Burning Plasma Simulations: Edge Issues

1
Burning Plasma Simulations Edge Issues

• D. P. Coster
• Max Planck Institute for Plasma Physics,EURATOM
  Association, Garching, Germany

2
Outline

• Why are edge issues important for burning plasma
  simulations?
• All the external sources pass through the edge
• The non-neutron waste products are all removed
  via the edge
• Much of the performance of the core is determined
  by the edge
• The edge provides constraints on the core

3
Outline

• Why are edge issues important for burning plasma
  simulations?
• All the external sources pass through the edge
• Particle
• Gas puff
• Pellets
• NBI
• Energy
• NBI
• RF
• The non-neutron waste products are all removed
  via the edge
• Much of the performance of the core is determined
  by the edge
• The edge provides constraints on the core

4
Outline

• Why are edge issues important for burning plasma
  simulations?
• All the external sources pass through the edge
• The non-neutron waste products are all removed
  via the edge
• Particles
• D, T
• He
• Energy
• Much of the performance of the core is determined
  by the edge
• The edge provides constraints on the core

5
Outline

• Why are edge issues important for burning plasma
  simulations?
• All the external sources pass through the edge
• The non-neutron waste products are all removed
  via the edge
• Much of the performance of the core is determined
  by the edge
• H-mode barrier
• He concentration
• Impurities
• The edge provides constraints on the core

6
Outline

• Why are edge issues important for burning plasma
  simulations?
• All the external sources pass through the edge
• The non-neutron waste products are all removed
  via the edge
• Much of the performance of the core is determined
  by the edge
• The edge provides constraints on the core
• Minimum separatrix density
• ELMs
• Particle throughput

7
What tools are currently available?

• Mainly a European perspective
• Edge transport codes
• Global (in the sense of whole domain, consistent)
• SOLPS (B2-Eirene)
• EDGE2D-NIMBUS
• Fluid plasma (2d)
• Kinetic neutral (2d or 3d)
• Multiple species
• Local (limited domain, against a fixed
  background)
• ERO
• Kinetic trace impurities
• 2d/3d
• ASCOT
• Fast ions or electrons
• PIC codes

8
Global codes, I

• Fluid plasma (2d)
• kinetic plasma effects not properly treated
• Flux limiters
• Should we investigate fluid/kinetic hybrids?
• 2d
• Ignore 3D effects
• Localised recycling, sources etc.
• Multiple species
• Mixed material issues (later)

Th. Pütterich
9
Global codes, II

• Fluid plasma (2d), cont
• Solution domain somewhat limited
• Doesnt extend all the way to the vacuum vessel
• What are the appropriate boundary conditions
  there?
• Implications for
• heat/particle loads in main vessel
• Sources of impurities

10
Global codes, III

• Kinetic neutral (2d or 3d)
• Kinetic coupling to a Monte-Carlo code
• More accurate, slower
• Monte-Carlo Noise
• (Fluid usually not full Navier-Stokes
• Not as accurate
• Own temperature?)
• Still discussion of role of physics
• He elastic collisions
• Hydrocarbon break up chains
• Vibrational excitations
• Often details of bypasses, wall out-gassing
  neglected
• Neutral-neutral collisions and optically thick
  regions

11
Some issues, I

• Impurities
• Intrinsically produced
• Somewhat simplified erosion/deposition models
• Plasma solution domain usually doesnt extend to
  vacuum vessel
• Problem with main chamber sources
• PSI models usually too simple
• Usually not affected by the plasma in the
  simulations
• Heating
• Out-gassing/absorption
• Isotope exchange
• ELM target cooling with Ar at inner target
• Hydrocarbon chemistry including T co-deposition
• How can we improve them?

Th. Pütterich
12
Some issues, I solution path

• Effects can be studies using local codes
• ERO
• Local erosion / deposition analysis
• Detailed model for processes
• Including CxDy
• If effects are important need to, either
• Couple to global code
• Incorporate physics in global code

A. Kirschner
13
Some issues, II

• Effects of fast particles(?)
• Differences in energy seen by Langmuir probes and
  Thermography?
• Look at with kinetic codes
• ASCOT
• Monte-Carlo treatment of fast ions or fast
  electrons
• PIC
• Problem with time-space scales (particularly if
  more than one space dimension)
• Need to have recycling physics

Chankin, SOLPS modelling of ASDEX-U H-mode
Plasma, Edge Physics Forum, Jan 2005
Unpuffed, H-mode, AUG
14
Some issues, III

• Transport coefficients are inputs to the global
  (and local) codes
• Need 1st principles treatment of radial transport
• 3 options
• Parametrize radial and poloidal dependence of
  transport coefficients
• Non-diffusive nature ( pinches)
• Couple transport and turbulence codes
• Move all the physics currently in the transport
  codes into the turbulence codes

Nishimura (while a PostDoc at IPP)
15
Some issues, IV

• Drifts
• Still not complete agreement on the equations to
  be implemented
• Not nearly as robust as the non-drift versions
• Often have additional time-step limitations
• Sometimes have problems with extending to far
  into the core
• Sometimes have problems with reduces solution
  domain in terms of plasma parameters
• Need to develop a technique for moving from drift
  treatment (2d) to a neoclassical (1d radial)
  treatment

16
Integration

• Edge with core
• Edge with PFC

17
Integration

• Edge with core
• Couple core and edge
• COCONUT
• JETTO-SANCO-EDGE2D-NIMBUS
• Add core physics to edge codes
• SOLPS efforts are a start
• Edge with PFC

Fichtmueller et al, EPTSW
Buerbaumer, Bonnin
18
Integration, II

• Edge with core
• Source from ASTRA
• Use B2.5-I to fit to exp. (Y.P. Chen)
• Can then use to examine W erosion (Kukushkin)
• Used as a starting point for ELM simulations
• Or for disruptions (Konz)
• Edge with PFC

2
1
4
3
19
Integration

• Edge with core
• Edge with PFC
• Couple edge codes
• globaland local
• Add physics from local codes to global codes

20
New developments

• Working towards more integrated models
• Integrating plasma and plate
• SOLPS5-B2
• Fluid neutrals
• Thermal model for the plate ???
• Latest Roth chemical sputtering formula
• Working towards modelling mixed materials

21
New developments

• Add accounting of erosion and deposition (B2 side
  only so far)
• Added a simple mixed materials model

Without re-erosion of deposited material
With re-erosion of deposited material

DCHe, SOLPS-B2, AUG
22
New developments

• Add accounting of erosion and deposition (B2 side
  only so far)
• Added a simple mixed materials model
• Can now treat Be main chamber wall and C divertor
  plates
• New time-scales!

DCHeBe, SOLPS-B2, AUG
23
Summary

• Need to think about the edge for Burning Plasma
  Simulations
• Progress is being made
• But we still have a way to go!
• Challenges
• The same as more many other areas
• Time-scales
• Space-scales
• Dimensionality