The primary activities at Alcator C-Mod during the first quarter of FY01 were: continuation of the year-2000 run campaign; analysis of data from that campaign; initiation of a short maintenance period in order to assess and repair the 4-strap ICRF antenna; and continued participation in the critical assessment of fusion science. A more detailed account of these activities follows.
 
 

Scientific Results

Core Confinement and Transport Studies

A major focus of this quarter's experimental campaign has been an investigation of the internal transport barriers obtained in Alcator C-Mod with off-axis ICRF heating. These plasmas feature a combined EDA H-mode edge and an ITB, as indicated by formation of peaked density profiles. The regime is accessed in C-Mod using ICRF heating at 80 MHz with a central toroidal field <4.5 T, which places the minority hydrogen resonance well off-axis to the high-field side. The mechanism for generation of the barrier is unknown, but previous experiments have indicated a correlation with changes in core rotation, and a relationship to other core barrier regimes including PEP mode and the spontaneous formation of an ITB at the H-mode to L-mode transition (the so-called Enhanced Neutron or EN mode). We have now documented that the "off-axis-heating" ITB is a barrier for the transport of energy as well as for particles. This is shown in Figure 1, where the TRANSP-derived c_eff is shown as a function of space and of time. The effective heat diffusivity is reduced inside the barrier, which is located at r/a~0.4.

Figure 1.

Experiments were performed to study the conditions leading to the formation of the ITB. In particular, a sequence of similar shots allowed a radial scan (out to r/a of 0.5) of T_i and plasma rotation. (This utilized high-resolution x-ray spectroscopy, and those data are still being analyzed.) We have also found that a higher target density, n_e^bar>1.5x10²⁰m^-3, is beneficial in sustaining the H-mode edge along with the ITBs. We found that with higher target density we could create sustained EDA H-modes with pronounced central density peaking. Central densities in these cases were up to n_e0=8x10²⁰ m^-3. The higher target density shots did not exhibit the H-mode to L-mode back-transitions that were observed at lower target density. It is likely that the higher target density effectively dilutes the plasma impurities, leading to less radiation, and thus avoiding a collapse of the H-mode phase.

Figure 2.

An obvious way to extend this promising regime of reproducible ITBs is to employ C-Mod's available two-frequency ICRF capability. In this scenario, use of off-axis (80 MHz) heating results in the formation of the barrier, while application of ICRF at 70 MHz heats the region inside the barrier. This scenario also allows the possibility for changing the local temperature and density gradients by varying the on-axis and off-axis ICRF heating powers. Such a scenario is shown in Figure 2. There are obvious increases in the neutron rate and heating when the central (70 MHz) ICRF heating is applied. Again, radial profiles (r/a < 0.5) of T_i and plasma rotation were obtained during the on-axis heating phase of the density barrier mode.

It is apparent that for the shot shown in Figure 2 the central density continues to rise even after the application of the on-axis heating. This behavior was not universal. In some shots the density increase during barrier formation was found to be arrested with the central heating. Thus it is possible that the on-axis heating causes the density barrier to "leak" under certain plasma conditions.

________________________________________________________________________________

It has been established that triangularity, especially of the active x-point, affects the character of the H-mode, with higher triangularity lending itself to the advantageous EDA operation, and lower triangularity to more ELM-free behavior, with the associated difficulties of impurity accumulation. The closed geometry of the lower divertor makes triangularity scans equivalent to scans of divertor geometry, from an inner strike point that nearly touches the inner nose through to operation on the top of the outer divertor. Operating the plasma into the upper divertor chamber, which is open, avoids most of the strike-point geometry changes, and has enabled us to establish much more clearly whether the effects we observe are a consequence of plasma shape or are linked more intimately to divertor configuration and recycling location. Reverse-field operation is required to maintain the sense of the grad-B drift with respect to the active x-point. A series of dynamic triangularity scans were carried out, with the upper triangularity varied from 0.8 to 0.4. The variation of x-ray pedestal width as well as the appearance of the quasi-coherent mode, as observed on the PCI diagnostic, confirmed our previous results. The "EDA-ness" depends on the triangularity. Ramps of current and toroidal field were employed to maintain the q approximately constant during the triangularity variation, confirming that the shaping rather than the safety factor was the dominant effect. It therefore appears that the upper diverted case is the same as the lower diverted case, and the effects we see are not due to the proximity of divertor hardware, but is truly a result of shaping.

________________________________________________________________________________

We have observed what appear to be beta-limiting MHD modes in some shots at high power on C-Mod. By combining all three ICRF antennas, we have a number of shots with up to 4.5 MW of ICRF power, and we have achieved normalized beta values up to at least 1.6 for several hundred msec. At moderately high density, 3 - 4 x 10²⁰ m^-3 in EDA H-mode, small chaotic ELMs appear on top of the enhanced D alpha emission. At lower density, less than 2.5 x 10²⁰ m^-3, we sometimes find large amplitude low frequency MHD activity that begins as sawtooth precursors or postcursors, but then continues throughout many sawteeth. The frequency may reach 20 kHz or so, but when the amplitude becomes large enough ~ 3 G or more at the wall, the mode begins to slow down and finally locks. The confinement is substantially degraded when these large modes appear and the appearance of the modes is well correlated with the peak in normalized beta ~ 1.4 - 1.6. The poloidal beta values at the same time reach values of 0.8. Mode analysis has shown that in some cases, the modes have m=2, n=1 or m=3, n=2. The low m, low n, large amplitude, and the correlation with the peak normalized beta suggest that these modes may be the first neoclassical tearing modes observed on C-Mod, particularly since they are also correlated with lower density operation at high power in a lower collisionality regime. However, more analysis is required to confirm this speculation.

________________________________________________________________________________

During previous experiments, it was observed that the x-ray pedestal location at the plasma top is found to be very close to the separatrix (typically 1-2 mm), in contrast with the x-ray pedestal on the outboard side, which is usually 10-12 mm inside of the separatrix. The emissivity is from recombination of highly charged impurities drifting out from the core, and therefore may be subject to the same up/down asymmetry observed for Ar⁺¹⁷, which was attributed to Bxgrad(B) effects on neoclassical transport. This explanation was tested during experiments with reverse-field H-modes, for which the Bxgrad(B) drifts are reversed. Contrary to expectations, the observed relative displacement of the x-ray pedestals at the top and outboard were the same as with normal field operation. The up/down argon asymmetry, on the other hand, was reversed from the normal field case, as had been observed previously. These results indicate that the mechanism for the shift of the x-ray pedestals is not the ion grad-B drift effect.

Edge/Divertor Physics

The report from the last quarter detailed the new picture of wall vs. divertor recycling that has been emerging from C-Mod results. Similar trends have also been seen on Asdex-Upgrade and more recently on DIII-D. Briefly, that new picture is that a large fraction of ions in the SOL flow to the walls rather than flow into the divertor (although the C-Mod divertor usually receives most of the conducted and convected energy fluxes from the SOL and does entrain/compress recycling impurity and fuel gases). The reason for such high ion fluxes to the vessel walls rather than into the divertor appears to be primarily the existence of strong cross-field particle transport in the main-chamber SOL.
Figure 3.

The causes of large cross-field SOL transport are also being investigated, both with probe diagnostics and with optical diagnostics. As is true in other machines, we find the SOL to be quite turbulent, with normalized density fluctuations, n^RMS/n, of order 1 (at least in the far SOL). We also observe the far SOL transport to be quite "bursty", i.e. there are random large, relatively long-lived bursts of density. This is shown is Figure 3(b), where the auto-correlation functions of normalized ion saturation current are plotted. The curves are color-coded according to the measurement's radial position in the SOL, which is shown in Figure 3(a) along with the measured SOL density profile. It is apparent that the auto-correlation times are increasing with distance into the SOL. The "bursty" nature of the SOL turbulence is also evidenced by ~1ms snapshots of emission from the SOL. Shown in Figure 4 is such an exposure where, instead of smoothly-varying emission, bright (and dark) features with radial and poloidal size scales of ~1 cm are typically seen.
 
 

Figure 4.
 

New results from a gap scan run are also consistent with the picture of dominant main-chamber recycling. The midplane pressure is found to be insensitive to variations in the outer gap. This result is in contrast to its stronger dependence on density and confinement mode. In the gap-scan experiment we compared the effect of changing the outer gap on three different sets of probes - the A scanning probe (ASP) which is inserted between two outer poloidal limiters, the F scanning probe (FSP) which measures the SOL profile near the divertor entrance and the Outer divertor probes located in the outer divertor surface. We found that varying the gap has a clear effect only on the ASP SOL profiles. As the gap was decreased, the width of the SOL at the ASP location decreased accordingly. However, the same was not true for the other probe measurements. We found that they were relatively unaffected by gap changes. This result implies that the SOL 'fills in' in regions where the field line length is long enough (3-5 m) between obstacles, but not for short connection lengths as is the case for the ASP (.8 m). This result was further emphasized for the case of varying the inner gap. When that gap was reduced to zero, all the probes indicated that the SOL was 'filling in'. This 'filling in' effect is a clear demonstration that the perpendicular transport in the SOL is very large, leading to large interactions with the wall. In fact, as the outer gap is reduced the interaction with the far wall stays the same as the plasma 'fills in' the SOL, leading to the independence of midplane pressure on gap.

RF Research

In the last quarterly report, details of the improved performance of Alcator C-Mod's 4-strap antenna, the so-called J-port antenna, were given. In essence, the performance of the 4-strap antenna had been improved to the point where it was equal to that of the two dipole antennas. The 4-strap antenna exhibited good heating efficiency and freedom from impurity generation, similar to the two double strap antennas. The good performance was observed up to heating powers of ~2.0 MW. However, during this quarter additional experiments have shown that satisfactory performance with the J-port antenna was limited to powers less than about 2 MW. In fact, experiments in December of 2000 showed that the performance was degrading with time. RF-plasma interactions (hot spots and injections) were observed to occur on the antenna protection box and Faraday screen. At power levels above 2MW this interaction became disruptive to the plasma, even after antenna conditioning. For [0,pi,0,pi] phasing, the power level at which the plasma disrupted was ~2.5 MW. For other phases, the limit decreased: for [0,.85pi,0,.85pi] and [0,1.15pi,0,1.15pi] the limit was ~1.8 MW and for [0,0,0,0] the limit was 0.8MW. In addition, the hot spot location on the antenna side limiter and Faraday screen changed with plasma current, consistent with the field-line pitch. This suggested that there is an electric field developing along a field line connecting the corner of the antenna box, side plate and Faraday screen. These observations and trends led to a machine vacuum break in order to diagnose their cause or causes.

Operations and Diagnostics

The Alcator C-Mod run campaign for the year 2000 continued during the months of November and December. The final run of the campaign was December 22, 2000. After that run, we began warming up the machine in preparation for a maintenance period whose primary purpose is to repair the 4-strap antenna. (See below and the section on RF Research.) Manned access to the vacuum vessel was made during the first week in January, 2001. Except for parts of the 4-strap antenna system, one mirror used for the BES (Beam Emission Spectroscopy) and MSE (Motional Stark Effect) diagnostics, and some melting on leading edges of some divertor tiles, the rest of the internal hardware was essentially undamaged. Pictures documenting the state inside the machine can be found at

http://www.psfc.mit.edu/cmod/operations/EngImages/CmodImages.asp?ref=/INVESSEL/2001/Up_to_air_Survey/

Arc sites were found on the antenna-protection-box housing the 4-strap antenna, and this observation is consistent with along-field-line paths between exposed metal components. This damage was not a surprise, since strong plasma interaction had been seen on video images of the antenna during plasma operation. Extensive tracking and arcing marks were also discovered between the strip lines (which run from the co-ax feeds at the vacuum interface to the antenna straps) and septum (which is grounded and separates the strip lines vertically). The damage occurred in areas where there exists an along-field (E parallel to B) path between the strip line and the septum ground. Meetings to review changes to the antenna were held with our PPPL collaborators on January 19 and 26. A conceptual design for changes to the strip line feeds was made as a result of those meetings. The strip lines will be rotated approximately 90 degrees relative to their old orientation in order to take advantage of magnetic insulation. In addition, the separation between the strip line and the ground plane will be increased by 50% to reduce the electric fields. The conceptual design solid model can be viewed at

http://www.psfc.mit.edu/people/irby/jport_antenna_stripline3a.jpg

A schedule has been developed to implement the changes to the antenna. This is turn determines the length of the present maintenance period. We expect installation of the antenna with the new components to be complete by late March. Following pumpdown, vacuum conditioning, and plasma startup, plasma physics operation will begin in early May and continue through early June. The short term schedule may be found at

http://www.psfc.mit.edu/cmod/operations/short_term.pdf

The Diagnostic Neutral Beam

The damage on the MSE/BES mirror consisted of chipping of glass at the mounting points on the back of the mirror. It appears to be a result of shaking during disruptions. The mirror has been removed and has now been epoxied to its metal mounting fixture. This fix occurred after a series of tests that showed that a rugged bond could be made in this way between the mirror and the fixture.

The on-going maintenance period has enabled maintenance of the DNB that should improve its performance and reliability. A water leak in the gas feed line was fixed and the entire beam was baked. The DNB is back on the air and is operating reliably at beam voltages of 50 kV and extracted ion currents over 5 A. In the beam component mix the full energy component fraction has been improved from 13% to 27%.

Lower Hybrid Project

Work on this major project is continuing. Many of the details have been listed in the weekly reports and will not be repeated here. The procurement process for the second largest purchase of the project, the circulator, is progressing. Proposals were received at the end of December. A source selection committee consisting of Monty Grimes here at MIT, the head of the RF group at Bates Laboratory, and an outside consultant is in the process of reviewing the proposals. The largest item being purchased for the project, the HV mod/reg supply, is still on schedule for an October 2001 delivery.

A design review of the project is scheduled for February 7th and 8th.

Collaborations and Participation in the Fusion Science Community

Jim Terry has submitted for the Alcator group an abstract to the general meeting of the APS entitled "Experiments in Plasma Physics and Fusion Science on the Alcator C-Mod Tokamak".

This meeting, held in Washington D.C. at the end of March, brings together all divisions of the APS. It is, therefore, a broad forum for communicating the general aspects of the C-Mod program's research to the physics community and the government community.
 
 

Domestic Travel

During the first week in November, 2000, Steve Wukitch attended the US/Japan technology exchange at PPPL. At the exchange, a fast ferrite tuner prototype was discussed; a performance test of this device will be made on C-Mod in the near future.

Earl Marmar attended the 2000 Conference on the Applications of Accelerators for Research and Industry, held in Denton, Texas, where he presented an invited talk entitled "Diagnostics on the Alcator C-Mod Tokamak using Neutral Hydrogen Beams". The presentation was on November 4.

Miklos Porkolab and Dmitri Mossessian participated in the DIII-D Research Opportunities (aka "Brainstorming") Meeting at General Atomics in San Diego during the second week in November. Amanda Hubbard and Martin Greenwald also presented research proposals via the video-conferencing system from MIT PSFC.

Miklos Porkolab attended the FESAC meeting in Washington on November 14 and 15.

On November 30 Paul Bonoli attended the DoE Workshop on US - Tore Supra collaborations. The meeting was held at the DoE headquarters in Germantown, MD and was attended (in part) by Jean Jacquinot and M. Chetalier from Tore Supra, Wendell Horton from University of Texas, Ron Stambaugh from GA, John Hogan from ORNL, and Erol Oktay, John Willis, Rostom Dagazian, and Walter Sadowski from DoE. Paul spoke about the possibility of C-Mod - Tore Supra collaborations involving a general exchange of information in the following four areas: (i) Hard x-ray measurements during LH current profile control experiments in C-Mod to determine the spatial and energy distribution of fast electrons. (ii) Long pulse and cooling issues related to the C-Mod divertor and its antenna systems (LH and ICRF). (iii) LH launcher design (to be coordinated with PPPL). (iv) Internal transport barrier physics.

Bruce Lipschultz spent November 28 - 30 at GA. The primary purposes of the visit were: (1) to discuss C-Mod and DIII-D data on the subject of radial transport in the SOL; and (2) to discuss techniques used for measurement of T_i and v_flow using spectroscopy. He also gave a talk entitled "Wall recycling in Alcator C-Mod; Is it occurring or not?". This covered a variety of results from C-Mod on radial transport. Considerable time was spent looking at DIII-D probe data with Rick Moyer and Jose Boedo, and Thomson scattering data with Dennis Whyte. They found a number of discharges that exhibited flat density profiles in the far SOL. Whyte, Boedo, Peter Stangeby, Gary Porter and Bruce spent considerable time discussing the DIII-D miniproposal for studying the SOL transport under conditions of varying density and 'gap' between the separatrix and the wall. Finally, Bruce discussed the fitting techniques for CXRS spectroscopy with Richard Groebner, and the hardware characteristics of the DIII-D equipment with Keith Burrell. This information will be useful in both our current analysis and a new system for poloidal CXRS that we are considering.

Earl Marmar, Joe Snipes, Spencer Pitcher, and Miklos Porkolab participated in the Burning Plasma Science Workshop in Austin, Texas from 11 - 13 December.

Yuri Rokhman and David Terry attended Labview software programming courses from December 18 through 22 at National Instruments in Woburn, MA.

During the second week in January, 2001, Steve Wolfe and Ron Parker participated in the DIII-D Advisory Committee Meeting at General Atomics in San Diego. Steve Wolfe also presented a talk on recent C-Mod results to the DIII-D group.

Paul Bonoli and Bob Granetz co-chaired RF and MHD topical group sessions at the NSTX Ideas Forum at PPPL held during the third week in January. Paul gave two short presentations at the forum entitled "Full-wave simulations of HHFW in NSTX Using Exact MHD Equilibria" and "Future Modelling Plans for HHFW Heating in NSTX Using TORIC & TRANSP". Bob gave a talk discussing disruptions in spherical tokamaks.

Eric Melby spent January 23-25 at the Los Alamos National Laboratory, where he gave a talk entitled "Phase Contrast Imaging Observations of Ion Bernstein Waves in the Alcator C-Mod Tokamak."

________________________________________________________________________________

International Travel

Tom Fredian visited NIFS in Toki, Japan to install MDSplus on the CHS experiment. Tom together with Jeff Schachter from General Atomics installed MDSplus and MDSplus related utilities such as ReviewPlus and provided training on the use of the software.

Ian Hutchinson began a 5-month long sabbatical at the Australian National University at the beginning of January. Earl Marmar has taken on his responsibilities while he is away.

John Rice is spending ~ 6 weeks (January and part of February) working at the LHD experiment in Japan.

Joe Snipes spent the week of 22 - 26 January at JET taking part in experiments that attempted to obtain Enhanced D alpha H-mode on JET.

________________________________________________________________________________

Near Term Plans

The duration of the present maintenance is being determined by the refurbishment of the 4-strap antenna. Machine pumpdown is scheduled for April with a short (~5 weeks) run campaign schedule to begin in May. The goals of that campaign will be 1) to evaluate the changes made to the 4-strap antenna and 2) to evaluate the DNB and its diagnostics. Assuming that the 4-strap antenna modifications result in its anticipated good performance, we plan for a) more heating power, allowing stronger ITB core heating (~ 3MW), b) experiments on Mode Conversion flow drive, current drive, c) further investigations of 2-frequency ICRF, with prolonged control of the "off-axis-heating" ITB so that the primary mechanism(s) responsible can be identified.

Those with access to the online OPS bulletin board are advised to check there for the most up-to-date schedule information.

________________________________________________________________________________

Remote participation in all C-Mod staff meetings is available using video and audio conferencing equipment. The system includes high quality room audio and video images. Remote participation from the community is encouraged.

________________________________________________________________________________

Live streaming video and audio from the control room is available over the WEB during C-Mod operations for off-site collaborators and other interested parties. The video shows the control room activity between shots and switches to video of the discharge itself during each pulse. The URL is

http://www.psfc.mit.edu/cmod/online/control_room_rv1.html