The primary activities at Alcator C-Mod during the second quarter of FY99 were: execution of the present run campaign, analysis of data from the campaign, preparation of the DNB, and continued participation in the critical assessment of fusion science. A more detailed account of these activities follows.

Scientific Results

Core Confinement and Edge Pedestal Results

Measurements of large plasma rotation in C-Mod, in the absence of direct momentum inputs, but with ICRF auxiliary heating, have been reported extensively in previous reports. Recently core plasma rotation has been observed to change from the counter- to co-current direction during the transition from L- to H-mode, in plasmas that are heated purely ohmically. This ohmic case is shown is Fig. 1. Measurements of Doppler-shifts of highly charged argon x-ray lines, viewed along a tangential chord through the plasma center, provide measurements of the E×B velocity. Simultaneously, the rotation of sawtooth postcursors, observed on magnetic diagnostics, provides an independent measurement of rotation, characteristic of the q ~ 1 radius, but possibly systematically offset by diamagnetic effects from the E×B velocity itself. The changes across the L-H transition of the toroidal velocities from the two measurements are quantitatively similar. The magnitude of the change is observed to be typically of order 30 km/s. This velocity change, which corresponds to roughly 10% of the sound speed, and an E_r of about +15 kV/m (at q = 1, ignoring v_q×B_t), is consistent with the previously documented [J.E.Rice, et al, Nucl. Fusion 38 (1998) 75] scaling for rotation in ICRF-heated H-modes. The rotation in this ohmic experiment is obviously not an RF effect. Its quantitative consistency with ICRF experiments strongly suggests that the observed rotation is not caused by an RF effect, but is controlled by transport, even in RF-heated H-modes.

---The absolute density calibration has been applied to the core Thomson scattering measurements, and core plasma density profiles have been obtained for L-mode, various types of H-modes, L-H and H-L transitions. Hollow density profiles were observed at L-H transition, with edge density growing faster than the core density, and sharp peaking of the profile was observed at the collapse of the H-mode.

---One run during this campaign was devoted to the careful measurement of conditions just before the L-to-H threshold in a range of conditions. Both scanning probe data and B_t sweeps were obtained shortly before the transition. Measurements from the soft x-ray edge arrays and edge Thomson scattering system were used as well. A current scan was carried out at fixed density, up to 1.2 MA. The L-to-H threshold was at an RF power of 1.2 MW at 1.0 MA and only 0.56 MW at 1.2 MA. Again, L-mode T_e profiles were obtained close to threshold. It appears there is at most a weak I_p dependence of the T_e threshold. A density scan at 1.2 MA was also performed. Several high density points, up to n_e = 2.5 ×10²⁰m^-3, were obtained. A surprise was that even a small density increase, from 1.6 to 2.0×10²⁰m^-3, caused a large increase in the L-to-H power threshold (to 1.75 MW from 0.5 MW). This effect may have been related to the large gas fueling required and/or high neutral pressures. The threshold edge temperatures seemed quite constant with density. The power threshold was 2.2 MW at the highest density.

Edge Pedestals

A comparison of temperature, density, and x-ray emissivity profiles in the pedestal region of an Enhanced D_a H-mode plasma is shown in Figure 2. The temperature profile is show in the first panel. The ECE temperature profiles are measured using small B_t ramps and are consistent with those from the last campaign, with typical widths of ~ 11 mm, and with the foot of the pedestal located at the separatrix. During this run campaign the temperature pedestal heights are moderate ( ~ 350 eV) and vary with input power. The ECE pedestal temperature profile is compared with that measured by a new edge Thomson scattering system. The edge Thomson temperature profiles are in reasonable agreement with those from ECE, but are noisier, since the larger temperatures obtained in H-mode are above the range for which the spectrometer was optimized. The pedestal density profile measured by this system is shown in the second panel of Figure 2. A clear density pedestal is observed. Its typical width is 4-6 mm, centered on the separatrix.

Three additional edge profiles are shown in the third panel. Two are the soft x-ray emission profiles, the third is the total radiated power. The x-ray emissivities are measured by two high resolution arrays, one of which views the edge plasma across the top, the other of which views down the outside edge of the plasma. (Both are mapped to midplane coordinates along surfaces of constant flux.) A pronounced pedestal is seen on both arrays, as expected. The center of the pedestal is 10-12 mm inside the separatrix at the outboard midplane, consistent with previous measurements at the outboard midplane. However, it is located within 2 mm of the separatrix at the top of the plasma. This implies either that the x-ray emissivity varies significantly on EFIT flux surfaces just inside the separatrix or that there are substantial systematic offsets in the EFIT reconstruction. The possibility of the former interpretation is supported by previous observations by J. Rice et al., showing large up-down asymmetries in x-ray emission from hydrogen-like argon just inside the last closed flux surface. Our conclusions based on these observations are:

1

- the density and x-ray emissivity profiles have much narrower pedestal widths than the temperature profile,

2

- the density and x-ray emissivity profiles at the top of the plasma have a similar position and width.

---New, high spatial resolution, magnetic fluctuation measurements of Type III ELM precursors have verified that the toroidal and poloidal mode numbers are typically n = 10, m £ 16 with oscillation frequencies of 160 kHz. The modes are strongly ballooning in the sense that they are visible only on the outboard coils and not on the inboard coils even though the rational surface is closer to the inboard coils than to the outboard coils. Since the fluctuations are measured on outboard coils that are beneath molybdenum tiles, just like the inboard coils, the lack of observable oscillations on the inboard coils is not an effect of attenuation at high frequency due to the molybdenum tiles.

---After the first boronization of this campaign, we observed the following changes in the carbon and oxygen impurities: The carbon concentration was reduced from the pre-boronization level by a factor of ~ 2.7, and the oxygen concentration was reduced by a factor of ~ 6. These measurements were made spectroscopically by looking at the impurity resonance lines. While the carbon and oxygen emissions were decreased by the boronization, the atomic deuterium emission from the divertor region increased after boronization, typically by a factor of ~ 2.5 (at the same plasma density).

Divertor and Edge Results

Ten sets of mechanical flaps which, when open, allow gas conduction from the private flux region of the divertor to the region outside the main plasma were installed for this run campaign. The open-or-closed state of these bypass flaps can be changed as desired in ~ 50 ms. Estimates of conductances and neutral particle pressures indicate that the flows set up with the bypass flaps open are quite large and should directly modify neutral particle flows between the divertor volume and main chamber volume. These bypass flaps are being used routinely. It has been shown that the quality and character of H-mode plasmas are unchanged by the bypass state, even though the divertor pressure drops by a factor of two when the flaps are open. The midplane pressure rises slightly. These effects are shown in Figure 3. It is also clear from Figure 3 that the state of the bypass strongly influences the divertor plasma (the neutrals and the temperatures and densities) and influences to some degree the main plasma scrape-off-layer, e.g. SOL D_a and bulk flows. As discussed in the next paragraph, the retention of impurities in the divertor is also affected.

The measurements of divertor compression (i.e. the ratio of impurity atoms in the divertor to impurity ions in the main plasma) and enrichment (i.e. the ratio of impurity compression to deuterium compression) have been extended. Impurity gases in addition to argon have been used and the effects of the divertor bypass on impurity behavior have been investigated. A density scan with puffed argon and krypton in Ohmic plasmas was performed, starting with low density and increasing until the detachment threshold, ~ 2.2 ×10 ²⁰ m^-3, was passed. At an intermediate density, the divertor bypass was cycled. The scaling of argon compression with density is similar to that obtained in the previous run campaign, although the absolute values are reduced by a factor of ~ 2, probably as a result of a `leakier' divertor with the (closed) bypass installed. The krypton data are being analyzed. As shown in fourth panel of Figure 3, the divertor bypass does indeed have an effect on the impurity behavior. The difference on the argon and krypton compression with the bypass open and bypass closed is a factor of about two. The compression is higher with the bypass closed. This is due to less argon/krypton in the core and more argon/krypton in the divertor. Opening and closing the bypass during a discharge reproduces these results dynamically.

---The tangentially viewing divertor camera system is installed and operating. The new system contains two cameras viewing the divertor tangentially, each filtered for a different deuterium Balmer spectral line (e.g. D_a and D_g). Images from recent shots have been analyzed and compared to the previous campaign. Most inversions of images show features very similar to those from the last campaign, although the spatial resolution from the recent inversions is now approximately 5mm, a factor of about 2 better than the resolution in previous inversions. The typical features are: strong emission on the inner divertor ``nose,'' and strong emission along the outer divertor leg when the outer divertor is detached. However, images recorded during Ohmic L-mode shots imply the existence of a plasma in the private flux region of sufficient temperature and density as to dominate the deuterium emission. This observation is confirmed by the divertor Langmuir probes showing a higher plasma density below the inner strikepoint than nearly on it. Analysis of the mechanism generating the private flux plasma is continuing.

RF Research

Because of the trade-off between ICRF wave absorption and the minority tail formation, an H/D ratio in the range from 5 to 10 % is necessary for good RF heating using the H minority in deuterium (majority) plasmas at a toroidal field of 5.4 T. For this run campaign a spectrograph has been dedicated to making routine H/D ratio measurements using Balmer_a emission. Early in the campaign, during the ``clean-up'' phase, it was observed that wall fueling is an important particle source, and the levels of hydrogen and deuterium in the recycling gases are approximately equal in this phase. After the clean-up phase, the hydrogen fraction has decreased to levels somewhat greater than the optimum.

--- The location of peak electron heating due to mode conversion of the injected RF wave is directly related to how much of the minority ion species is present. To improve efforts in estimating the Helium-3 fraction in D(³He) plasmas, the predictions of such peak heating from TORIC, a full wave code including temperature effects, were compared with a cold plasma model. The comparison indicates that the simpler cold plasma model is adequate for concentrations above about 15%, but differs from TORIC by up to 1.2 cm (out of a 22 cm minor radius) for lower concentrations.

Operations and Diagnostics

Alcator C-Mod began the CY99 run campaign in late January. Full duration discharges were being obtained in the first week of February. High plasma start-up reliability has been maintained since.

Although the RF power systems for the old antennas were brought on-line quite rapidly this campaign (3 MW were delivered to the antennas one day after reliable plasma operation began), as the run continued, reliability and power delivery capability of the D and E-port ICRF systems decreased. Arcing in the transmission lines to both D-port and E-port was identified. The antenna conditioning problems were also observed. Unlike previous campaigns, the antenna voltage condition has deteriorated during the course of a day and deteriorated after boronization. The faults that have appeared consistently are phase balance faults. Most of March was spent analyzing RF power transmission systems and maintaining them so they will operate more safely and reliably.

Progress on the DNB and the power supplies for the DNB continues and is still approximately on schedule. The technical details of work on the DNB can be found in the weekly reports.

Two runs in early February were devoted to calibration of several spectroscopic diagnostics. The sensitivity of the grazing incidence VUV spectrometer was calibrated using the branching ratio technique. This technique relies on observation along the same line-of-sight of two spectral lines originating from the same upper state, one in the VUV and one in the visible. The intensity ratio of such a pair of lines can be calculated quite accurately. Thus a calibration in the visible is transferred to the VUV.

There are two filtered-diode arrays which measure chord integrals of the soft x-ray emissivity in the edge of the plasma with extremely high spatial resolution. Improvements have been made to the algorithm which inverts these chord integrals. Previously, most of the CPU time involved in this routine was spent on mapping pixels to their flux surfaces (and then to the midplane). A routine has been developed which performs this mapping roughly 6 times faster. This improvement allows us to invert data from both edge arrays inbetween shots. In addition, the inversion algorithm has been changed to eliminate a numerical problem which previously limited the spatial resolution to 1.7 mm. Inversions can now be performed routinely with a spatial resolution of 1.2 mm. This improvement is particularly important for ELM-free H-modes, where the pedestal width of the soft x-ray profile is very narrow.

One run day was dedicated to calibration and testing of the fast-scanning gas-injection probe. Controlled amounts of ethylene and freon gas were first puffed into the vacuum chamber to correlate quantity injected with plenum fill pressures and to optimize valve timing. Plasma experiments were performed in which freon was injected into 0.8 MA discharges with line-averaged density around 1.5 ×10²⁰ m^-3. The amount of gas injected and the insertion depth of the scanning probe were varied. Measurements of fluorine were made by observing Li-like flourine lines and measurements of chlorine were made using x-ray lines from He- and H-like charge states.

Romik Chatterjee (UT-FRC), John Heard (AU) , and Perry Phillips (UT-FRC) completed the installation of the high spatial resolution ECE radiometer. Thirty of the thirty-two channels are operating with low noise, and shakedown data are being acquired at the ``slow'' 10 kHz rate. Qualitatively, the data are consistent with that expected of plasma temperature during current rise and sawteeth. Spatial localization of the channels is not consistent with expectations. This discrepancy is attributed to the sensitivity of the second-harmonic mixer to third-harmonic radiation. A low-pass filter is being install to overcome this problem.

Radiated power measurements are routinely available, even though all bolometer systems were relocated during the last shutdown and numerous changes were made. Two systems are global measurements, one using a standard foil bolometer, the other using a wideband silicon detector (AXUV diode series). A third system is based on 2 arrays of wideband detectors (AXUV diodes). The diode single view agrees with the integrated radial emissivity profile, within a few percent. However, a significant difference (up to a factor of 2-3) is found between the standard foil bolometer and the diode. On a single impurity injection event, though, the difference amounts to approximately 30% only, indicating that a measurable amount of power is carried by neutrals (which are not measured by the diodes). Difficulties arise with neutrals as the plasma is not ``optically thin'' and careful inversion/analysis is required. Additional checks and calibrations are under way, and a future run in helium is planned which should help in elucidating the difference.

A new CCD array detector has been installed on the tokamak to measure visible continuum emissivity profiles with sub-mm spatial resolution. In the spectral region covered by the transmission filters, the emission is dominated by free-free bremsstrahlung. The camera views the plasma tangentially, in the midplane, with major radius coverage 0.60m < R < 0.93m. The camera has a 2048x1 pixel detector, and laboratory calibrations of the optical system show that chordal resolution of 1 mm or better should be achieved along the entire length of chords passing tangentially near to the location of the steep density gradient seen in H-Mode. Depending on signal strength, the camera is capable of time resolution between 0.25 and 3.3 ms. Initial results show that the time histories of the central chord are in excellent agreement with the single channel view which uses a narrower spectral filter, indicating that the broader spectrum is dominated by free-free bremsstrahlung. Profiles in ohmic and low power L-mode ICRF discharges were obtained with 2 msec time resolution.

First results from a new visible spectrometer system (Kaiser) have been obtained. The system consists of 14 spatial chords looking tangentially at the plasma boundary (radial spatial resolution ~ 3 mm). In addition to observing D_a and BII profiles, the spectrometer views three He I lines emanating from a helium gas jet which is fixed to the side of the AB split limiter. Using the ratio of the three He lines, it is possible to derive the local n_e and T_e profiles. This work is in collaboration with PPPL (which has loaned the spectrometer) and TEXTOR (which has supplied the theoretical line ratios). Full profiles of n_e and T_e from several centimeters outside the separatrix to approximately 1 cm inside have been obtained. There is reasonable quantitative agreement, particularly with respect to the density profile, with a reciprocating probe at approximately the same location.

First results from the recently installed Ly_a detector array have been obtained. This diagnostic consists of a 20 channel silicon detector, AXUV-20 EL, capable of measuring radiation in a very wide bandwidth. Using a Ly_a filter (centered around 1216 Å), we can now image the Ly_a emission (from neutral deuterium) in a horizontal plane slightly below the midplane in a tangential view. Covering a 4 cm radial range near the separatrix, this instrument has a nominal 2 mm resolution, with a temporal resolution of a few milliseconds (due to the high amplifier gain needed). Results indicate a peaking of the radiation near the separatrix, increasing with plasma and neutral densities, and decreasing in the presence of an H mode, similarly to the D_a emission. This instrument will give us a measurement of the neutral density and ionization profiles in the main chamber near the separatrix, important parameters for the edge dynamics.

An inner wall rangefinder has begun operation. This instrument measures differential wall movement between two closely spaced positions on the inner wall. Initial results during low current disruptions, with barely measurable halo currents, indicate 2 to 6 microns of relative motion during the current collapse. Results from this new diagnostic will be correlated with data from strain-gauges at the same location. The data will be used to benchmark codes being developed to model the inner wall upgrade.

Collaborations and Participation in the Fusion Science Community

The yearly C-Mod Program Advisory Committee meeting occurred at the PSFC on Feb. 4-5. The PAC members attending, D. Hill, E. Synakowski, E. Strait, B. Carreras, C. Gormezano, and P. Efthimion, heard presentations on the status of the C-Mod facility, near-term program, and long-range plans. Our OFES contract monitor, R. Dagazian, also attended the PAC meeting. R. Hawryluk and N. Sauthoff from PPPL participated in the discussions on the first day.

Ron Bravenec from U Tx. was at C-Mod for one week visits in both February and March working on the BES diagnostic.

Stewart Zweben (PPPL) visited the PSFC the first week in February and in late April, and presented an idea for a 2D turbulence imaging diagnostic, which could be implemented on C-Mod in the next run campaign.

Our PPPL collaborators, scientists Gerd Schilling (RF), Gary Taylor (ECE), Dave Johnson (Edge Thomson Scatteriing), and RF technicians, Chris Brunkhorst and Gary D'Amico, were on-site frequently during the quarter.

This was also the case for Bill Noonan, our collaborator from U Md, and for Ricky Maqueda, our LANL collaborator. Dr. Maqueda finished bringing two diagnostics fully on-line. One diagnostic was the fast-framing visible camera (frame rate=1 khz, gateable to 10 ns). This is being used to look at ``snapshots'' of striations (filaments) in visible light. The other diagnostic is the IR camera, imaging the inside of the vessel from above in 4.24-4.42 micron light. The view is made somewhat tangential by using a mirror mounted within the vacuum vessel. As such, the outer strike point (beneath the nose of the outer divertor) is imaged. The observed image is close to what was expected.

Dave Mikkelsen (PPPL) spent the second week of March at MIT becoming familiar with the MDSplus tools for generating and displaying TRANSP runs based on C-Mod data. These will be used as the starting point for comparisons with various transport models.

Raffi Nazikian (PPPL) visited in March to work on the reflectometry diagnostic on C-Mod with Yijun Lin and Jim Irby. During his visit, he helped identify the source of a phase error which had been previously noted in reflectometer profile data. The problem appears to have been magnetic field effects in a ferritic component; local shielding to remedy this problem is being implemented. Raffi plans to continue to help with hardware upgrades to the C-Mod system, bench marking of the PPPL and MIT 2D codes, and data analysis.

In mid-March Romik Chatterjee of UT-FRC began a 1 month visit to C-Mod to work on hardware and data analysis for the FRC/AU ECE heterodyne radiometer. Ashley Shugart, a UT-FRC graduate student and his thesis advisor, Gary Hallock, arrived to kick off Ashley's first extended visit to C-Mod. While here, Gary worked with Ashley and Alex Mazurenko (PSFC) to remove oscillation problems in the PCI amplifiers supplied by UT-FRC for interim use while new amps are being designed and constructed at UT-FRC.

Selected Domestic Travel

Bruce Lipschultz visited Columbia January 29 and presented a talk - ``Exploring the synergism between atomic and plasma physics in Alcator C-Mod''.

Miklos Porkolab attended the Fusion Power Associates Meeting in Marina del Ray, California, on Jan. 25-27, and presented a paper: ``Requirements and Development Path for the Advanced Tokamak Concept''. He also attended the FPA Board Meeting on January 25.

Ron Parker, Dave Gwinn, and Steve Wukitch visited PPPL in early February to meet with Stefano Bernabei, Neville Greennough, and Elmer Fred to discuss, survey, and assess the MIT lower hybrid equipment at PPPL and the proposed LH experiment on C-mod.

Bob Granetz, Jay Kesner, and Jesus Ramos attended an MHD Critical Issues Workshop at PPPL in late February. While there, they saw Energy Secretary Bill Richardson dedicate the NSTX, which just recently began operation.

Also in late February Martin Greenwald and Tom Fredian attended a workshop on software sharing in the MFE community at PPPL. The main area of discussion was the sharing of analysis applications (such as TRANSP and EFIT), database applications (such as LOGBOOK and LOCUS) and graphical tools (such as REVIEW+, FAS and IDL/RPLOT). This software sharing has become feasible since many sites have standardized on the MDSplus data system, a common relational database interface and the IDL data scripting language.

On Feb. 22 Paul Bonoli was at PPPL for their Science Focus Group Meeting on Wave Particle Interactions. He gave a 30 minute talk on non-inductive current drive for Advanced Tokamak Applications.

Paul Bonoli attended a March 9-11 workshop at General Atomics on the Physics Requirements for Advanced Tokamaks. Paul was one of the meeting organizers, along with Ed Synakowski (PPPL) and Alan Turnbull (GA), and gave a talk on ``Lower Hybrid Current Profile Control Studies in Alcator C-Mod''.

Miklos Porkolab also attended the Physics Requirements for Advanced Tokamaks workshop on March 9-11 at General Atomics, where he made the following presentations: 1. Overview of the C-MOD AT Program 2. ITB formation Experiments in C-Mod: PEP Mode During Current Ramp-Up and ITB Induced by H/L Transitions. He also led the Discussion Session: ``What Issues Remain for MHD Stability?''

Miklos also had discussions with Ed Synakowski(PPPL) and Keith Burrell(GA) about the role of sheared electric fields and flows in the termination phase of past PEP experiments with ICRF and pellet injection. A new interpretation of the collapse of the PEP has emerged from these discussions, which may be tested in future experiments on C-Mod and DIII-D. If the interpretation is correct, it may be possible to extend the PEP duration by spinning the plasma in the counter current direction with either beams (DIII-D) or mode converted IBW (C-Mod).

In early March Ian Hutchinson participated in the Fusion Facilities Coordinating Committee (FFCC) meeting in San Diego, and also attended some sessions of the AT Workshop.

Joe Snipes, Paul Bonoli, Antonio Bruno, Xavier Bonnin, Miklos Porkolab, Eric Nelson-Melby, and Sergei Krasheninnikov attended the APS Centennial Meeting and the associated Sherwood Theory Meeting, in Atlanta March 21-26.

In mid-April Martin Greenwald presented a seminar, ``Highlights of the C-Mod Transport Program'', at NYU.

International Travel

Tom Fredian visited the Australia National University during the first part of March. There he installed the MDSplus data system for the H-1 Heliac device. His local contact is Boyd Blackwell. With this installation, MDSplus will be in use on four continents.

Near Term Plans

Our near term plans are focussed on the present run campaign, which is scheduled to last into June. A number of MiniProposals have already been submitted, but proposals are still being solicited. Fifteen new MiniProposals have been approved. The C-Mod Run Schedule on the Web has been updated. Overall plans for the 1999 Campaign can be found at http://www.psfc.mit.edu/cmod/run_schedule.html. Daily operations information can be found on the C-Mod Operations Calendar at http://www.psfc.mit.edu/cgi/calendars/cmod, which is updated as information becomes available. Those with access to the online OPS bulletin board are advised to check there for the most up-to-date schedule information.

We expect to be delivering significant heating power through the new ICRF antenna during this campaign, and we plan continued investigation of the effects of the divertor bypass flaps. Most of the new diagnostics are now operational.

M.V.Umansky, S.I.Krasheninnikov, B.LaBombard, B.Lipschultz and J.L.Terry ``Modeling of particle and energy transport in the edge plasma of Alcator C-Mod'', to be published in Phys.Plasmas.

May M.J. et. al, Plasma Physics and Controlled Nuclear Fusion, 41, No 1, (1999) 45-63.

J.A. Stillerman and T.W. Fredian, ``The MDSplus data acquisition system, current status and future directions'', Fusion Engineering and Design 43 (1999) 301-308.

T.W. Fredian and J.A. Stillerman, ``MDSplus remote collaboration support-internet and world wide web'' Fusion Engineering and Design 43 (1999) 327-334.

J. Lewis and R. Childs, ``Heat treating a Superconducting Coil for a Fusion Reactor'', Advanced Materials and Processes 155 (1999) 67-69.