FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
Liquid metal divertor options for FNSF
Dick Majeski, PPPL
FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
Engineering features of liquid metal divertors
Continuously renewed as new fluid enters the system Neutron damage not a concern for liquid metals
– Liquid metal is only subject to PMI – Substrate is only subject to neutron damage
PMI limited to sputtering + evaporation – No significant erosion/redeposition issue for liquids
Several design approaches – Fast flowing jets, open-channel systems – Slowly flowing systems with capillary restraint (porous refractory
metals) Liquids commonly considered: lithium, gallium, tin BUT – engineering of liquid metal systems is still at an early
stage of development – No attempt at a full flowing liquid metal divertor design since the
ALPS program investigated candidates for NSTX
FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
Lithium and high recycling liquids (gallium, tin) have different fusion applications
Low recycling lithium walls shown to increase energy confinement – Smaller scale size, driven tokamaks/spherical tokamaks possible
» Exploratory reactors with high tritium burnup fraction » Fusion-fission hybrids » Neutron sources at high wall loading (>10 MW/m2) for material
testing But: implementation of lithium walls in a nuclear tokamak designed for
high recycling walls carries few advantages – For the FNSF divertor, a reasonable choice may be a high recycling
liquid metal – gallium or tin If modest improvements in confinement or density control are desirable,
a tin-lithium eutectic might be a reasonable choice – Self-cooled gallium, tin, Sn-Li jet – J x B driven flows – Evaporatively cooled porous system
FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
Evaporation limits the operating temperature range for liquid metals
Gallium-1100 C
Tin-1300 C
But - 450 °C lithium limiter operation in CDX-U did not result in a lithium plasma – Zeffective remained close to 1
Only experimental data available
Lithium~450 C
SnLi
Total flux (sputtering + evaporation) imposes a temperature limit (Rognlien)
– 500 – 600 °C for SnLi
– 800 – 900 °C for Sn Operating limit for lithium is much lower
– <450 °C in a conventional divertor – No estimate has been performed for the Super-X geometry
FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
Flow rates for a liquid metal divertor Flow rate is set by limits in erosion, temperature or liquid D-T inventory
– Temperature, hydrogenic inventory limits most restrictive
Capillary or thin-film systems rely on cooling from behind the liquid substrate. – Flow rate of liquid not determined by heat removal – For lithium, flow rate is determined by requirement that liquid be
removed before LiD(T) forms, precipitates For gallium, tin thin-films, flow rate determined by erosion replacement
Required flow rate is high for all “self-cooled” concepts (thermal limit) – Jets or fast open-channel flow (e.g. J x B driven flows) – Flow rate determined by heat flux, flow path – Typical flow rates: 5-10 m/sec for 10 MW/m2 power flux (lithium)
» Flow path: 10s of cm at most » Estimate assumes only heat conduction, not convection » Power limits higher for gallium, tin
FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
Gallium jet experiment in ISTTOK R.B.Gomes et al., Fus. Eng. Des. 83 (2008) p. 102
Gallium jet
Discharge behavior is similar with
gallium jet and graphite limiters
FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
High power handling tests of lithium systems
Two approaches developed for very high power handling Both approaches successful at exceeding conduction-limited power
density limits – First approach (Red Star, Russian Federation) uses evaporation
of lithium in a porous mesh target » Employs heat of evaporation » Evaporating lithium provides vapor shielding of target
– Second approach employs naturally generated (convective) flows in free surface liquid lithium for redistribution of heat (PPPL)
Both approaches have issues for application in a tokamak – Lithium influx with evaporative technique may be prohibitive – High magnetic field may suppress self induced flows
But both techniques have demonstrated heat handling capability in excess of 50 MW/m2
FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
Evaporative cooling could provide very high transient heat handling capability
Evaporative flux must be confined to divertor region – Candidate for a Super-X divertor target?
t, o C 100 500 900 1300 1700
Q,
MW/m2
Hg
Na
Li
Ga
10-1
100
101
102
M.N. Ivanovskiy et al., Evaporationand Condensation of Metals, Moscow,Atomizdat, 1976
FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
The organization of works in Russia on Lithium Capillary-Pore Systems problem
ROSATOM Federal State Unitary Enterprise “Red Star”
Very high power handling demonstrated - >50 MW/m2 (25 MW/m2 steady-state) ~60 MW/m2, 300 sec. demonstrated with a 3 mm liquid lithium film on CDX-U
Liquid Lithium Limiter on FTU
Beam spot
IR image of e-beam heated lithium tray limiter in CDX-U
FNSF/PFC session 4 August 2010
PRINCETON PLASMA PHYSICS LABORATORY
PPPL
Summary Liquid metals for the FNSF divertor
Liquid metal PFCs are at an early stage of development Implementations of lithium in tokamaks primarily for wall conditioning
– Very few experiments with liquid lithium as a PFC » First-generation experiments started in NSTX, HT-7 » Second-generation tests starting in LTX
Few (two!) experiments with gallium in tokamaks; none with tin Liquid metal PFC development may offer
– In-situ renewal of PFCs – Self-repairing walls (disruptions, ELMs) – Separability of neutron damage, PMI issues – Improved power handling – Access to new tokamak confinement regimes (lithium)
» But lithium is less attractive than gallium or tin as a LM PFC in “conventional” reactor designs