a range of plasma parameters. The first set of experiments conducted was designed to test predictions of H-mode core transport stiffness at fixed pedestal density and temperature. In low triangularity lower single null plasmas, a factor of 3 variation in neutral beam injection (NBI) heating was obtained, with modest changes to pedestal conditions that slowly increased with applied heating. The measurements and trends with increased NBI heating at both low and high injected torque are generally well-reproduced by the quasilinear trapped gyro-Landau fluid (TGLF) transport model at the lowest heating levels, but with decreasing fidelity (particularly in the electron profiles) as the heating power is increased. Complementing these global stiffness studies, a second set of experiments was performed to quantify the relationship between the local electron energy flux Qe and electron temperature gradient by varying the deposition profile of electron cyclotron heating about a specified reference radius in low density, low current L-mode plasmas. Modelling of these experiments using both the TGLF model and the nonlinear gyrokinetic GYRO code yields systematic underpredictions of the measured fluxes and fluctuation levels.
    No preview · Article · Jul 2013 · Nuclear Fusion
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The first real-time profile control experiments integrating magnetic and kinetic variables were performed on DIII-D in view of regulating and extrapolating advanced tokamak scenarios to steady-state devices and burning plasma experiments. Device-specific, control-oriented models were obtained from experimental data using a generic two-time- scale method that was validated on JET, JT-60U and DIII-D under the framework of the International Tokamak Physics Activity for Integrated Operation Scenarios (Moreau et al 2011 Nucl. Fusion 51 063009). On DIII-D, these data-driven models were used to synthesize integrated magnetic and kinetic profile controllers. The neutral beam injection (NBI), electron cyclotron current drive (ECCD) systems and ohmic coil provided the heating and current drive (H& CD) sources. The first control actuator was the plasma surface loop voltage (i. e. the ohmic coil), and the available beamlines and gyrotrons were grouped to form five additional H& CD actuators: co-current on-axis NBI, co-current off-axis NBI, counter-current NBI, balanced NBI and total ECCD power from all gyrotrons (with off-axis current deposition). Successful closed-loop experiments showing the control of (a) the poloidal flux profile, Psi(x), (b) the poloidal flux profile together with the normalized pressure parameter, beta(N), and (c) the inverse of the safety factor profile, (i) over bar (x) = 1/q(x), are described.
    Full-text · Article · Jun 2013 · Nuclear Fusion
  • [Show abstract] [Hide abstract]
    ABSTRACT: A critical gradient for long wavelength (kθρs≲0.4) electron temperature fluctuations has been observed in an experiment in the DIII-D tokamak [J. L. Luxon, Nucl. Fusion 42, 614 (2002)], where below a threshold value of LTe−1 = |∇Te|/Te electron temperature fluctuations are constant and above they steadily increase. Above the critical gradient, the electron heat flux inferred by power balance also increases rapidly. Critical gradients are a predicted attribute of turbulence arising from linear instabilities and are thought to be related to transport stiffness. The presented results are the first direct, systematic demonstration of critical gradient behavior in turbulence measurements in a tokamak. The experiment was performed by changing the deposition location of electron cyclotron heating shot-to-shot to locally scan LTe−1 at r/a = 0.6 in L-mode plasmas; rotation was also varied by changing the momentum input from neutral beam injection. Temperature fluctuations were measured with a correlation electron cyclotron emission (CECE) radiometry system. In addition to the CECE measurements, an array of turbulence measurements were acquired to characterize fluctuations in multiple fields and at multiple scales as LTe−1 and rotation were modified: long wavelength (kθρs≲0.5) density fluctuations were acquired with beam emission spectroscopy, the phase angle between electron temperature and density fluctuations was measured by coupling the CECE system and a reflectometer, intermediate scale (kθρs ∼ 0.8) density fluctuations were measured with a Doppler backscattering (DBS) system, and low frequency flows were also measured with DBS. The accumulated measurements and trends constrain identification of the instability responsible for the observed critical gradient to the ∇Te-driven trapped electron mode.
    No preview · Article · May 2013 · Physics of Plasmas
  • B.A. Grierson · K.H. Burrell · W.M. Solomon · R.V. Budny · J. Candy
    [Show abstract] [Hide abstract]
    ABSTRACT: In tokamak plasmas with low levels of toroidal rotation, the radial electric field Er is a combination of pressure gradient and toroidal and poloidal rotation components, all having similar magnitudes. In order to assess the validity of neoclassical poloidal rotation theory for determining the poloidal rotation contribution to Er, Dα emission from neutral beam heated tokamak discharges in DIII-D (Luxon 2002 Nucl. Fusion 42 614) has been evaluated in a sequence of low torque (electron cyclotron resonance heating and balanced diagnostic neutral beam pulse) discharges to determine the local deuterium toroidal rotation velocity. By invoking the radial force balance relation the deuterium poloidal rotation can be inferred. It is found that the deuterium poloidal flow exceeds the neoclassical value in plasmas with collisionality , being more ion diamagnetic, and with a stronger dependence on collisionality than neoclassical theory predicts. At low toroidal rotation, the poloidal rotation contribution to the radial electric field and its shear is significant. The effect of anomalous levels of poloidal rotation on the radial electric field and cross-field heat transport is investigated for ITER parameters.
    No preview · Article · May 2013 · Nuclear Fusion
  • [Show abstract] [Hide abstract]
    ABSTRACT: A critical gradient threshold has been observed for the first time in a systematic, controlled experiment for a locally measured turbulent quantity in the core of a confined high-temperature plasma. In an experiment in the DIII-D tokamak where L_{T_{e}}^{-1}=|∇T_{e}|/T_{e} and toroidal rotation were varied, long wavelength (k_{θ}ρ_{s}≲0.4) electron temperature fluctuations exhibit a threshold in L_{T_{e}}^{-1}: below, they change little; above, they steadily increase. The increase in δT_{e}/T_{e} is concurrent with increased electron heat flux and transport stiffness. Observations were insensitive to rotation. Accumulated evidence strongly enforces the identification of the experimentally observed threshold with ∇T_{e}-driven trapped electron mode turbulence.
    No preview · Article · Jan 2013 · Physical Review Letters
  • [Show abstract] [Hide abstract]
    ABSTRACT: DIII-D is using its flexibility and diagnostics to address the critical science required to enable next step fusion devices. Operating scenarios for ITER have been adapted to low torque and are now being optimized for transport. Three ELM mitigation scenarios have been developed to near-ITER parameters. New control techniques are managing the most challenging plasma instabilities. Disruption mitigation tools show promising dissipation strategies for runaway electrons and heat load. An off axis neutral beam upgrade has enabled sustainment of high βN capable steady state regimes. Divertor research is identifying the challenge, physics and candidate solutions for handling the hot plasma exhaust with notable progress in heat flux reduction using the snowflake configuration. This work is helping optimize design choices and prepare the scientific tools for operation in ITER, and resolve key elements of the plasma configuration and divertor solution for an FNSF.
    No preview · Conference Paper · Jan 2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: A database of toroidal momentum transport on five tokamaks, Alcator C-Mod, DIII-D, JET, NSTX and JT-60U, has been constructed under a wide range of conditions in order to understand the characteristics of toroidal momentum transport coefficients, namely the toroidal momentum diffusivity (chi(phi)) and the pinch velocity (V-pinch). Through an inter-machine comparison, the similarities and differences in the properties of chi(phi) and V-pinch among the machines have been clarified. Parametric dependences of these momentum transport coefficients have been investigated over a wide range of plasma parameters taking advantage of the different operation regimes in machines. The approach offers insights into the parametric dependences as follows. The toroidal momentum diffusivity (chi(phi)) generally increases with increasing heat diffusivity (chi(i)). The correlation is observed over a wide range of chi(phi), covering roughly two orders of magnitude, and within each of the machines over the whole radius. Through the inter-machine comparison, it is found that chi(phi) becomes larger in the outer region of the plasma. Also observed is a general trend for V-pinch in tokamaks; the inward pinch velocity (-V-pinch) increases with increasing chi(phi). The results that are commonly observed in machines will support a toroidal rotation prediction in future devices. On the other hand, differences among machines have been observed. The toroidal momentum diffusivity, chi(phi), is larger than or equal to chi(i) in JET and JT-60U; on the other hand, chi(phi) is smaller than or equal to chi(i) in NSTX, DIII-D and Alcator C-Mod. In DIII-D, the ratio -RVpinch/chi(phi) at r/a = 0.5-0.6 is about 2, which is small compared with that in other tokamaks (-RVpinch/chi(phi) approximate to 5). Based on these different observations, parametric dependences of chi(phi)/chi(i), RVpinch/chi(phi) and chi(phi) have been investigated in H-mode plasmas. Across the dataset from all machines, the ratio chi(phi)/chi(i) tends to be larger in low nu(e)* at fixed T-e/T-i and rho(pol)*. An increase in chi(phi) is observed with decreasing n(e) and/or increasing T-e. The pinch number (-RVpinch/chi(phi)) is observed to increase with increasing R/L-ne at both q(95) = 5.5-7.2 and q(95) = 3.7-4.5. Here nu(e)*, nu(pol)*, T-e, T-i, R/L-ne and q(95) are, respectively, the normalized effective electron collision frequency, the normalized ion poloidal Larmor radius, the electron and ion temperatures, the inverse ratio of density scale length, L-ne, to the major radius, R, and the safety factor at the 95% flux surface.
    No preview · Article · Dec 2012 · Nuclear Fusion
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The first real-time profile control experiments integrating magnetic and kinetic variables were performed on DIII-D in view of regulating and extrapolating advanced tokamak scenarios to steady state devices and burning plasma experiments. Device-specific, control-oriented models were obtained from experimental data and these data-driven models were used to synthesize integrated magnetic and kinetic profile controllers. Closed-loop experiments were performed for the regulation of (a) the poloidal flux profile, Ψ(x), (b) the inverse of the safety factor profile, ι(x)=1/q(x), and (c) either the Ψ(x) profile or the ι(x) profile together with the normalized pressure parameter, β N . The neutral beam injection (NBI), electron cyclotron current drive (ECCD) systems and ohmic coils provided the heating and current drive (H&CD) sources. The first control actuator was the plasma surface loop voltage or current (i. e. the ohmic coil), and the available beamlines and gyrotrons were grouped to form five additional H&CD actuators: co-current on-axis NBI, co-current off-axis NBI, counter-current NBI, balanced NBI and total ECCD power from all gyrotrons (with off-axis current deposition). The control method was also applied on simulated ITER discharges using a simplified transport code (METIS).
    Full-text · Conference Paper · Oct 2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: The need for a validated predictive capability of turbulent transport in ITER is now widely recognized. However, to date most validation studies of nonlinear codes such as GYRO (Candy and Waltz 2003 J. Comput. Phys. 186 545) have focused upon low power L-mode discharges, which have significant differences in key dimensionless parameters such as ρ* = ρs/a from more ITER-relevant H-mode discharges. In order to begin addressing this gap, comparisons of the turbulent transport and fluctuations predicted by nonlinear GYRO simulations for a number of DIII-D (Luxon 2002 Nucl. Fusion 42 614) H-mode discharges to power balance analyses and experimental measurements are presented. The results of two H-mode studies are presented in this paper, this first of which investigates the importance of nonlocality at typical DIII-D H-mode ρ* values. Electrostatic global GYRO simulations of H-mode discharges at low and high rotation are shown to predict turbulence and transport levels lower than corresponding local simulations, and which are consistent with or slightly above experimental measurements and power balance analyses, even at 'near-edge' radii where gyrofluid and gyrokinetic models systematically underpredict turbulence and transport levels. The second study addresses the stabilizing effect of a significant density of energetic particles on turbulent transport. The results of local GYRO simulations of low-density QH-mode plasmas are presented, which model the fast beam ion population as a separate, dynamic ion species. The simulations show a significant reduction of transport with this fast ion treatment, which helps to understand previously reported results (Holland et al 2011 Phys. Plasmas 18 056113) in which GYRO simulations without this treatment significantly overpredicted (by a factor of 10 or more) power balance calculations. These results are contrasted with simulations of a high-density, low fast ion fraction QH-mode discharge, which predict transport levels consistent with power balance, regardless of the fast ion treatment.
    No preview · Article · Oct 2012 · Nuclear Fusion
  • [Show abstract] [Hide abstract]
    ABSTRACT: Comparisons of the bulk deuterium ion toroidal rotation to neoclassical theory reveal a significant discrepancy. The source of this discrepancy lies in the prediction of the main-ion poloidal rotation. In low toroidal rotation plasmas, Er is dominated by the pressure and poloidal rotation contributions; hence, an accurate determination of the poloidal flow is required in these conditions. We infer the main-ion poloidal rotation from measured main-ion toroidal rotation and the radial force balance relation. Inferred main-ion poloidal flow is significantly larger in the ion diamagnetic direction than NCLASS predictions. By experimentally performing scans of the plasma current, toroidal field and heating mix, the dependence of main-ion toroidal and poloidal rotation on these parameters can be understood. Comparisons of main-ion charge exchange recombination measurements of rotation with Mach probe data and several neoclassical rotation models will be performed.
    No preview · Article · Oct 2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: A range of experiments on DIII-D have investigated the role of externally applied torque and the associated toroidal rotation on confinement and stability in plasmas with high levels of normalized fusion performance. In standard H-mode plasmas, the confinement is initially enhanced by increasing rotation, but saturates at intermediate levels of rotation. However, the same effect is observed over a wider range of rotation in advanced inductive plasmas. Both ion and electron confinement improves with high rotation. Surprisingly, experiments in quiescent H-mode (QH-mode) plasmas have found the opposite trend, with improved confinement, performance and reduced turbulence levels at low rotation. The different behavior suggests that ExB shear, rather than rotation, is needed for improving confinement. In particular, for these QH-mode plasmas, it is found that the ExB shear near the edge is maintained or enhanced with torque from non-resonant magnetic fields, even at low rotation. In all scenarios, no major difference is observed in confinement whether the plasma is initiated with high rotation and slowed down, or formed with low rotation from the beginning.
    No preview · Article · Oct 2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: Application of static, non-axisymmetric magnetic fields (NAMFs) to DIII-D plasmas allows sustained quiescent H-mode (QH-mode) operation under reactor-relevant conditions of beta, collisionality and torque from neutral beam injection (NBI). QH-mode is an ideal plasma for next step devices, exhibiting H-mode confinement levels while operating without edge localized modes at constant density and radiated power. Peeling-ballooning mode stability theory suggests, and previous studies confirm, that QH-mode operation requires sufficient radial shear in the toroidal rotation near the plasma edge. In past experiments, this rotation shear was predominantly produced by torque from counter-directed NBI. In recent experiments, co-NBI torque was overcome by the counter torque due to neoclassical toroidal viscosity (NTV) produced by the NAMFs. The latest experiments have demonstrated that sufficient NTV torque can be created using NAMFs produced by coils outside the toroidal field coil. These new results open a path for QH-mode utilization in self-heated, burning plasmas, where toroidal momentum input from NBI will be small or absent.
    No preview · Article · Oct 2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: Low frequency Zonal Flows (ZFs) have been observed to trigger the L-H transition near the power threshold, by either an extended predator-prey limit cycle oscillation (LCO [1]) or a short (˜0.5-1.5 ms) ZF burst executing only part of one limit cycle. Localized turbulence suppression (kθρs˜0.5) is initiated as the ZF shearing rate approaches the turbulence decorrelation rate. Turbulence-flow correlations (via Doppler Backscattering) show that the ZF amplitude and shear initially lag the rms fluctuation level by 90^o during LCO, transitioning to 180^o as the increasing ion pressure gradient and resulting equilibrium ExB shear secure the final transition to ELM-free H-mode. In a separate experiment, localized suppression of electron-scale fluctuations (kθρs˜3) by ZF shear is also observed in an internal thermal electron transport barrier. However, in contrast to the L-H transition, here the density fluctuation level is always anti-correlated (180^o out of phase) with the ZF shearing rate. 4pt[1] L. Schmitz et al., Phys. Rev. Lett. 108, 155002 (2012).
    No preview · Article · Oct 2012
  • A. Marinoni · J. C. Rost · M. Porkolab · K. H. Burrell · J. Candy · T. C. Luce
    [Show abstract] [Hide abstract]
    ABSTRACT: The Phase Contrast Imaging (PCI) diagnostic on the DIII-D tokamak has recently been modified to image density fluctuations near the plasma mid-radius, thus enabling the investigation of core turbulence. Results are presented on core fluctuations in experiments exploring ion profile stiffness [1], i.e. the degree of sensitivity of ion temperature profiles to heat flux variations. In these experiments, plasmas were heated by neutral beams (NBI) configured to provide both high and low input torque; the injected NBI power was varied at constant torque to evaluate profile stiffness. A preliminary analysis indicates a decreased stiffness at high rotation in the outer half of the plasma. The toroidal rotation depends primarly on torque, with little or no dependence on input power. The amplitude of fluctuations increases with decreasing rotation, and the power spectra at similar torque have quantitatively similar shapes and values with little dependence on input power. Correlation lengths depend neither on torque nor input power. PCI power spectra and correlation lengths are evaluated and compared to non-linear gyro-kinetic simulations using the GYRO code. 6pt [1] J.E. Kinsey, et al., Bull. Am. Phys. Soc. 56, 282 (2011).
    No preview · Article · Oct 2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: Geodesic Acoustic Modes (GAMs) are coherent flows induced by plasma turbulence that in turn affect the turbulence and turbulent transport. Recently, in a neutral beam and electron cyclotron heated L-mode plasma in the DIII-D tokamak, strong GAM oscillations have been observed in electron temperature fluctuations Te in addition to the often-observed GAM density fluctuations. The mode frequency is constant over a radial range (δρ˜0.2), as expected of an eigenmode, with two different frequencies observed depending upon radius. Both modes exist at the location where one frequency transits to another as detected in Te. GAM oscillations in density and ExB flow peak at far edge (at ρ˜0.9) and have similar profile shapes. In contrast, the GAM oscillations in Te peak much deeper into plasma (at ρ˜0.7). After the auxiliary heating power is turned off for t,> 100 ms, the eigenmode feature evolves into a continuum. This observation of GAM properties may provide challenges for existing theories to understand GAMs and plasma turbulence.
    No preview · Article · Oct 2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: To improve poloidal rotation measurement capabilities on the DIII-D tokamak, new chords for the charge exchange recombination spectroscopy (CER) diagnostic have been installed. CER is a common method for measuring impurity rotation in tokamak plasmas. These new chords make measurements on the high-field side of the plasma. They are designed so that they can measure toroidal rotation without the need for the calculation of atomic physics corrections. Asymmetry between toroidal rotation on the high- and low-field sides of the plasma is used to calculate poloidal rotation. Results for the main impurity in the plasma are shown and compared with a neoclassical calculation of poloidal rotation.
    No preview · Article · Oct 2012 · The Review of scientific instruments
  • [Show abstract] [Hide abstract]
    ABSTRACT: The neutral-beam induced D(α) emission spectrum contains a wealth of information such as deuterium ion temperature, toroidal rotation, density, beam emission intensity, beam neutral density, and local magnetic field strength magnitude ∣B∣ from the Stark-split beam emission spectrum, and fast-ion D(α) emission (FIDA) proportional to the beam-injected fast ion density. A comprehensive spectral fitting routine which accounts for all photoemission processes is employed for the spectral analysis. Interpretation of the measurements to determine physically relevant plasma parameters is assisted by the use of an optimized viewing geometry and forward modeling of the emission spectra using a Monte-Carlo 3D simulation code.
    No preview · Article · Oct 2012 · The Review of scientific instruments
  • [Show abstract] [Hide abstract]
    ABSTRACT: A synthetic diagnostic has been developed that reproduces the highly structured electron cyclotron emission (ECE) spectrum radiated from the edge region of H-mode discharges. The modeled dependence on local perturbations of the equilibrium plasma pressure allows for interpretation of ECE data for diagnosis of local quantities. Forward modeling of the diagnostic response in this region allows for improved mapping of the observed fluctuations to flux surfaces within the plasma, allowing for the poloidal mode number of coherent structures to be resolved. In addition, other spectral features that are dependent on both T(e) and n(e) contain information about pedestal structure and the electron energy distribution of localized phenomena, such as edge filaments arising during edge-localized mode (ELM) activity.
    No preview · Article · Oct 2012 · The Review of scientific instruments

Publication Stats

11k Citations
1,024.84 Total Impact Points

Institutions

  • 1988-2015
    • General Atomics
      San Diego, California, United States
  • 2011
    • California College San Diego
      San Diego, California, United States
  • 1978-2011
    • Oak Ridge National Laboratory
      • Fusion Energy Division
      Oak Ridge, Florida, United States
  • 1998-2010
    • Lawrence Livermore National Laboratory
      • Physics Division
      Livermore, California, United States
  • 2005
    • University of Texas at Austin
      • Fusion Research Center
      Austin, Texas, United States
  • 2002
    • University of Wisconsin–Madison
      Madison, Wisconsin, United States
  • 1995-2002
    • University of California, Berkeley
      • Department of Nuclear Engineering
      Berkeley, California, United States
    • University of California, San Diego
      San Diego, California, United States
  • 1991-2002
    • University of California, Los Angeles
      • Department of Electrical Engineering
      Los Ángeles, California, United States
  • 2001
    • École Polytechnique Fédérale de Lausanne
      • Center for Research In Plasma Physics
      Lausanne, Vaud, Switzerland
  • 1995-2000
    • Massachusetts Institute of Technology
      • Plasma Science and Fusion Center (PSFC)
      Cambridge, Massachusetts, United States