ICRF Breakdown Studies M.L. Garrett, S.J. Wukitch, and P. Koert, *MIT Science and Fusion Center, Cambridge, MA, USA *Work supported by USDOE DE-FC02-99ER54512

Summary Design Arc Trails Electrode Damage • ICRF breakdown experiment has been designed and tested Double-ridge waveguide designed and optimized for Alcator C-mod ICRF frequency of 78 MHz. Arc trails formed in several locations, both Local melting and cratering observed on both types of Cu electrode damage • Arcing achieved with energy deposition between 1J and 200J The ridges taper from end to center, creating a peaked at the middle of the on the central region of the copper ridges electrodes. However, arcing is not always isolated to central (30 m surface finish) • Arcing only occurs with applied DC magnetic field waveguide. and on the aluminum waveguide side walls. test sample region. Possible causes: • Resolution of arc location insufficient to quantify results: evaluation of complex These trails, different from the cratering fields and electrical contact will continue • Simulations demonstrate a high, peaked field in the center of the double ridged waveguide. shown to the far right, only occurred for • Possible multipactor condition present • Excitation of 100kW produces ~9.0MV/m at electrode location. E-field scales as sqrt(PIN ). longer duration higher energy pulses. • BeCu fingerstock electrical contact Typical energy deposition was ~ 100 J. Arc • Complex interaction of DC / RF magnetic field Double Ridge Waveguide Frequency Response trails have been observed to form as Introduction 0 cathode spots move across electrode One of the challenges of ICRF operation in tokamak plasmas is reliable operation at high -5 Simulated and measured transmission    surfaces with retrograde, F J B motion.5 . On Alcator C-Mod, we observe arcing among ICRF antenna components at electric -10 and reflection coefficients are in good r 304 Stainless steel This may explain the movement observed fields of 1.5 MV/m for E ||B. Electrical breakdown is influenced by the antenna material, -15 agreement. (6 m surface finish) RF -20 here. However, more simulation must be neutral pressure, ionizing radiation, and the tokamak magnetic field. While copper is typically -25 • Excellent match at design frequency done. 200 m utilized due to its superior conductivity, it has a relatively low melting temperature (1085 C), -30

Amplitude [dB] Amplitude (78.0 MHz) low tensile strength, low breakdown field (170 MV/m), and strong material displacement -35 S11 simulation after breakdown. The low melting temperature and tensile strength lead to a reduced voltage -40 S21 simulation Pitting observed along breakdown limit. The material displacement after breakdown is consistent with observations -45 S11 measured surface scratches on both S21 measured that the antenna voltage limits degrade with operation. Due to high tensile strength and 60 70 80 90 100 re-entrant copper 1-3 Frequency [MHz] melting temperature, refractory metals could improve voltage limits. A test stand has been electrodes Tungsten Arc trails on aluminum WG walls 1.0 cm designed and built to characterize ICRF relevant voltage breakdown. The vacuum test stand (6 m surface finish) structure is a double-ridged, tapered waveguide which creates a peaked electric field of 7.5 Electric Field Along Ridge 12 Upper Observed surface wave pattern in area MV/m at the electrode location. Effects of electrode material, surface structure, magnetic Re-entrant electrode adjacent to crater damage resembles Tapered electrode electrode field and neutral pressure are investigated. 10 Shown to the right are arc trails on the Tonks-Frenkel instability.6 upper and lower electrodes, following a 8 ~200 J discharge [70kW, 3ms]. Note the Lower anti-symmetry, both at the sliding electrode 6 Electric field evaluated along ridge centerline for both tapered and re-entrant electrode transition and between upper and lower

ICRF Breakdown 4 configurations faces. Electric Field [MV/m] Field Electric The ICRF Breakdown Experiment is a 2 tapered double-ridge waveguide which Sliding transition 0 is used to study ICRF relevant 0 0.5 1 The height of both the voltage breakdown. Longitudinal Position [m] central cylindrical sample DC Magnetic Field and electrode can be RF voltage breakdown occurred at different input power levels for different magnetic field orientations. No RF breakdown observed without an applied DC magnetic field. • Compact design Tapered electrode Re-entrant electrode adjusted to make small • High E-fields between ridges changes in the gap spacing Breakdown Threshold • Source sees matched load until and frequency response 180 breakdown event (no tuning) 160 SEM images reveal fine surface arc • Provides robust method to benchmark Cu (present ICRF 140 damage on copper tapered electrode strap material) against various refractory materials 120 • Variable B-field and neutral pressure 100 • Voltage/current probe, optical diagnostics to analyze breakdown – 80

quantify dark current, field emission data 60 Input Power [kW] Power Input BeCu fingerstock locations 40

20 Future work No Breakdown Observed Breakdown No 0 • 1 2 3 4 Eliminate BeCu fingerstock from central region Re-entrant electrode: Magnetic Configuration • Eliminate possible multipactor / phenomena Motivation • Sliding BeCu fingerstock joints • Concentrate magnetic flux density in smaller region—iron core / smaller magnets • BeCu fingerstock ring around test sample 0.8 • Improve modeling to better understand magnetic fields, particle trajectories Evidence of arcing on Alcator C-mod ICRF antennas. and cathode spot movement Alternate materials have the potential to increase Neodymium magnets below channel 0.6 voltage handling of ICRF antennas and striplines. to create up to ~ .3T field across gap

0.4 • Refractory metals: high tensile strength, high *1+ Laurent, L. L., “High gradient RF breakdown studies,” PhD Thesis, Dept. ECE, UC, Davis (2002). melting point materials may breakdown at higher 0.2 [2] Norem, J., et al, Phys. Rev. Spec. Top. Accel. Beams 6 (2003).

surface fields2,3 BeCu fingerstock for optimum [3] Taborelli, M., et al, Proc. EPAC 2004, Lucerne (2004). Magnetic Flux Density [T] Density Flux Magnetic • B-field: strong dependence on B-field orientation electrical contact between 0 [4] Wukitch, S. J., et al, Plasma Phys. and Cont. Fus. 46, 1479 (2004). (ICRF operational limit ~ 1.5MV/m for ERF ll B and Antenna strap ground surface electrode and ridge [5] Kajita, S., et al, Nucl. Fusion 49 (2009). 1 Configuration 1 Configuration 2 Configuration 3 Configuration 4 ~3.5MV/m for ERF ┴ B on Alcator C-mod damage from arcing [6] Insepov, Z., et al, Proc. PAC09 (2009). (No magnets)