Alcator C-Mod Weekly Highlights May 22, 2000 Plasma operations continued at Alcator C-Mod last week. A fresh boronization was carried out on Monday night; an average deposition layer of 2048 Angstroms was laid down, followed by 7 hours of ECDC in D2. Following the boronization, three run days were completed. A total of 56 plasmas were produced with a startup reliability of 75%, including the post-boronization recovery and conditioning day. Physics experiments were devoted to systematic studies of SOL fluctuations in L-mode, and an investigation of the quasi-coherent mode in ohmic EDA H-mode plasmas. Plasma Physics operation is scheduled to continue this week. Physics -------- A run was dedicated to an exploration of the dependence of fluctuations in the SOL and attendant cross-field particle transport on collisionality in the SOL (MP#267). Key diagnostics employed in this study included two fast scanning langmuir probes, operated in fluctuation mode; the divertor probes, also in fluctuation mode; fast visible diodes viewing the helium puff from the outboard midplane limiter-mounted capillary; tangential Lyman-alpha diode array; limiter-mounted particle flux probes; and edge Thomson scattering. A density scan was carried out at fixed current and toroidal field. The fluctuations recorded by the scanning probes clearly showed differences as the density was lowered. At low density, the fluctuation level at the separatrix appeared to be much lower, increasing with distance from the separatrix. As has been seen previously at low density, both mach probes record a strong 'reverse' flow, favoring higher Isat on the 'West' probe relative to the 'East' probe. It was noted that when this strong flow occurs, the down-stream probes of the Mach probes (East) see much more Isat fluctuation than the upstream probes (West). This is perhaps evidence that the 'wake' of the probe has a higher turbulence level. A run-day was devoted to study (MP#253) of the fluctuations characteristic of the so-called "EDA H-mode", which exhibits good energy confinement, moderate particle confinement and lack of large ELMs, and is therefore considered to represent a highly favorable regime for a tokamak reactor. The key to understanding EDA seems to be in the quasi-coherent fluctuations which have been observered by the reflectometer, Phase Contrast Interferometry (PCI), and probes. These fluctuations seem always to accompany EDA and have been shown to drive significant particle flux. The purpose of this experiment was to localize the mode relative to the density and temperature pedestals and to correlate both mode amplitude and convected flux with the local gradients. Because the range of the scanning probe is limited by incident heat flux, this run was carried out in ohmic EDA discharges at the lowest possible input power. This was done by reducing the plasma current and toroidal field, while keeping a fixed time-evolution of q. By fine tuning the Bt and Ip ramps, we were able to get shots with an L/H transition followed by an EDA transition. On several shots, we had probe plunges well inside the LCFS while in the EDA phase. With PCI and the reflectometer, the fluctuations were similar to those seen before. PCI saw k_R ~ 5-6 cm^-1. The A-side scanning probe saw considerably lower k_theta ~ 2-3 cm^-1; this result differs from previous experiments using the vertically scanning FSP, which observed k-values in the same range as the PCI. The quasi-coherent fluctuations were seen over a distance of ~7.5mm, but we were unable to access the inner edge of the mode with the probe. A new analysis provides qualitative spectroscopic evidence that the separatrix location at the outer midplane differs from that reported by the EFIT code. Previously, the SOL profiles as measured by the "midplane" diagnostics, specifically the radially scanning probe (ASP) and HeI line-ratio measurement have indicated that the outer separatrix location is ~5mm inside of its EFIT-determined location. This is the case even in plasmas with an L-mode edge. The spectroscopic evidence supporting these measurements comes from the observation that the atomic He puffed from the outboard gas jet should be almost fully ionized after passing through a line density of plasma of n_e x l=2x10^17 m^-2 [Schweer, et al., J. Nucl. Mater. 196-198 (1992),p174]. Assuming the Thomson edge density profile is mapped correctly by EFIT from the top of the plasma to the midplane, then too much atomic He is observed at radii inside of the n_e x l value quoted above. However the observation is consistent with this n_e x L value if the actual separatrix on the outboard midplane is ~5 mm INSIDE of the EFIT-determined one. ICRF Systems ------------ Further ICRF antenna plasma conditioning and antenna comparison experiments were performed this past week. Although the impurity production associated with the J-port antenna is greatly reduced compared to the last campaign, its heating efficiency appears to be less than that of the D-port antenna. A MW from J-port heats as effectively as 0.3-0.5 MW from D-port; a fresh boronization of the machine improved the comparison but the relative efficiency remains low. The Mo source rate from J-port is about 2-3 times that of D-Port at the same power, but the plasma core radiation increased by a much smaller factor. Antenna loading into plasma appears to be insensitive to antenna-plasma gap variations, suggesting parasitic loading is a major component of the antenna loading. A possible source of parasitic loading may be the antenna back feed straps. We are considering replumbing the antenna to reduce the parasitic load by isolating those elements with the suspected largest parasitic load. If the parasitic load is the cause of the poor heating efficiency and our analysis is correct, the heating efficiency should improve. In addition, we are planning for antenna modifications that should allow us to improve the overall antenna performance. The E-Port antenna was inoperable last week due to arcing inside the transmission line, near the machine. The transmission line has been repaired, the antenna vacuum conditioned, and E-port is now ready for resumed operation. DNB Systems ------------ The DNB was operated into plasmas on all three plasma-operation run days. The filament and snubber 480V inputs were switched to a new circuit. This switch provided a substantial improvement to the filament current during C-Mod pulses, essentially eliminating the portion of the arc current droop due to the filaments. Additional perveance scans were done at 34 kV and 37 kV. The gradient grid voltage was varied as a fraction of the accelerator grid voltage to determine its effect on accelerator current and perveance. The suppressor grid voltage was also varied to determine its effect for each of the gradient grid settings. The scans were done at 39 kV. Travel and Visitors ------------------- Y. Nakamura has arrived from Japan for a two week visit to continue the JAERI/MIT collaboration on disruption research. Our neutral point experiment is scheduled to be run on C-Mod later this week. Chris Watts of Univ. Auburn visited for two days. He helped with work on the e-mode heterodyne radiometer. Gary Taylor was at C-Mod for three days, installing faster digitizers and improving shielding on GPC2, which is now no longer bothered by RF pickup from the J-Port antenna. David Mikkelsen spent the week here for discussions and work on transport modeling.