Status and Plans for APEX Task A Plasma/Liquid-Wall Interactions

Transcription

1 Status and Plans for APEX Task A Plasma/Liquid-Wall Interactions T.D. * with J. Brooks, T. Evans, A. Hassanein, & M. Rensink* APEX Fall Meeting PPPL Nov. 4, 2002 Work performed under auspices of US DOE; for * by Univ. of Calif LLNL, under contract no. W-7405-Eng-48

2 Outline Highlights from FY02 Plans for FY03 Status and examples of present work

3 Highlights for FY02 - APEX related plasma/wall modeling CLIFF flinabe temperature limits based on core impurity intrusion* CLiFF divertor heat flux with tilt* ELM heat-flux characterization Spheromak edge-plasma modeling* * Planned FED papers NSTX lithium module impurity influx DIII-D DiMES and CDX-U plasma modeling

4 Dai-Kai s question: What should be done to confirm the allowable surface temperature? More detailed comparison between simulation and plasma experiments Most comparison done with carbon, but the source and sinks uncertain Numerous impurity gas-puffing experiments exist, and more comparison needed Experiments with plasma interacting with liquids - CDX-U, Model improvements are needed Separate ion temperatures for hydrogen and impurities Verification of kinetic transport corrections Coupling to edge-plasma turbulence simulations for radial impurity transport - nondiffusive transport

5 Task I related Explore implementing flowing liquids in a major experimental physics device Color code: blue (ALPS funded) red (APEX funded) green (Moir/APEX) Continue coupling between UEDGE and WBC for C-MOD, NSTX edge-plasma for ALIST liquid-module design (LLNL, ANL) Analyze effect of scrape-off layer currents in edge region for DiMES probe (LLNL, GA)

6 Task II related Exploration of high-payoff exploratory concepts Edge-plasma modeling for CDX-U and benchmark with data; include core region (LLNL/ORNL) Determine impact on core impurity intrusion when impurity ions are thermally decoupled from hydrogen; applies especially to low recycling including the diverted flux-bundle concept of Kotschenreuther. Requires separate energy equations in UEDGE; benchmark with GA MCI code as available (LLNL,, Rensink; 10k; GA, Evans; also ALPS related) Analyze a consistent liquid-wall power plant based on the kinetic tandem mirror concept (LLNL, Moir; 15k) Continue assessment of high power capability of liquid drops and streams (LLNL, Moir; 5k).

7 Task III related - Integrated design of liquidwall system Analyze self-consistent mantle radiation by extending UEDGE core boundary inward; eventually couple with core transport code (LLNL,, Rensink 15k; also ALPS related) Continue divertor integration work for CLIFF to analyze plasma heat loads for geometrical variations; add helium and ExB drifts as time permits (LLNL,, Rensink, 10k) Assess impurities from Ga walls for CLIFF-like design (LLNL,, Rensink 5k) Modeling heat loads from ELMs in scrape-off layer for small, type III ELMs (LLNL,,?k) and large type I ELMs (ANL, Hassanein?k)

8 Future APEX-related work - A. Hassanein, ANL Horizontal plasma oscillations of liquid first-wall Study the effect of halo and induced currents Calculations of surface wave growth Corresponding aerosol formation Stability of liquid flow during normal operation Effect of flow instabilities and splashing Surface wave dynamics and growth

9 Future APEX-related work - T. Evans, GA Continue with Monte Carlo impurity ion transport in whole SOL Provide assessment of kinetic effects and benchmark corrections to fluid models Determine impact of toroidal asymmetries via 3D simulation

10 Large mantle impurity radiation is needed to give tolerable divertor heat loads even for liquids The Problem: General issue for MFE devices; brought to the fore in CLIFF work 2-3 GW fusion power results in MW of alpha power to remove UEDGE simulations give peak heat on orthog. divertor for SOL input: 40 MW gives 27 MW/m 2 80 MW gives 104 MW/m MW gives 186 MW/m 2 The approach: Extend UEDGE simulations to include mantle inside separatrix (expand from 1 cm to 10 cm) Investigate different impurities (Neon, Argon, Krypton) Core Outer midplane domain Mantle Separatrix SOL Since heat flux needs < 50 MW/m 2, more than 80% of alpha power must be radiated 10 cm Radial 6 cm

11 Low density, high temperature for low recycling plasma limits single T i assumption for all species Previous SOL fluid transport modeling uses a summed ion energy equation based on high collisionality Low recycling leads to low collisionality Resulting lower impurity temperature can reduce influx to core Separate T i equations have been implemented in UEDGE to assess this effect

12 Time history and width of energy deposition on divertor surface helps determine damage Electron conduction time is fast Ion conduction and convection are much slower [~(m e /m i ) 1/2 ] Experiments indicate substantial power comes out on the ion timescale In/out divertor asymmetries and parallel currents play important roles in which plate receives most of the energy

13 Significant ion ELM energy is transmitted to the electron channel via Coulomb collisions UEDGE model is used in 1D and 2D with kinetic corrections for heat transport & sheath loss ELM is simulated by energy and density injected at midplane for 200 microsec For the heat flux at the plate, a fast electron response is seen, but a second electron response from (T e -T i ) collisional coupling is observed Plate heat flux 200 ms ELM Base case Plate power deposition time is several times the input duration

14 Collisional coupling remains large for high power ELM Increasing ELM energy more clearly separates the electron and ion time scales The second-stage slower electron responses is caused by T i T e coupling Base case x6 ELM energy

15 Time-dependent ELM modeling is being done by UEDGE UEDGE considers SOL currents and ExB drifts Characteristic two-time-scale seen: electron conduction and ion flow Parallel currents can dominate heat flow, causing outboard or inboard peaking, depending on current direction ExB drift causes strong reverseflow near separatrix Can currents be manipulated to advantage?

16 Spherical tori have large variations of B-field in the edge-plasma region that can effect flows Low recycling by lithium (or external divertor) leads to weaker collisionality via lower n i and higher T i Ions stay hotter and have slower relaxation times than electrons Ions acquire significantly different T perp, T temperatures, giving -(T perp -T ) db/ds force on ions Impurity intrusion can be affected by the changes in the hydrogen edge plasma and their own T i anisotropy We have completed the coding and initial debugging for separate T per and T equations in UEDGE

17 Lithium contamination of core from NSTX module modeled by coupling UEDGE & WBC Heat and particle flux to module computed by UEDGE Temperature rise of Li surface from Ulrickson s model Sputtering of Li from U. Ill. composite model WBC calculates lithium source near the divertor plate UEDGE uses this Li source to calculate lithium density at core boundary

18 Core-edge lithium concentration from 40 cm module is ~0.2% for 6 MW case UEDGE takes lithium ion source from WBC 5 cm above the plate Lithium density vs. distance from plate along flux surface 0.3 cm in SOL Full SOL hydrogen/lithium plasma is then evolved to steady state with hydrogen core-edge density 4x10 19 For the planned 40 cm (toroidally) module, only 0.2% Li at core edge For full toroidal coverage with 13 times more module gives 2.5% Li Even for full coverage, Li SOL radiation is only 6.4x10 4 W, or 1% of core input power; thus, no need to iterate WBC/UEDGE here

19 UEDGE simulation domain extended radially inward to model the edge mantle region Use the single-null variant of ARIES-RS Mantle Inner boundary has normalized poloidal flux of 0.75 (1 on sep.) Consider fixed-fraction impurity models for Kr, Ar, & Ne Krypton has large radiation at both low and high T e Most radiation is in the divertorleg region owing to high densities there

20 Liquid wall research plans -Ralph Moir 1) Adapt and analyze liquid walls for the kinetic stabilized tandem mirror (K-S/T-M). This study will be patterned after that of the FRC and Spheromak efforts, covering the same issues such as developing models for surface temperature of the liquid and estimate the evaporative flux. The study will integrated plasma physics, fluid mechanics and fusion technology. Substantial work has been done on this new version of the tandem mirror: R. F. Post, Fusion Technology 39, 253 (2001).

21 The kinetic tandem mirror is asymmetric with MHD stabilization via injected ions at ends Liquid-walled version of kinetic stabilized tandem mirror* /10/2002 * R. F. Post, Fusion Technology 39, 253 (2001) Liquid-vacuum interface In Out Side-view End-view 5/13/2002

22 Moir plans cont.: 2) Liquid droplet divertors for tokamak-like applications This study involves the large heat transfer enhancement with streams of liquid droplets in a geometry relevant to a tokamak. A major issue will be the maximum impurity flux criterion in the case of an irregular surface of droplets rather than the familiar case of a smooth surface. Write a report on this work. Magnetic flux carries plasma to divertor Y BP X Jets in divertor plate plane (X=Z) Droplets Z BT