M. Reinelt, K. Schmid, K. Krieger WG High-Z Ljubljana 01.10.200 Max-Planck-Institut für Plasmaphysik EURATOM Association, Garching b. München, Germany Extended grid DIVIMP erosion deposition modelling
Mar 27, 2015
M. Reinelt, K. Schmid, K. Krieger
SEWG High-Z Ljubljana 01.10.2009
Max-Planck-Institut für PlasmaphysikEURATOM Association, Garching b. München, Germany
Extended grid DIVIMP erosion deposition modelling
Outline
Question:Steady state surface composition of the ITER first wall ?
Our conceptual approach & strategy
Standard and extended grids for DIVIMP
Modeling of material mixing Modeling of plasma impurity generation Modeling of chemical phase formations
"Work in progress"
Question:Steady state surface composition of the ITER first wall ?
Our conceptual approach & strategy
Standard and extended grids for DIVIMP
Modeling of material mixing Modeling of plasma impurity generation Modeling of chemical phase formations
"Work in progress"
Motivation
What are the steady state surface concentrations of the ITER first wall ?
Initial surface composition
Initial surface composition
Plasma impurityconcentration
Plasma impurityconcentration
Erosion by hydrogen
Bulk material
Bulk material
Temperature
Re-deposition
Erosion by impuritiesand self sputtering
Deposition
Plasma transport
Sublimation Diffusion
Phase formations
Layer growth
Dynamic surface composition
Dynamic surface composition
Steady state surface:Total flux balance
Steady state surface:Total flux balance
Simplifications
Assumption 1: Plasma transport is instantaneous
Erosion by hydrogen
Re-deposition
Erosion by impuritiesand self sputtering
Deposition
INSTANTPlasma transport
Sublimation
Dynamic surface composition
Dynamic surface composition
Bulk material
Bulk material
Temperature
Diffusion
Phase formations
Layer growth
Simplifications
Erosion by hydrogen Temperature
Re-deposition
Erosion by impuritiesand self sputtering
Deposition
INSTANTPlasma transport
Sublimation Diffusion
Phase formations
Layer growth
CONSTANTbulk composition
Dynamic surface composition
Dynamic surface composition
Assumption 1: Plasma transport is instantaneousAssumption 2: Bulk composition is constant
All processes depend primarily on the concentrations in the near surface region.
All processes depend primarily on the concentrations in the near surface region.
Conceptual approach
DIVIMPDIVIMP
Plasma transport ofimpurities
Expected results:
* Steady state wall concentrations & erosion fluxes* Plasma impurity concentrations
Benchmark results with JET experiments Extrapolate to ITER
ERODEPDIF:Flux balances
ERODEPDIF:Flux balances
Background plasma
OEDGE(OSM)
OEDGE(OSM)
SOLPS(B2+Eirene)
SOLPS(B2+Eirene)
CARRE,recent codes
CARRE,recent codes
Grid
• Diffusion• Sublimation• Chemical phase formation
Impurity generation
SDTrimSDTrim
Materials properties
Materials properties
Conceptual approach
DIVIMPDIVIMP
Plasma transport ofimpurities
ERODEPDIF:Flux balances
ERODEPDIF:Flux balances
Background plasma
OEDGE(OSM)
OEDGE(OSM)
SOLPS(B2+Eirene)
SOLPS(B2+Eirene)
CARRE,recent codes
CARRE,recent codes
Grid
• Diffusion• Sublimation• Chemical phase formation
Impurity generation
SDTrimSDTrim
Materials properties
Materials properties
Conceptual approach
DIVIMPDIVIMP
Plasma transport ofimpurities
ERODEPDIF:Flux balances
ERODEPDIF:Flux balances
Background plasma
OEDGE(OSM)
OEDGE(OSM)
SOLPS(B2+Eirene)
SOLPS(B2+Eirene)
CARRE,recent codes
CARRE,recent codes
Grid
• Diffusion• Sublimation• Chemical phase formation
Impurity generation
SDTrimSDTrim
Materials properties
Materials properties
Extended grid (EG)
JET SG(Standard grid)
JET SG(Standard grid)
JET EG [1](Extended grid)
JET EG [1](Extended grid)
[1] By S. Lisgo
Extended grid (EG)
... to be filled with plasma
Conceptual approach
DIVIMPDIVIMP
Plasma transport ofimpurities
ERODEPDIF:Flux balances
ERODEPDIF:Flux balances
Background plasma
OEDGE(OSM)
OEDGE(OSM)
SOLPS(B2+Eirene)
SOLPS(B2+Eirene)
CARRE,recent codes
CARRE,recent codes
Grid
• Diffusion• Sublimation• Chemical phase formation
Impurity generation
SDTrimSDTrim
Materials properties
Materials properties
Material mixing model
Material mixing
Plasma
Each tile receives a flux due to erosion & re-deposition from other tilesPlasma transport is characterized by a re-deposition matrix:
i on tile up ends that j tilefrom melement offlux eroded ofFraction , mjir
Flux of material m on tile i:
n
j
mji
Dj
mj
mjm
i rYxN
tt
1,
Result: Set of n coupled differential / algebraic equationsResult: Set of n coupled differential / algebraic equations
Concept: The first wall is divided into n tilesConcept: The first wall is divided into n tiles
Mixed material formation
Plasma
BulkReactionzone
Backgroundplasma
Concept: Each tile is composed of a thin reaction zone and a bulk materialConcept: Each tile is composed of a thin reaction zone and a bulk material
Allows layer growth and erosion, sublimation and simplified chemistry. No diffusion!Allows layer growth and erosion, sublimation and simplified chemistry. No diffusion!
* Constant thickness Collision cascades:
< 50 nm* Variable composition
* Constant source / sink
* Constant composition
Mixed material formation
Bulk
For n-tiles and k-chemical phases: kn coupled differential equationsFirst tests with Mathematica: Works for >1000 coupled equations
For n-tiles and k-chemical phases: kn coupled differential equationsFirst tests with Mathematica: Works for >1000 coupled equations
dσX / dt =
Plasma
+Influx (by re-deposition matrix)
– Erosion flux (by hydrogen and impurities)
– Flux of sublimation (from vapor pressure of the chemical phase)
± Balancing flux (with bulk material)
k Chemical phases or elements [X] [Y] [Z] ...
Chemical reactions
+Flux of formation reactions (X is Product)
– Flux of dissociation reactions (X is Reactant)
Concept: Each tile is composed of a thin reaction zone and a bulk materialConcept: Each tile is composed of a thin reaction zone and a bulk material
Prove-Of-Principle (w/o chemical reactions)
10 20 30 40 50 60 70
10
20
30
40
50
60
70
Destination bin
So
urc
e b
in
1.00E-4
6.31E-4
3.98E-3
2.51E-2
1.58E-1
1.00E0
Flux fraction (LOG scale)
BeTotal
Numerical solution for 69 tiles, re-deposition matrix and C wall + Be evaporationNumerical solution for 69 tiles, re-deposition matrix and C wall + Be evaporation
Initial Be coverage Re-deposition of Be
Prove-Of-Principle (w/o chemical reactions)
10 20 30 40 50 60 70
10
20
30
40
50
60
70
Destination bin
So
urc
e b
in
1.00E-4
6.31E-4
3.98E-3
2.51E-2
1.58E-1
1.00E0
Flux fraction (LOG scale)
BeTotal
Numerical solution for 69 tiles, re-deposition matrix and C wall + Be evaporationNumerical solution for 69 tiles, re-deposition matrix and C wall + Be evaporation
Initial Be coverage Re-deposition of BeBe is covered by C
Prove-Of-Principle (w/o chemical reactions)
Numerical solution for 69 tiles, re-deposition matrix and C wall + Be evaporationNumerical solution for 69 tiles, re-deposition matrix and C wall + Be evaporation
[Be
/ Ǻ2 ]
Time [s]
Tiles with Be at surface
Tiles with C at surface
All Be is covered by C
Conceptual approach
DIVIMPDIVIMP
Plasma transport ofimpurities
ERODEPDIF:Flux balances
ERODEPDIF:Flux balances
Background plasma
OEDGE(OSM)
OEDGE(OSM)
SOLPS(B2+Eirene)
SOLPS(B2+Eirene)
CARRE,recent codes
CARRE,recent codes
Grid
• Diffusion• Sublimation• Chemical phase formation
Impurity generation
SDTrimSDTrim
Materials properties
Materials properties
Model of surface chemistry
ITER first wall
HeHe BeBeWW
CC
OO
HH
NNElementsElements
ITER first wall
HeHe
Nitrides:WNBe3N2
Nitrides:WNBe3N2
Hydrides:BeH2
CXHY
OH2
Hydrides:BeH2
CXHY
OH2
Carbides:WC, W2CBe2C
Carbides:WC, W2CBe2C
Beryllides:Be2W, Be12W
Beryllides:Be2W, Be12W
Oxides:WO3
BeOCO2
Oxides:WO3
BeOCO2
BeBeWW
CC
OO
HH
NNElementsElements
Binary phasesBinary phases
ITER first wall
HeHe
Nitrides:WNBe3N2
Nitrides:WNBe3N2
Hydrides:BeH2
CXHY
OH2
Hydrides:BeH2
CXHY
OH2
Carbides:WC, W2CBe2C
Carbides:WC, W2CBe2C
Beryllides:Be2W, Be12W
Beryllides:Be2W, Be12W
Oxides:WO3
BeOCO2
Oxides:WO3
BeOCO2
BeBeWW
CC
OO
HH
NN
Tungstates:BeWO3, BeWO4
Hydroxides:Be(OH)2, W(OH)X
…
Tungstates:BeWO3, BeWO4
Hydroxides:Be(OH)2, W(OH)X
…
ElementsElements
Binary phasesBinary phases
Ternary phasesTernary phases
Simplified description of ITERs first wall chemistry
Be W C Be2C W2C WC Be2W Be12WBegas BeO Oads
WO3 WO3,gas
Chemical phases
Chemical phases
2 Be + C → Be2CBe2C → 2 Be + CW + C → WCWC → W + C2 W + C → W2CW2C → 2 W + CW + 2 Be → Be2WBe2W → W + 2 BeW2C → WC + WWC + W → W2CBe + O → BeOBeO → Be + OW + 3 O → WO3
WO3 → W + 3 O
Sublimation:Be → Begas
WO3 → WO3,gas
O-Adsorption:O2,gas → Oads
Oads → O2,gas
Elementary reactions
Elementary reactions
sm
smk
kT
EkCW
4
2 11
121
1 withexp][][
ssmk
kT
EkCW 1
22
221
2 withexp][2
…
Equations for reaction fluxesEquations for reaction fluxes
...][
212
dt
CWd...22
][21
dt
Wd
Reaction balancesReaction balances
Change of areal density of chemical phase = + all formation reaction fluxes – all dissociation reaction fluxes
Couple to plasma transport code Couple to plasma transport code
Benchmarking example: W/Be/O
XPS
XPS experimental data• 2.1 nm Be on • W (Substrate,pc)• 10-10 mbar O2
Layered system
XPS experimental data• 2.1 nm Be on • W (Substrate,pc)• 10-10 mbar O2
Layered system
Model
Model of coupled reaction equations
Elementary processes:• O adsorption• Be and W oxidation• BeO and WO3 dissociation• Be and WO3 sublimation• Be2W formation and dissociation
Not included: Depth profiles (Homogeneous distributed phases)
Model of coupled reaction equations
Elementary processes:• O adsorption• Be and W oxidation• BeO and WO3 dissociation• Be and WO3 sublimation• Be2W formation and dissociation
Not included: Depth profiles (Homogeneous distributed phases)
SummarySet up a scalable model for JET (and ITER) that describes the first wall material evolution as a combination of:
+ Dynamic impurity generation (Parametrised TRIDYN)+ Plasma transport via a static background (DIVIMP)+ Some temperature dependent processes (Chemical phase formations, sublimation, directly benchmarked by XPS data)
Method: Numerical solution of a set of coupled algebraic differential equations (Mathematica)
Result: Time evolution of
Surface concentrations (incl. layer growth) Plasma impurity concentrations Erosion and re-erosion fluxes Benchmark the results with JET experiments
(e.g. post-mortem analysis of layers, spectroscopy of erosion fluxes)
Erosion and re-erosion by impurities
Assumption: Individual sputteryields Yj of a mixture of elements scales linearily with the surface concentration
Assumption: Individual sputteryields Yj of a mixture of elements scales linearily with the surface concentration
Works well for Be / C but only fairly good for W / C, W / BeWorks well for Be / C but only fairly good for W / C, W / Be
0 200 400 600 800 10000.0
0.2
0.4
0.6
0.8
1.0
Be
su
rfa
ce
co
nc
en
tra
tio
n
Fluence (#/A2)
TRIDYN Analytical model
50 eV D + 100 eV Be on C(Precalculated Yields)
10 20 30 40 50 60 70
10
20
30
40
50
60
70
Destination bin
So
urc
e b
in
1.00E-4
6.31E-4
3.98E-3
2.51E-2
1.58E-1
1.00E0
Flux fraction (LOG scale)
BeTotal
Re-deposition matrix (JET SG)
Promtre-deposition
... ... ...
Simple (unverified) OSM plasma background
Re-deposition matrix by DIVIMP
2 3 4
-2
-1
0
1
2
Outer plasma grid cells Projection of the grid to the vessel wall Areas for homogeneous
impurity launch
Z [
m]
R [m]
Lauch flux of Be impurity ionsand map points of re-deposition (Charge resolved)
Re-deposition matrix, n ~ 70
static BGP
Bin
static BGP,standard grid