(at the moment: TEXTOR, ASDEX, JET, ITER, PISCES, MAGNUM

1
Status and plans of ERO modelling
A. Kirschner, D. Borodin, S. Droste, M. Airila
Institut für Plasmaphysik, Forschungszentrum
Jülich GmbH

• Examples of ERO modelling
• TEXTOR tungsten testlimiters, 13CH4 injection
• JET hydrocarbon transport
• ITER predictions of target lifetime tritium
  retention
• Plans
• Modelling further benchmarking predictions
• Code development

2
The ERO code

• ? 3-dimensional, Monte Carlo
• ? Various experimental geometries are possible
• (at the moment TEXTOR, ASDEX, JET, ITER,
  PISCES, MAGNUM)
• ? Plasma-wall-interaction
•   - physical sputtering, chemical erosion (CH4 or
  higher hydrocarbons)
• - C-deposition from background, re-deposition of
  eroded particles
• - simple material mixing model
•  
• ? Local particle transport
•   - ionisation, recombination, dissociation
• - Ehrhardt/Langer or Janev data for CH4 reaction
  chain
•   - friction (Fokker-Planck), thermal force,
  Lorentz force (E and B-field), diffusion D?

3
TEXTOR tungsten testlimiters (1)
Experiment
net erosion

• Spherically shaped tungsten testlimiter
• deuterium plasma with main impurity carbon

carbon layer
carbon layer
Modelled compounds in interaction layer (in
equilibrium)
net erosion
Carbon
Tungsten
4
TEXTOR tungsten testlimiters (2)
Modelled carbon and tungsten distribution above
the testlimiter
Neutral Carbon
Neutral Tungsten
5
TEXTOR tungsten testlimiters (3)
WI emission near the testlimiter modelling vs.
experiment
e-folding lengths
radial profile
Te(LCFS) 80eV, ne(LCFS) 4.81012cm-3
Good agreement between simulated and measured
WI-profiles in radial direction
6
TEXTOR tungsten testlimiters (4)
CII emission near the testlimiter modelling vs.
experiment - radial profile -
Te(LCFS) 80eV, ne(LCFS) 4.81012cm-3
Good agreement between simulated and measured
CII-profile in radial direction if negligible
chemical erosion is assumed (transient C-layer)
7
TEXTOR tungsten testlimiters (5)
CII emission near the testlimiter modelling vs.
experiment - e-folding lengths -
Good agreement between simulated and measured
CII-profiles in radial direction
8
TEXTOR 13CH4 injection (1)
13C deposition (108s exposure)
Experimental set-up
B
radial
46.0 cm (LCFS)
13CH4
D, 12C4
Al-plate
9.21019 13CH4/gas puff
13C-deposition efficiency ? 0.5
toroidal length 115 mm
Plasma parameter exponential decay in radial
direction Te(LCFS)54eV, ne(LCFS)1.91012cm-3
9
TEXTOR 13CH4 injection (2)
ERO-Modelling - Janev data for rate
coefficients - effective sticking for
hydrocarbons S 1 or S 0
10
TEXTOR 13CH4 injection (3)
ERO Modelling Conclusion

• Low sticking of hydrocarbons necessary in
  simulations, but
• modelled 13C deposition efficiency is still too
  high
• 2.5 with Ehrhardt-Langer data
• 14 with Janev data
• measurement 0.5
• additional assumption high re-erosion of
  deposits (8 vs. 1.5 for substrate
  material)
• low efficiency of 0.5 is reproduced

low effective sticking of hydrocarbons (self-re-er
osion, enhanced re-erosion through background
plasma)
11
JET hydrocarbon transport (1)
Carbon transport in inner divertor MkIIA
Plasma OSM for standard gas fuelled ELMy H-mode,
12 MW (44029)
Large amount of carbon found at inner louver Can
not be modelled with standard assumptions (high
sticking for CxDy)
12
JET hydrocarbon transport (2)
Carbon deposition at louver of inner divertor
MkIIA
Measured carbon deposition at inner louver can be
simulated if zero effective sticking for CxHy and
enhanced re-erosion of re-deposited carbon.
13
JET hydrocarbon transport (3)
Quartz microbalance (QMB) at inner louver of
divertor MkIIGB SRP
Principle Resonance frequency of quartz changes
with mass Thickness resolution 0.2 nm
Shot-resolved measurement of deposition (erosion).
14
JET hydrocarbon transport (4)
Deposition on QMB in dependence on plasma
configuration
Large deposition for strike point moved
downwards to QMB. Further increase of
deposition for strike point on horizontal target.
15
JET hydrocarbon transport (5)
QMB modelling with ERO plasma parameter for
DOC-L (D. Coster)
In addition plasma paramater for DOC-U and base
configuration available.
16
JET hydrocarbon transport (6)
Carbon deposition on QMB modelling experiment
DOC-U vs. DOC-L vs. BASE
Particles to louver 1/cm2s
Dependences of QMB deposition data are
reproduced. Reasonable agreement of absolute
values with assumption of effective sticking S
0 (high re-erosion), and including erosion by
atoms.
17
ITER predictions of target lifetime tritium
retention (1)
Plasma parameter for outer divertor
B2-Eirene - semi-detached - fuelling mostly by
puffing from top (A.S. Kukushkin)
18
ITER predictions of target lifetime tritium
retention (2)
Erosion re-deposition along outer target Roth
vs. 1 yield
Effective sticking 0 for hydrocarbons, MolDyn
for atoms
with Roth yield 94 re-deposition 6 loss to PFR
with fixed yield 80 re-deposition 20 loss to
PFR
? Gross erosion 10 times larger with fixed 1
than Roth yield. ? Local re-deposition
significantly smaller with fixed yield.
19
ITER predictions of target lifetime tritium
retention (3)
Long term tritium retention ERO prediction
T/C 0.2
Assumption T retention by non-local
re-deposition of hydrocarbons (outer divertor) ?
long-term tritium retention
T-retention with S 0 much larger than with S
1 (up to factor of 35!!). With Roth formula
significant decrease of T-retention.
20
Modelling plans (1)
Benchmarking, predictions code development
TEXTOR - 13C hydrocarbon injection through
different substrate materials (graphite vs.
high-Z) and testlimiter geometries (roof vs.
mushroom) - hydrocarbon injection through nozzle
in comparison with limiter (effect of
hydrocarbon recycling on D/XB) - castellated
limiter structures (deposition in
gaps) JET - 13C injection experiments - QMB
studies for dedicated discharges - hydrocarbon
injection in dependence on plasma configuration
and puffing location (light emission and D/XB)
21
Modelling plans (2)
Benchmarking, predictions code development

• ASDEX-UG
• - hydrocarbon injection experiments (deposition
  and D/XB)
• tungsten transport
• ITER
• - predictions of target lifetime and tritium
  retention
• Further experiments PISCES, MAGNUM,
• Code development
• parallelisation, improvement of surface layer
  model (TriDyn, MolDyn?),
• unification of various versions (limiter,
  divertor, )

22
Surface model in ERO
Constant amount of particles in interaction
layer Homogenous distribution of different
species