Alcator C-Mod Weekly Highlights Oct 25, 1999 Plasma operation resumed last week at Alcator C-Mod. Magnet testing was carried out on Monday, followed by a series of plasma shots. Plasma runs were carried out Tuesday-Friday. This week's runs were devoted to ICRF conditioning, evaluation of the performance of the J-port antenna, and exploration of the sensitivity of heating to minority (H) concentration. A total of 64 plasma shots were produced, with a start-up reliability of nearly 90%. Following cooldown, a series of TF magnet tests were carried out to evaluate the single-turn resistance measurements in the vicinity of H-port. Prior to the recent warm-up these measurements had included some anomalously high readings at low current (<1kA) , while measurements performed during high-current pulses (>50kA) showed no unusual features. Initial low-current measurements on Monday were nominal, as were those during high-current pulses. A modification to the normal TF current waveform was adopted which provides a 50KA plateau at the end of the shot to allow continued monitoring of the resistances during plasma operations. The J-port antenna was operated with the new feed configuration, which permits adjacent current straps to be driven 180 degrees out of phase. Both 0,pi,pi,0 and 0,pi,0,pi phasing were used in these experiments. A phase feedback circuit operated successfully to maintain the programmed phasing during plasma operation. The cavity modifications permitted increased power from FMIT#1 & 2 (D- and E-port antennas), with a maximum of about 1.8MW injected from each system; the limit on coupled power from these systems is now set by the voltage at the antennas, rather than the source. With all transmitters operating simultaneously, a total of over 5MW was injected for a short time. On Thursday a study of heating and impurity generation from each antenna was carried out. For the same power levels, the radiated power associated with the J-port antenna is higher than for either D- of E-port. Central heating appeared to be similar, but the temperature rise in the edge region is smaller with J-port, and it was more difficult to access the H-mode and EDA regimes. The impurity influx from J-port appears to be dominated by titanium, which presumably comes from the TiC coating on the Faraday screens and/or antenna protection tiles. Further aggressive antenna conditioning in plasma might help to alleviate this problem. On Friday we carried out a systematic study of the ICRF heating efficiency as a function of minority (H) concentration. The H/D ratio was determined spectroscopically using Balmer alpha emission, which provides a measure of the concentration in the plasma edge and SOL. D- and E-port antennas were used to inject about 2.5MW into the plasma, with the power waveforms modulated to allow the absorbed power to be determined by the break-in-slope technique. Best results were obtained at the lowest H concentration obtainable with no H2-puff, following overnight ECDC in D2; this condition corresponded to a measured H/D of 2.5%. H2 gas puffing was used to increase the concentration up to >20% during the course of these experiments. Physics and Analysis -------------------- Preliminary analysis of fluctuation data recorded during a recent ohmic H-mode run by the A-port and F-port scanning probes (ASP, FSP) indicates the presence of a coherent ~100 kHz mode near the separatrix during EDA H-modes. This mode is seen on both the ASP and FSP probes over approximately 1 mm of their final travel when they are scanned up to or slightly inside the last closed flux surface. At the same time that the probes detect a coherent oscillation, the reflectometer and phase-contrast imaging systems also see a coherent oscillation at the same frequency (+/- 10%). The oscillation is most noticeable on ion saturation current signals but is also seen on floating potential signals. The mode appears to propagate in the electron diagmagnetic direction. Fluctuation-induced particle fluxes are estimated from the FSP probe set, using two probes as plasma potential fluctuation monitors and one probe as a density fluctuation monitor. Cross-field particle transport is clearly enhanced by the presence of the ~100 kHz coherent fluctuation, suggesting that this mode is responsible for the reduced particle confinement time and enhanced D-alpha seen in EDA H-modes. In contrast, no coherent oscillations are seen in ELM-free H-modes. During the ELM-free phase, ion saturation current fluctuation levels normalized to the mean value decrease dramatically as the probe traverses the scrape-off layer towards the separatrix. Fluctuation-induced particle fluxes correspondingly decrease as the separatrix is approached. The magnitude of the fluctuation-induced particle transport in the scrape-off layer is consistent with simple estimates from ionization measurements and flows to limiter/divertor surfaces. A new edge fluctuation diagnostic is operational this run campaign. It relies on visible light emitted at the plasma boundary and detected by sensitive diodes, whose photo-current is amplified by a circuit with high-gain out to high frequencies (~250 kHz). The detection system is similar to the U. of Texas BES system also operating on C-Mod and described in earlier Weekly Reports. The experimental design is similar to the BES experiment in its attempt to relate fluctuations in light emission to fluctuations in electron density. Like the BES system, this new diagnostic has essentially-toroidal views of intrinsic emission. However it can also view a gas jet near the plasma boundary with extremely high radial (~3 mm) resolution. There are three detection systems allowing for three views at any one time. The gas jet and the fiber optic views are part of another experiment. The gas jet allows for a local, non-chord-averaged measurement. By puffing D2 from the jet, we observe that the relative fluctuation amplitude from the local measurement is increased by a factor of two compared to the chord-averaged measurement. (The D_alpha brightness was increased by a factor of four as well.) The observed RMS variation in the D_alpha emission from a D2 puff is typically 15-20% of the average emission in L-mode plasmas. The fluctuations are NOT a result of photon-counting statistics. The frequency spectrum of the D_alpha emission fluctuations from the puff is broadband out to about 100 kHz, dropping strongly at frequencies > 100 kHz. Since there is typically no intrinsic He in C-Mod, all of the measured signal is from the local He emission when He is puffed from the jet. The RMS fluctuation in the intensity of the neutral He 5876 A line is also measured to be 15-20% in L-mode. By looking at cross-correlations in the fluctuations from radially separated views we plan to measure the radial size of the fluctuations and to see if they propagate radially. These experiments have benefited from help and guidance from Stewart Zweben at PPPL and are part of an MIT/PPPL collaboration. Dominque Heubner, a summer student from the Univ. of Wurzburg in Germany, was instrumental in the experimental construction. Travel and Visitors ------------------- Miklos Porkolab attended the Fusion Power Associates Meeting in Washington, Oct. 19-21, and gave the talk "Advanced Tokamak: Promise and Near Term Plans". The topic of the meeting was : Fusion Power for the 21st Century: Science and Technology for the New Millennium". Last week C S Pitcher attended the Fourth International IEA Workshop on Helium Transport and Exhaust in Fusion Devices in Georgia. The workshop brought together approximately 40 researchers from the US, EU and Japan. Pitcher presented preliminary results from the divertor bypass experiments on Alcator C-Mod. The topics of the workshop included RI discharges, the behavior of helium and other recycling gases in divertors, edge fluid and Monte Carlo modelling, core helium transport, helium exhaust in ST's and stellarators, ergodic divertors and means of enhancing helium core transport using RF techniques. The following programs were represented, ITER, JT-60U, ASDEX-Upgrade, DIII-D, TFTR, JET, TEXTOR and C-Mod. Dave Mikkelsen of PPPL was on-site C-Mod last week to work on transport modeling.