Report on EUPWI SEWG on Transient Loads

1
Report onEU-PWI SEWG on Transient Loads

• Alberto Loarte
• European Fusion Development Agreement
• Close Support Unit - Garching

Contributors to SEWG CEA F.
Saint-Laurent CRPP R. Pitts ENEA G.
Maddaluno IPP G. Pautasso, A. Herrmann, T.
Eich ITER G. Federici, G. Strohmayer FZJ K.H.
Finken, M. Lehnen, J. Linke, T. Hirai FZK I.
Landman, S. Pestchanyi, B. Bazylev UKAEA V.
Riccardo, P. Andrew, W. Fundamenski, G. Counsell,
A. Kirk
2
Outline

• Summary of work in 2006
• Effects of transient loads on materials
• Characterisation of ELM loads
• Characterisation of Disruption loads
• Disruption mitigation
• 2. Plans for 2007
• 3. Conclusions

3
Expected loads in ITER transients (I)
Expected transient loads at ITER divertor/first
wall are uncertain but have strong implications
for PFC lifetime
Be
Be
Raclette - G. Federici G. Strohmayer
4
Expected loads in ITER transients (II)

• As guideline for experiments the following energy
  ranges and plasma impact energies have been
  defined
• Divertor target (CFC and W without/with Be
  coatings)
• Type I ELM 0.5 4 MJ/m2, Dt 300-600 ?s, Ee
  Ei 3 5 keV
• Thermal quench 2.0 13 MJ/m2, Dt 1-3 ms, Ee
  Ei 3 5 keV
• Main wall (Be)
• Type I ELM 0.5 2 MJ/m2, Dt 300-600 ?s, Ee
  100 eV, Ei 3 keV
• Thermal quench 0.5 5 MJ/m2, Dt 1-3 ms, Ee
  Ei 3 5 keV
• Mitigated disruptions 0.1 2.0 MJ/m2, Dt
  0.2-1 ms, radiation

5
FZJ e-beam Judith facilities
J. Linke T. Hirai
6
TRINITI facilities ? QSPA (I)
QSPA facility provides adequate pulse durations
and energy densities. It is applied for erosion
measurement in conditions relevant to ITER ELMs
and disruptions
View of QSPA facility
The diagram of QSPA facility

• Conditions for ITER ELMs disruptions not
  easily reproducible in tokamaks
• QSPA reproduces
• Energy density Timescale
• with plasma pressure 10 too high
• nT3/2QSPAnT3/2ITER but TITER 10-100 x TQSPA

• Plasma parameters (ELMs Disruptions)
• Heat load 0.5 2 MJ/m2 / 8
  10MJ/m2
• Pulse duration 0.1 0.6 ms
• Plasma stream diameter 5 cm
• Magnetic field 0 T
• Ion impact energy 0.1 keV
• Electron temperature lt 10 eV
• Plasma density 1022 m-3/ 1022 m-3

7
TRINITI facilities ? QSPA (II)
The energy density distribution on CFC surface,
Typical energy density profile on CFC surface
The energy density distribution on W surface,
Typical energy density profile on W surface,
8
TRINITI facilities ? QSPA (III)
Typical micrographs of the tungsten droplets
tracks

• During the first shot droplets ejected mainly
  from the edges of the tiles.
• As a result of edge smoothing and bridging of
  gaps the droplet ejection was reduced and mass
  losses were decreased.

9
TRINITI facilities ? QSPA (IV)
W and CFC erosion at 1.5 MJm-2

• QSPA can reproduce plasma-interaction processes
  at ITER-like load levels
• Melt layer displacement under plasma pressure
• Vapour shielding formation and effects on damage
  development
• Extrapolation to ITER requires modelling
  (Pressure too high, no magnetic field, etc.)

10
CFC results

• Under ITER-like heat loads erosion of CFC was
  determined mainly by the erosion of PAN-fibers

CFC

• Noticeable mass losses of a sample took place at
  an energy density of 1.4 MJ/m2
• Severe crack formation was observed at energy
  densities 0.7 MJ/m2(cracking of pitch fibre
  bundles)

11
W results

• 1. Under ITER-like heat loads erosion of tungsten
  macrobrush was determined mainly by melt layer
  movement and droplets ejection

W

• Noticeable W erosion mainly due to droplet
  formation took place at wmax 1.6 MJ/m2. The
  average erosion was approx. 0.06 µm/shot (1
  µm/shot during the first shot, and then decreased
  to 0.03 µm/shot after 40th pulse).
• Cracks formation was observed at energy densities
  0.7 MJ/m2.Metallographic sections show crack
  depths ranging from 50 to 500 µm.

12
ELMs erosion/deposition and impurity influxes
ELMs in JET cause significant impurity influx (
deposition) particularly when 1MJ ELMs is
reached

• Impurity generation and deposition by ELMs can
  dominate in ITER even if target lifetime is OK ?
  0.15 g-C/ELM ? 150 g-C per shot
• Determination of impurity influx and
  C-deposition during ELMs (W C comparison)

13
Main
plasma ELM energy loss
DWELM correlated with nped, Tped (? ltngt, ltTgt)
transport loss mechanism
Conduction
Convection
Convective ELMs obtained so far in regimes not
compatible with ITER QDT 10 scenario i) q95 3
(Ip 15 MA) but too high n ( n/T2) or ii) low
n but q95 gt 4 (Ip 11 MA)
14
ELM power fluxes
to PFCs (I)
During ELM event energy flows to divertor target
and main chamber PFCs
Kirk PPCF
ASDEX Upgrade Herrmann PPCF 2004
i losses across B to main wall vELM km/s
e,i losses along B to divertor
e,i losses along B to divertor
15
Energy balance of ELM divertor power pulse
qELM,div (t) ? more than 60 of DWELM,div arrives
after qELM,divmax ? smaller DTsurfELM
Fundamenski PPCF 2006
Eich JNM 2005 Loarte PoP 2004
in agreement with PIC simulations
16
In/out
asymmetries of ELM divertor power fluxes
ELM energy deposition at divertor in/out
asymmetric asymmetry depends on Bf direction but
extrapolation of observations to next step
devices remains unclear
Eich PSI 2006
17
ELM energy
fluxes to main chamber PFCs (I)
Formation and dynamics of ELM filaments and
energy deposition at main chamber starts to be
well diagnosed

• Vtor 0 before filament leaves LCFS
• vr goes from 0 at LCFS to 13 km
• Filaments leave LCFS at different times

MAST-Kirk
Herrmann-AUG

• Energy flux to the wall by individual filaments

• Energy per filament lt 2.5 DWELM (MAST)

18
ELM energy
fluxes to main chamber PFCs (II)
ELM energy deposition at main chamber given by
competition of parallel and perpendicular
transport (JET-Fundamenski Pitts validated
model) ? larger VELM (MELM)? larger DWELMwall
AUG-Kirk
JET-IR Eich

• Correlation between vELM and DWELM found
  experimentally
• vELM/cs (DWELM/Wped)a with a gt 1 (deduced from
  DIII-D, Loarte IAEA 2006)
• vELM/cs (DWELM/Wped)b with b 1/2 (JET,
  Fundamenski PSI 2006)
• vELM/cs (DWELM/Wped)g with g 0 (Kirk, AUG)

19

Pre-disruption energy confinement degradation (I)
Wplasma at thermal quench usually much smaller
than Wplasmafull-performance (except for VDEs and
ideal-b limits) caused by tE deterioration
H-L transition
thermal quench
Riccardo NF 2005
Riccardo NF 2005, Pautasso EPS 2004
VDEs b-limits (ITBs)
Size scaling and/or disruption amelioration
actions ?
20

Pre-disruption energy confinement degradation (II)
Confinement deterioration takes place in
timescales tE except for fast H-L transition
growth/locking of modes but lp does not change
much
Most disruptions ? largest divertor surface
temperature rise is caused by power fluxes during
thermal quench rather than pre-disruption
events DT qdiv t1/2
Does this hold across devices ?
21
Timescale of
thermal quench power fluxes
timescale of thermal quench fluxes increases with
R but large disruption-to-disruption variability
22
Footprint of
thermal quench power fluxes
Power flux during thermal quench broadens
significantly (even after radiation correction)
can develop toroidal asymmetries ( 2-3)
A. Herrmann - ASDEX Upgrade
G. Counsell - MAST
23
Power fluxes
on PFCs during ITER ELMs disruptions
Extrapolation of power fluxes to PFCs based on
experimental evidence models
toroidal symmetry assumed
ITER PFCs lifetime can be evaluated from these
loads ? tolerable DWELM Wt.q.
24

Calculated ELM-driven/disruption erosion in ITER
Material erosion by ELM/disruption transient
loads - no vapour shielding no redeposition
(Raclette, Federici Strohmayer)
CFC target lifetime requires qELMmax lt 1.5-2.0
GWm-2
CFC target lifetime requires qdis,max lt (2-4)
GWm-2
DWELM/Wped lt 0.05 (convective ELMs)
Wped/Wplasmafull-performance lt 0.4 (typical for
JET ELMy H-modes)
qUpper-BeELM lt 50 MWm-2 ? No Be melting
qUpper-BeELM 100-400 MWm-2 ? No Be melting for
qexperimental(t)
25

Disruption mitigation (I)
Massive gas injection systems available in
ASDEX-Upgrade, JET, TEXTOR and TORE-Supra
F. Saint-Laurent Tore Supra
M. Lehnen TEXTOR
G. Counsell - MAST

• He injection in Tore-Supra very effective in
  suppressing e runaway generation in disruptions
• Time to t.q. depends on pressure but He
  penetration does not depend on pressure
• He injection does not suppress e runaways
  already produced

• no neutral penetration in MGI shots
• dynamics of disruption correlated with impurity
  mass
• Ar D produces fast termination reduction of
  thermal loads and runaway suppression (pure Ar
  produces runaways)

26

Disruption mitigation (II)
ECRH has been used to suppress disruption by
affecting evolution of MHD
Density limit G. Maddaluno FTU
Mo-injection G. Maddaluno FTU
ECRH power and localisation requires optimisation
for different disruption type (central for DL and
peripheral for Mo-injection)
27
Plans for 2007 (I)

• Proposed joint activities
• Comparison of models for material damage during
  ELMs and subsequent plasma evolution with
  existing experimental data (mainly from JET)
  FZK, FZJ, JET, CSU Garching, IPP
• Analysis of pre-disruptive thermal confinement
  deterioration and associated power fluxes on
  PFCs for similar disruptive triggers (density
  limit, low q disruption, ideal b limits, etc.)
  and pre-disruptive regimes (L-mode, H-mode, ITBs,
  ) FZJ, CRPP, ENEA, UKAEA, CEA, IPP, JET,
  CRPP, CSU Garching
• Determination of spatial and temporal
  characteristics of power fluxes during
  disruption thermal quenches for comparable
  disruptions FZJ, CRPP, ENEA, UKAEA, CEA, IPP,
  JET, CSU Garching

28
Plans for 2007 (II)

• Proposed joint activities
• Comparative studies for the optimisation of
  disruption mitigation by massive gas injection
  for runaway suppression and thermal load
  minimisation FZJ, CEA, IPP, JET, CRPP, HAS,
  CSU Garching
• Determination of spatial and temporal
  characteristics of power/particle fluxes during
  ELMs for comparable plasma conditions CRPP,
  UKAEA, IPP, JET, CSU Garching

29
Conclusions

• Experiments and modelling of material damage
  under ITER-like transient loads are providing
  firm basis to determine maximum tolerable
  ELM/disruption loads for acceptable lifetime
• Coordinated experiments and data analysis on
  disruptions and ELMs are starting to provide a
  physics-based extrapolation of expected transient
  loads in ITER ? Further progress in 2007
  expected in by coordinated experiments and data
  analysis
• Many EU devices are now equipped with systems for
  disruption mitigation by massive gas injection ?
  significant progress in 2007 expected in this
  area by inter-machine comparison