Challenges in edge modeling
IPP-Teilinstitut Greifswald, EURATOM Association, Wendelsteinstraße 1, D-17491 Greifswald, Germany
Outline:1. Motivation2. Plasma model3. Neutral model4. Turbulence5. Multi-scale strategies6. Plasma-wall interaction7. Remarks on Integrated Modeling
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Ralf Schneider
Motivation
Max-Planck-Institut für Plasmaphysik, EURATOM Association
• ultimate goal: predictive quality for tokamaks and stellarators (‘integrated modeling’)
What are the problems to be solved in edge modeling?
Which strategies should be followed?
What are the consequences of that?
Motivation
Max-Planck-Institut für Plasmaphysik, EURATOM Association
• edge modeling made considerable progress: yes!but: long list of topics
Summary report of the American Divertor and Edge Plasma Theory Working Group, 22.12.1992
J. Neuhauser: ITER Workshop 17.-21.7.1989
Motivation
Max-Planck-Institut für Plasmaphysik, EURATOM Association
• edge modeling made considerable progress: yes!but: long list of topics
A. Kukushkin: ITER meeting 1993
Plasma edge physics
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Max-Planck-Institut für Plasmaphysik, EURATOM Association
drifts and currents: open questions
• Bruce Scott: Physics of Plasmas 10 (2003) 963only anomalous viscosity – either Reynolds stress or correction from gyroviscosity – can create anomalous transportanomalous resistivity cannot create anomalous transport or radial electric field, because resistivity is a momentum conserving friction between electrons and ions-> classical equation for the potential (div j = 0)
• numerical performance: time-step limits for complete Newton solvers (UEDGE) and equation sub-cycling (B2);semi-implicit solvers? (Zagorski et al.; see also talk by A. Kalyentev)
Plasma models
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Plasma models
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Plasma models
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Plasma models
Max-Planck-Institut für Plasmaphysik, EURATOM Association
kinetic corrections: fluid corrections or coupling with kinetics
Plasma models
Coupling with kinetic code (BGK): 2 kinetic equations for thermal conductivity and viscosity(Kukushkin, Runov, Igitkhanov 1994)
Max-Planck-Institut für Plasmaphysik, EURATOM Association
3D and ergodic effects (see also talk by A. Kalyentev)
Plasma models
Max-Planck-Institut für Plasmaphysik, EURATOM Association
• Monte Carlo vs. fluid: different level of accuracy and complexity; flux limits necessary for making the fluid model realistic
Neutral models
D. Coster et al., EPS2005
• atomic and molecular data: below 5 eV??
Molecular physics
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Franck-Condon atoms: low plasma temperature -> mostly molecules reflected from saturated walls
MAR
Max-Planck-Institut für Plasmaphysik, EURATOM Association
MAD and MAI
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Max-Planck-Institut für Plasmaphysik, EURATOM Association
THE central problem
Turbulence
Max-Planck-Institut für Plasmaphysik, EURATOM Association
• fitting of transport coefficients
Turbulence
Max-Planck-Institut für Plasmaphysik, EURATOM Association
• fitting of transport coefficients
Turbulence
• physics-based scaling• full coupling
Multi-scales
sputtered and backscattered species and fluxes
Plasma-wall interaction
Moleculardynamics
Binary collisionapproximation
KineticMonte Carlo
Kineticmodel
Fluidmodel
impinging particle and energy fluxes
Max-Planck-Institut für Plasmaphysik, EURATOM Association
• Chemical Erosion of carbon by hydrogen produces hydrocarbon species (CxHy)
• Dissociation & Recombination's leads to amorphous hydrocarbon layer formation
• Carbon acts as sponge for hydrogen• Tritium is retained by co-deposition with carbon, on the plasma facing
sides or on remote areas.
HydrogenG F Counsell, Plasma Sources Sci. Technol. 11 (2002) A80–A85
Hydrocarbon-codeposition
Max-Planck-Institut für Plasmaphysik, EURATOM Association
2eV CH3 onto amorphous hydrocarbon
Classical MD
Max-Planck-Institut für Plasmaphysik, EURATOM Association
MD studies of interaction of hydrocarbons with amorphous carbon:• empricial Brenner potential• reflection coefficients of hydrocarbons on amorphous hydrocarbon(collaboration with K. Nordlund, Univ. Helsinki, U. v. Toussaint, IPP Garching, D. Naujoks, IPP Greifswald)
PhDs (HGF funded), Amit Raj Sharma, Abha Rai
Energy (eV)
CH
x R
efle
ctio
n co
effic
ient
1.2
1
0.8
0.6
0.4
0.2
00.01 0.1 1 10
CHCH2 CH3CH4
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Multi-scale strategies
WWW Uppsala Univ. Sweden
Classification:
A) Microscale info. local
B) Microscale info. global
C) Combination of (A) and (B)
D) Self-similarity in scales
http://www.math.princeton.edu/multiscale/review.pdf
Serial coupling
Concurrent “on the fly” coupling
Renormalization group
Multi-scaling Paradigms:
Multi-scale approach
Microscales
Molecular Dynamics (MD)
Mesoscales
Kinetic Monte Carlo (KMC)
Macroscales
KMC and Monte Carlo Diffusion (MCD)
Max-Planck-Institut für Plasmaphysik, EURATOM Association
´Intelligent´ coupling necessary
Meso/macro-scales
)s/cm(D 2
T1000 / )( 1K
Strong dependence on void sizes and not void fraction
Large variation in observed diffusion coefficients
standardgraphites
highly saturatedgraphite
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Diffusion coefficients without knowledge of structure are meaningless
Multi-scale modeling of hydrogen transport in porous graphite• inclusion of molecules for re-emission, • extension to chemical sputtering (Küppers-Hopf-cycle) (collaboration with M. Warrier, Max-Planck-India-Fellowship)
PhD (HGF funded): Abha RaiExperiment: P. Franzen, E. Vietzke, J. Vac. Sci. Technology A12(3), 1994
• H-atom release limited by detrapping process, not by diffusion
Modeling:
• results matches very well exp.
Hydrogen re-emission
Max-Planck-Institut für Plasmaphysik, EURATOM Association
Temperature (K)
Ree
mitt
ed F
lux
(%)
0
20
40
60
80
100
120
0
20
40
60
80
100
200 600 1000 1400 1800 2200
USB15
EK96
H2
H
Re-
emitt
ed F
lux
(%)
Temperature (K)600 1000 1400 1800
0
0.2
0.4
0.6
0.8
1.0
1.2
-0.2 800 1200 1600 2000
H 5% Void
H2 5% Void
H 9% Void
H2 9% Void
Max-Planck-Institut für Plasmaphysik, EURATOM Association
• a VERY personal view!
• technical remarks:• open source• standard interfaces• benchmarking
Integrated modeling
• physics remarks:• hierarchical models needed (upgrading, downgrading)• validation (experiment, theory)
• Combination and coupling of more and more codes will not improve the reliability and predictive quality• Depending on the problem and the question one needs very different tools (intelligent interpolation, interpretation, basic physics studies, …)
T. Angot et al., University of Provence, Marseille
STM of graphite surface Simulation
Modelling of Hydrogen bombardement of single crystal
Max-Planck-Institut für Plasmaphysik, EURATOM Association
surface science and low-temperature plasma physics
Model systems needed
Summary
Max-Planck-Institut für Plasmaphysik, EURATOM Association
• Multi-scale physics: combination of methods • ´Intelligent´ coupling necessary !??
• Hierarchy of models (downgrading, upgrading)
• Real structure of the material to be included
• Model systems needed
• Standards for the modules: open source code, interfaces, benchmarking, model validation