C-Mod ICRF Program. Alcator C-Mod PAC Meeting January 25-27, 2006 MIT Cambridge MA. Presented by S.J. Wukitch

Transcription

1 C-Mod ICRF Program Alcator C-Mod PAC Meeting January 5-7, 006 MIT Cambridge MA Presented by S.J. Wukitch Outline: 1. Overview of ICRF program. Antenna performance evaluation and coupling 3. Mode conversion physics and Current Drive

2 Context of the C-Mod ICRF Program ICRF provides the bulk auxiliary heating power on C-Mod. Antenna power density is similar to ITER requirements. Standard scenarios are fundamental and second harmonic minority ion cyclotron.» ITER will utilize fundamental 3 He and second harmonic tritium. H minority has single pass absorption similar to expected in ITER. Participate in ICRF heating and current drive benchmarking/modeling one of the International Team High Priority Physics items assigned to the US. The antenna is the most critical element to the success of ITER s initial 0 MW of ICRF heating power. Understanding the underlying physics that limits antenna power and voltage handling is critical. RF-plasma edge interaction is a key issue with respect to impurity production, enhanced sputtering, and localized hot spots.

3 Overview of the ICRF System D & E Antenna J antenna Thru 006 Frequency Power Antenna Phase D & E-Antenna ~ 80 MHz x MW x Strap fixed J-Antenna MHz 4 MW 4 Strap variable

4 C-Mod ICRF Operation Overview Achieved record stored energy and high performance with metallic limiters. Systems operated fairly reliably. Delivered 4-5 MW routinely before dust event. Lost MW source as a result of Ti dust in D antenna (early May). Lost MW source when T/R set had shorted rectifier stack (8/16) P RF (MW) W MHD (MJ) R DD (x10 14 s -1 ) 3 1 T e0 (kev) n e (x10 0 m -3 ) P Rad (MW) B T =5.4 T, I P =1 MA Time (sec)

5 ICRF Antenna Research Future C-Mod program will require 6 MW through ports and of course needs to be reliable and fault free. Issues regarding antenna operation are applicable to other present day, ITER, and future experiments. Goal: Develop reliable antenna coupling with minimum negative impact on the plasma. Move antenna design from art form to one based on scientific principles and good engineering practice. RF-plasma edge interaction Impurity production is particularly important for machines with metallic first wall. Antenna operation/performance issues: Power density and pulse length limits. Neutral pressure limits (may be the underlying physics of antenna ELM interaction). Long pulse fault free operation is desired what is the nature of antenna arcs?

6 Experimental Benchmarking of TOPICA Γ reflection Comparison of E Reflection Coefficient J Model Discharge Measured (green) TOPICA (blue) TOPICA is a sophisticated and relatively new electromagnetic solver that can simulate antennas in plasmas. Has been benchmarked in vacuum against other codes. Imports antenna CAD. Coupled to 1-D full wave solver. Includes transmission line network. Will need to have modules for arcing, multipactor, RF plasma edge interactions. Collected new experimental data to benchmark antenna characteristics with plasma. Need to compare simulation and experiment. APS 05 R. Maggiora - Politecnico di Torino

7 RF-Plasma Edge Interaction: Boronization Erosion Standard boronization degrades with ~0-50 MJ RF Corresponds to 0-30 high RF power plasmas). Mo levels rose shot by shot, presumably because Mo surfaces exposed following the B erosion Shot number Some RF plasma edge mechanism is probable ohmic H-modes with similar input energy had slower performance degradation. Enhanced sputtering due to RF sheaths that result in higher ion energies? Enhanced sputtering rate resulting from increased radial transport from convective cells? How sensitive to single pass absorption strength? W MHD (kj) P RF = 4 MW P RF = 3 MW

8 RF Plasma Edge Interaction: Plans Investigate boronization lifetime dependence on antenna phasing and absorption mechanism. Antenna phasing should influence sheath effects.»[0,π,π,0], [0,±π/, π,±3π/] and [0,0,0,0] phasing should have significantly stronger sheath effects than [0,π,0,π]. Lower single pass absorption by scanning H concentration. Eliminate sheath effects with insulating limiters. Si 3 N 4 and pyrolytic BN have been identified as candidate materials. Additional diagnostics: Resurrect capacitive probes to measure plasma potential. Filament imaging on field lines connected to antenna. Measure local density profile near antenna (reflectometry).

9 Antenna Performance Issues: Neutral Pressure Limit Practical Consideration: D and E antenna neutral limit is ~1 mtorr. 3 1 J antenna Neutral pressure limit 0.8 for J antenna is ~ mtorr. Neutral Pressure Limit 0.4 Impacts machine operation: 0. Limits initial target 0 densities, Time (sec) And slows boronization recovery.» Neutral pressure is higher following boronization for a given target density. Wider implication is that this phenomena may be the underlying physical explanation for antenna arcing with ELM s. P neutral (mtorr) P RF (MW) T. Graves APS 05

10 Antenna Performance: Plans Status: Multipactor initiated discharge is responsible for neutral pressure limit.» Neutral pressure lowers antenna multipactor susceptibility.» B-field further modifies the antenna s multipactor susceptibility. Tests indicate reducing secondary electron coefficient (SEC) can modify or eliminate this limit. Plans: Implement modified SEC material on J antenna. Investigate limitations on 6 MW discharges. Upgrade RF test stand to include 0.1 T B-field to investigate:» initial RF trip at neutral pressure limit,» Role of B-field in lowering the onset of multipactor induced discharge, and» Influence B-field has on high voltage breakdown. Obtain and modify multipactor and breakdown codes to include neutrals and B-fields.

11 Experiments to Investigate Faraday Screen-less Operation Motivation: Faraday screen can be significant source of impurities. Faraday screen thermal and disruption loads can result in antenna failure. Antenna design can be simplified without screen. Mixed results from operation without Faraday screen. Successful operation without FS on TEXTOR, Phaedrus-T and ASDEX-U Unsuccessful operation without FS on DIII-D. Purpose of FS is thought to be two fold: Prevent plasma from entering antenna box. Set wave polarization. What role does a Faraday screen play in antenna operation?

12 Replaced FS with Slotted Mo Septums septums Straps Mo septums added to prevent plasma from entering the antenna box. Plasma is scraped off and has short decay length (~3 mm) in shadow of limiter. Septum design balanced RF against plasma transparency.

13 Faraday Screen-less Assessment and Plans Status: Loading, voltage and power handling were unchanged unchanged. Heating effectiveness was decreased ~10% in L-mode and 15-0% in H-mode discharges. relative Cu density that shows a strong correlation with J operation and power level RF Power (MW) 4 5 an interaction was observed where the strap ground and feed pass near the middle of the antenna antenna resilience to boronization was remarkably decreased and the nominal neutral pressure limit was reduced. Plans: Model antenna with and without screen to compare field distributions. Change antenna strap to minimize radial fields. Relative Cu Density J-Port E-Port D+E-Port M. Reinke

14 Absorption and Propagation: D( 3 He) Minority Heating Primary ICRH scenario for 8 T discharges. May anticipate greater impurity production due to lower single pass absorption for D( 3 He) than D(H). Status: Confirmed previous greater sensitivity to 3 He concentration than D(H). Found L-mode heating effectiveness and H-mode threshold were similar to D(H) heating T e0 [kev] P rad [MW] B T =5.1 T, I P =1.0 MA D( 3 He) D(H) P RF [MW] H-mode J Antenna Threshold E Antenna W MHD [MJ] n line [x10 0 m - ] D α [a.u.] Time [sec]

15 5. T, D( 3 He) Heated H-modes Limited number of discharges investigated thus far. Good H-modes obtain >165 kj for 3 MW. Best from D( 3 He) discharges are kj. Direct H-mode comparison compromised by loss of D antenna. Other differences compared to D(H) heated H-modes. Boronization appeared to erode more quickly. Between shot boronization did not improve discharge performance B T =5. T, I P =1 MA P RF (MW) W MHD (MJ) T e0 (kev) n e (x10 0 m -3 ) R DD (x10 14 s -1 ) P Rad (MW) Time (sec)

16 Parasitic Absorption could reduce H-Mode Performance H-mode performance can be greatly affected by radiated power fraction. Power absorbed by parasitic mechanism could result in increased impurity production. D( 3 He) scenario has additional cyclotron resonances including minority B and fundamental majority D. Use of Be in ITER will result in similar situation. B can reach 1-% post boronization R major [m] Additional resonances are not present in D(H) minority. RF fields fill the torus for D( 3 He) compared to D(H). Far field sheaths could become more important than in the case of D(H) f=80 MHz, B T =5.4 T, I P =1 MA f ch R major [m] f c11b f=50 MHz, B T =5 T, I P =1 MA f cd f c3he

17 Plans: D( 3 He) Minority Heating Scan B-field to move 11 B and D resonance out or into core of the plasma. Investigate boronization lifetime dependence on 3 He concentration (minority versus mode conversion). Limited versus diverted electron power deposition profiles in minority and mode conversion scenario. far field sheaths should be more important in limited discharges. Investigate whether between shot boronization maintains optimal plasma performance. Investigate high power, high performance at 8 T with D( 3 He) minority heating under optimized conditions.

18 Absorption and Propagation: ICRF Mode Conversion Long wavelength Fast wave (~10 cm) converts to two short wavelength modes Ion Cyclotron (1- cm) and Ion Bernstein (~0.3 cm) waves. Have potential to provide» localized pressure, current and flow profile control and» synergistically interact with LH waves. Validate physics and computational models thru comparison of experiments and simulations: Electron power deposition profile measured by FRCECE And PCI wave measurements. Details of mode converted wave propagation and absorption can influence the current or flow drive efficiency. Rapid k up-shift could result in loss of spectrum control and reduce current drive efficiency. Strong shear flow can be obtained if significant mode converted power is absorbed near the ion cyclotron resonance.

19 Shown Good Agreement between Experiment and Simulation Status: Simulated perturbed density profile agrees with experiment. Suggests TORIC physics kernel and computation algorithm are describing the mode conversion physics properly. Power deposition profile also agrees with experiment. Plans: Probe up-down asymmetry by varying poloidal launch position. Use masking of the phase plate to allow localization of the modes. Investigate experimental and simulation amplitudes disparities. [a.u.] [a.u.] [a.u.] ~ Re( ne dl) ~ Im( ne dl) ~ ne dl Experimental Synthetic MC layer R (m)

20 Current Drive: Mode Conversion Current Drive Mode converted waves damp primarily on electrons near the mode conversion layer. Could be used to control sawteeth. Initial k spectrum imparted by the antenna is lost due to rapid parallel wavenumber upshifts and downshifts. Initial experiments found sawtooth period evolution consistent with localized current drive from mode converted waves. Using TORIC + Ehst-Karney parameterization, a model scenario where ~100 ka is driven for 3 MW RF power was identified. Since wave interacts with electrons near the trapped-passing boundary, calculation details for driven current could be complicated. Fokker Planck solver needs to treat boundary region accurately. Wave polarization is another variable loosely accounted for in parameterization. P RF (MW) 1 D-port D+J-port D-port D+J-port j (MW/m ) T e0 (kev) Time [sec] ctr-cd co-cd Time [sec] Total driven current for 1 MW IBW + ICW + FW 33 ka/mw IBW Contribution (x10) Ohmic profile current for modeled discharge (EFIT) r/a

21 Mode Conversion Current Drive: Experimental Status and Plans Status: Clear sawtooth variation with phase and deposition location. Sawteeth complicate driven current determined loop voltage. Heating phase ST behavior could be a result of inherent asymmetry due to mode conversion.» Modeling also consistently suggests heating phase will have some driven current. Plans: Use ramp discharges (sawtooth free period) to deduce driven current from loop voltage. Revisit with increased auxiliary power and better conditioned machine. A. Parisot APS 05

22 Mode Conversion Current Drive Simulations: Status and Plans Status: Preliminary results from Fokker- Planck simulations show differences with Ehst-Karney predictions.» About 0% decrease in peak current density and» Near elimination of reverse current. Plans: Utilize another Fokker-Planck (DKE) code to verify results.» Code is specialized to deal with trapped-passing boundary. Implement Porcelli ST model in TRANSP to model ST period.» May allow estimate of driven current.

23 Plan Summary Antenna coupling: Investigate RF role in boronization layer erosion. Modify antenna SEC to reduce or eliminate neutral pressure limit. Continue to benchmark TOPICA with experimental antenna measurements. Wave Propagation and Absorption: D( 3 He) minority heating Directly compare H-mode discharges heated by D( 3 He) to D(H). Evaluate role of parasitic ion absorption. Use results to investigate 8 T operation. Wave Propagation and Absorption: Mode conversion Test TORIC physics model with poloidal launch. Current Drive: Investigate mode conversion current drive Identify experimental conditions where significant MCCD can be obtained. Couple additional Fokker Planck package to TORIC to investigate MCCD.

24 Reference Material: Summary of Tasks RF Sources: Upgrade ICRF transmitter control and system reliability. Monitor tube lifetime issues. Matching network: Test and implement prototype fast ferrite matching network. Develop compact, high capacitive DC break. Antennas: Neutral pressure limit with modified SEC material. Eliminate RF sheaths through insulating materials. Investigate screen-less antenna performance with shielded radial feeders. Codes: TOPICA (now on MARSHALL) 3-D modeling of ICRF antenna code (U. Torino). TORIC (on MARSHALL) coupled with Fokker Planck code for current drive calculation. (Sci-DAC) Finite banana width Fokker Planck code with self consistent RF wave fields. (Sci-DAC initiative) Diagnostics: Upgrade PCI for localization. Implement active charge exchange for H minority energy distribution.

25 C-Mod ICRF Research Area Antenna (coupling) Propagation and absorption (heating) Current drive C-Mod Investigate antenna operation with metallic limiters and without a Faraday screen. RF role in erosion of boronization layer and enhanced sputtering. Influence of neutrals and B-field on antenna voltage and performance. Benchmark antenna analysis code. Details of mode conversion Core RF wave diagnostics: phase contrast imaging and compact neutral particle analyzer Compatibility of strong and weak single pass absorption with high performance Benchmark full wave and finite banana width Fokker Planck codes Localized mode conversion for sawtooth control Inclusion of Fokker Planck solver to predict driven current Fast wave below all cyclotron frequencies Investigate ion cyclotron current drive Other tokamaks NSTX: parametric decay Other tokamaks are dominated by ELMs DIII-D: double layer screen JET: load tolerant antenna design ASDEX-U has program to investigate RF breakdown Tore Supra concentrates on long pulse issues (excellent set of IR cameras) JET role of phasing and multiple frequency on heating effectiveness NSTX and DIII-D have concentrated upon investigating high harmonic fast waves absorption on electrons and fast ions at high harmonics. DIII-D and NSTX concentrated on fast wave current drive at 4-8 cyclotron harmonic. JET has done significant work on ion cyclotron current drive

26 Priorites Compare Favorably with ITPA and Scientific Priorities ITPA Experimentally test TORIC s physics kernel and computational algorithm for both heating and current drive. Continue experiments to investigate sawtooth stabilization and control and test developing codes. Develop real time current profile control using heating and CD actuators: assess predictability, in particular for off-axis CD. MDC5- Sawtooth stabilization for NTM control. Scientific Priorities Test TORIC s physics kernel and computational algorithm thru comparison of wave measurements with simulations. Investigate MCCD. Identifying operational space under which monster and small sawteeth are obtained. Identify physics that sets sawtooth period. Compare experimental sawtooth data with sawtooth models Understanding the propagation of waves and their nonlinear interactions with plasmas will lead to new techniques to control plasma behavior, and will be key to optimizing the conditions for burning of the fusion fuel. The interaction of the energetic particle population with the background plasma is a complex process, and a basic understanding of this interaction is critical to practical applications of fusion.

27 Antenna Performance: Power and Pulse Limits Have achieved antenna power densities of: 1.5 MW (10 MW/m ) for D and E antennas and 3 MW (11 MW/m ) for J antenna. Discharges with 6 MW of ICRF power have been achieved sec at 6 MW 0.7 sec at 5. MW 1 sec at 5 MW Would like to achieve fault free operation. Faults are instances when the reflected to forward power ratio exceeds 5%. Need to understand physical mechanism responsible for faults P RF [MW] W MHD [MJ] n line [x10 0 m -] B T =5. T I P =0.8 MA T e0 [kev] Time [sec] 0.35 sec, 6 MW L-mode discharge achieved during machine conditioning.

28 Antenna Performance: Metallic Limiters Observations: RF plasma interaction along the antenna face was tolerable.» Increased interaction on the Faraday screen.» Operated in current drive phases without impurity events as well. RF interaction localized to upper side protection tiles. Results: Achieved record stored energy and high performance with metallic limiters. Suggests tile alignment is more important than septum. Plasma Backlit

29 Antenna Performance Issues: RF Ground Status: Vessel damage suggests RF supported discharge.» Potential source of Fe and Ni. Verified RF ground prevented vessel damage. Not unique to C-Mod antennas, noticed similar damage on DIII-D 0 and 180 FW antennas. Plans: Model antenna to estimate RF voltage developed using TOPICA. Identify RF discharge character.» Is this multipactor initiated? up

30 Bench top Experiment (CMX) Demonstrated Neutral Pressure Modifies Multipactor Characteristics At 1- mtorr, multipactor-induced glow discharge in coaxial geometry is observed and is well below nominal Paschen pressure. Once discharge is established, power must be removed to quench discharge.

31 Magnetic Field Modifies Antenna Multipactor Susceptibility Limit Limit Voltage handling decreases with increasing pressure. Total loss of power at glow onset Magnetized J-antenna multipactor-induced glow discharge at 0.5 mtorr Magnetized E-antenna multipactor-induced glow discharge at 1 mtorr. Remarkable agreement with experimentally observed neutral pressure limits.