The primary activities at Alcator C-Mod during the fourth quarter of FY98 were: data analysis and presentation of physics results, the C-Mod Ideas Forum, maintenance of the TF magnet, reassembly of the machine, continued work on major systems upgrades (the RF transmitters and antennas and the Diagnostic Neutral Beam and its associated diagnostics), and continued participation in the critical assessment of fusion science. A more detailed account of these activities follows.

Scientific Results

Core Confinement Results

Alcator C-Mod Enhanced D_a (EDA) discharges which have a normalized beta above about 1.2 show an apparently new type of ELM behavior. These discharges have substantial edge pedestals and are well above the threshold for type III ELMs. The observed ``high beta'' ELMs exhibit a large signal on the D_a intensity (often bipolar, with the D_a level dipping well below the equilibrium EDA level), are quite short, perhaps 100 ms or less, seem to be localized within the transport barrier, and do not lead to any significant loss of particles or energy. Enhanced particle fluxes can, however, be seen by the divertor probes. The perturbation on the ion saturation current is intermittent, suggesting that the particle flux can be interrupted temporarily. This observation is consistent with the bipolar D_a signal and the observation of ECE signals which are either refracted or cutoff by high density regions that develop in the pedestal during these events. Otherwise similar high beta ELM-free discharges do not develop this type of ELM activity.

In our continuing study of the EDA/H-mode, we have compared it with the Low Particle Confinement H-mode (LPCH) (Bures, et al, Nucl. Fus., Vol. 32 (1992) 539.) observed on JET in the early 90's. At least in the case of strong gas puffing into the X-point of JET (before they had a divertor), the LPCH-mode and the EDA H-mode do appear to be quite similar. They are both obtained with high gas puffing. There is increased D_a emission and main chamber recycling. The particle confinement is reduced relative to ELM-free H-mode with t_p^{LPCH or EDA} = 1/3 t_p^ELM-free. The energy confinement is only 20% lower than in ELM-free H-mode. In the JET case the magnetic configuration was double null with I_p = 3 MA, B_t = 2.8 T, k = 1.8, d =0.54, D(H) ICRF heating with P_RF < 10 MW, [`(n<sub>e</sub>)] = 5-6 ×10<sup>19</sup> m<sup>-3</sup>, and T<sub>i</sub> =T<sub>e</sub>. The walls were mostly carbon with some beryllium. On C-Mod, the EDA is regularly obtained in a single null X-point configuration with I<sub>p</sub> < 1.4 MA, B<sub>t</sub> = 5.3 T, k= 1.6, d = 0.5, D(H) ICRF power < 3.5 MW, [`(n_e)] = 4 ×10²⁰ m^-3, and T_i = T_e. However, there is a very closed divertor, and there are molybdenum walls. Both regimes are at relatively low b_N < 2 .

An analysis of ICRF H-modes which occur during the current rise has been carried out in order to determine how early H-mode can be obtained, and whether or not the H-mode threshold is increased by an upward current ramp. Several current-rise H-modes were seen during runs in which ICRF heating was applied in the current rise in an attempt to obtain reversed shear conditions. The earliest H-mode found had an L-H transition at about 0.24 sec. However, the global power threshold during the current rise was as low as P_threshold/([`(n_e)] B_t A_s) = 0.037, which is higher than the typically low C-Mod values of 0.022, but still lower than the ASDEX-Upgrade values of 0.044. A systematic power scan is required during the current rise to be sure that these are indeed the lowest threshold values. These examples of current rise H-modes show that it is possible to obtain H-modes quite early in C-Mod discharges. They provide a good starting condition for the studies of confinement and fueling in the current rise for comparison with various ignition scenarios.

Divertor and Edge Results

Analysis of CCD images of D_g from the divertor region has produced interesting results. These images are inverted to obtain the local emissivity through a method developed by a previous graduate student, Aaron Allen. The method has been further improved by Chris Boswell. In general, D_g in the divertor region is produced by volume recombination, as opposed to ground state excitation. Thus we find it to be a measure of the recombination rate of ions and electrons away from the divertor plates. The CCD inversions show that during detachment the recombination region expands from the outer divertor strike point and private flux regions towards the X-point. We have found that, as the density is increased in detachment, this expansion continues. At high enough densities the recombination crosses the separatrix into the core near the X-point. This means that a T_e of ~ 1 eV exists on closed flux surfaces just above the X-point. The corresponding distance inside the separatrix (when mapped to the midplane) is about 5 mm. According to the ECE diagnostic, in these cases the edge temperatures of the core T_e profile are depressed to ~ 30 eV approximately 2 cm inside the separatrix (again referenced to the midplane). Thus a temperature gradient is sustained along closed flux lines, and we calculate that a modest amount of radiational cooling ( ~ 200 kW) is required along the appropriate flux tube to maintain such a gradient. As the plasma current or input power is increased, the density at which this recombination region crosses the separatrix increases (as does the detachment threshold). A sequence of these 2-D emissivity profiles is shown in Fig. 1. Evidence from other diagnostics supports the observation of a very cold region at and above the X-point in these discharges. The VUV spectrometer, viewing across this region, observes emission lines from O⁺⁵, N⁺³, N⁺², N⁺, and D0. The time histories of those lines are consistent with a strong cooling above the X-point to a temperature at which deuterium atoms recombine, i.e. ~ 1 eV.

Measurements of Stark broadening of the D⁰ 8®2 line have been analyzed. The measurements are made along 16 viewing chords through the divertor region. A number of detached divertor discharges were examined, covering a range in density and heating power. It was found that the measured Stark density, averaged over either the inner or outer divertor, scaled as approximately as [`(n<sub>e</sub>)]<sup>0.75</sup>  P<sub>SOL</sub><sup>0.3</sup>. This scaling can be roughly explained if one assumes that the pressure in the divertor region is a fraction of the upstream pressure (detachment), that the divertor T<sub>e</sub> in the emitting regions is essentially constant (0.5 eV), and that the upstream density and T<sub>e</sub> scale as [`(n_e)] and P_SOL^2/7 respectively.

D_g/D_a intensity ratios from the divertor have also been analyzed. This ratio has been used previously by McCracken and by Isler to determine whether or not the Balmer series emission is due to volume recombination. We applied the technique to a number of shots and found a wide range of cases. In Ohmic detached plasmas D_a is mostly produced by recombination. For the detached H-mode plasmas, the opposite is typical. D_a is due in large part to excitation from the ground state (although higher-n lines remain populated by recombination). This knowledge was used to interpret flow measurements from detached H-mode shots, where we have Doppler shift, flow measurements for HeII and D_a. The relative velocity is ~ 5×10³ m/sec. Because D_a is dominated by ground state excitation (and thus is reflective of the neutral velocity) in these discharges, the measured difference in the ion and neutral velocities indicates that that the transfer of momentum from ions to neutrals is efficient.

Analysis of data from two reciprocating probes shows systematic differences during ohmic operation, suggesting a possible poloidal asymmetry in the scrape off layer conditions. The A-port probe (scanning horizontally, 10 cm above the midplane) indicates a much flatter electron temperature profile than the F-port probe (scanning vertically upwards, from near the divertor), reaching a maximum of only 30 eV at the EFIT separatrix. Such low values of electron temperature at the separatrix are inconsistent with power balance estimates based on the standard two-point model. In addition, the radial electric field derived from electron temperature and floating potential profiles appears to not map between the probes, being not more than 500 V/m at the location of the A-port probe, but up to 10kV/m at the location of the F-port probe. The possible influence of a poloidal limiter located near the horizontal probe is being considered. Plans are underway to repeat the measurements with nearly identical probe geometries.

A technique has been developed for using neutral molecules and atoms to filter out the neutral deuterium atom flux from the plasma from being measured by the bolometers. This technique is based on increasing the neutral pressure inside the bolometer camera housing until the neutral-neutral mean free path is short compared to the length of the housing. Since the bolometers are sensitive to both neutral and photon fluxes from the plasma this enables us to filter out the neutral component for portions of the shot. We have used no-plasma shots to measure the cooling effect of the injected neutrals on the bolometer. This cooling or pickup shows up as negative power and is very repeatable. It is due to the conduction of heat from the camera housing to the bolometer foil which is ~ 20^o C hotter. We find that the neutral component can be as large as 30% of the measured bolometer power in the divertor. The percentage does not seem to be strongly affected by density variation or detachment. We have some limited data with H-modes which shows a similar effect. The neutral component location does seem to vary.

RF Research

Since Alcator C-Mod utilizes ICRF as its sole auxiliary heating method, it is crucial that the RF power be efficiently absorbed. Efficient RF absorption has been measured in both of the main heating schemes used to date on C-Mod, hydrogen minority heating - D(H) at B = 5.4 T and ³He minority heating - D(³He) at 7.9 T. This is shown in Fig. 2, where the absorption is plotted versus minority fraction. At optimal minority fraction, 80% to 90% of the launched power goes into plasma heating.

Nearly fifty shots at 7.9 T with a D(³He) plasma have been analyzed to investigate the balance between direct electron heating through mode-converted ion Bernstein waves and ³He minority ion heating. Electron power absorption profiles were calculated from the break-in-slope of electron temperature at times of RF power transitions. The location of peak power absorption was compared with a cold plasma model to estimate the ³He concentration present in the bulk plasma. The resulting relationship of power to electrons vs. ³He concentration was compared with the full-wave code TORIC, with which it agrees qualitatively.

Analysis of data from the loop probes on the inner wall, opposite the E-port antenna, indicates that the RF power detected by the probes during minority heating depends on the minority concentration in the D(H) scenario but not in the D(³He) scenario. In D(H), for which the single-pass absorption is expected to be high, the detected RF power decreases monotonically with increasing H fraction. This agrees well with the analytic single-pass transmission calculation, and the output of the FELICE code (run with radiating boundary condition at the high-field side). In D(³He), for which the single-pass absorption is expected to be low, the measured RF fields are much larger than in the D(H) scenario, and independent of the ³He fraction.

Professor Yuichi Takase, from the University of Tokyo, visited the Alcator C-Mod Group in August. He collaborated with Paul Bonoli on code development problems related to the MHD equilibrium and current drive code ACCOME. (The development of this code is part of an official joint effort between MIT and JAERI.) Problems related to solution convergence in the ACCOME equilibrium solver (SELENE) were discussed. It was discovered that the convergence properties of high beta equilibria with non-monotonic current density profiles could be greatly improved by starting with a well-converged equilibrium at lower beta and low current drive power and then increasing the off-axis current drive power in small steps. Professor Takase will be using ACCOME to study high harmonic ICRF fast wave current drive in spherical tokamak configurations.

Operations and Diagnostics

The primary operational effort during this quarter has been directed towards completing repairs and improvements of the TF magnet and towards machine reassembly which began in mid-September. The number of technical tasks accomplished in the TF refurbishment is quite large and includes (a partial list):

1

- perform current-carrying-lifetime tests on feltmetal samples,

2

- qualify and perform brush plating of the bare copper fingers,

3

- silver plate the new TF feltmetal,

4

- perform wear tests, optical microscopic evaluation, peel and bending tests, and SEM on the new feltmetal before soldering it to the copper fingers,

5

- copper plate and then silver plate the sliding joint fingers,

6

- solder the feltmetal to fingers,

7

- design, fabricate, test, and install spring plates for the TF sliding joints,

8

- carry out three electroforming runs to repair the upper TF core finger,

9

- clean, inspect, and hi-pot (where applicable).

The alternator has been brought up to full speed to verify that changes to the drive motor alignment were done properly. A two month long upgrade in air and oil handling systems for the alternator is complete. These changes will allow better control of water and particulate content in the alternator oil supply.

The initial fitup of the J-port antenna was done with the help of several PPPL collaborators (Les Gereg, Dave Cylinder, Randy Wilson, Gerd Schilling). The antenna fit inside the port once the planned modifications to internal magnetic probes near J-port were made. The antenna feed straps were modified slightly to avoid internal conduits, and the flange design and assembly procedure has been finalized. Approximately 40 components are presently being silver plated.

A final design review for the new K port horizontal flange was held on July 27th. The review was passed successfully and final drawings have been produced. Machining, welding, and grinding is underway. The new design was necessary as many diagnostics had to be relocated following the installation of the new J-port antenna and the DNB with its associated diagnostics. Included in the new design are ports for the impurity injector, core bolometric system (bolometers, AXUV diodes and ``2p'' bolometers), Z-meter array (profile information), a D_a reticon array, Chromex spectrometric view, divertor spectrometer (a Johns Hopkins collaboration), Low energy Neutral Particle Analyzer (TOF), and tangentially viewing neutron collimator and spectrometer. Additional modifications will be made to the RF protection limiters in order to allow for the changes brought about by new installations.

We plan to measure disruption induced distortions in the C-Mod vacuum vessel inner wall during the next run period. These measurements are needed to help model the inner wall as we plan upgrades to the divertor for high field, high current operation. A prototype diagnostic for making this measurement has been built in collaboration with Carlos III University and is being tested. The system relies upon retro-reflectors installed on the inner wall returning the 0th and 1st order beams from a diode laser after its beam has passed through a Bragg cell. The returning beams are combined with a beamsplitter, and the relative phase difference between the two beams is measured. Several retro-reflectors along the inner wall will allow maps of relative motion to be made on very fast timescales during disruptions. This technique avoids the need to make absolute position measurements relative to some external reference. Tests so far indicate the diagnostic can easily measure relative motions up to 3 mm with a 0.69 mm resolution (a laser wavelength).

Also in collaboration with Carlos III University, we are continuing to develop the tangential interferometer which we expect to use to make high spatial resolution density measurements of the plasma edge. We are exploring both the possibility of simply expanding the prototype to more channels, and the development of a differential interferometer that measures dNL/dR rather than NL (NL is the line density). The gradient in the line density is what is required to perform the Abel inversion, not the line density. The differential measurement is not sensitive to vibrations and allows diode arrays with high spatial resolution to be used as the detector.

The Alcator C-Mod Ideas Forum
The 1998 Alcator C-Mod Ideas Forum was held on August 19-20. There were a total of 117 presentations of ideas for experiments to be carried out on C-Mod during the next campaign. The ideas, presented by about fifty principal authors from eight institutions, covered topics including divertor flows, impurity transport, neutral effects, RF heating and current drive, advanced tokamak studies, internal transport barrier physics, turbulence, testing of theoretical transport models, toroidal rotation, similarity scaling, divertor detachment, recombination, RI-mode experiments, H-mode and edge-barrier physics, SOL transport, MHD stability, high-performance plasma experiments, and support for burning plasma initiatives.

The Ideas Forum sessions at MIT were well-attended, including a significant presence by representatives from collaborating institutions. In addition, the proceedings were broadcast (audio and video) on the MBone, allowing monitoring and participation from off-site. Initial feedback on the broadcast has been favorable. The following excerpt is from an e-mail from Dr. T. Casper of LLNL:

``Thanks for broadcasting the CMOD brainstorming session over the MBONE. I, for one, appreciated it. I attended most of the afternoon (morning my time) sessions and have a better feel for what you all are up to. It seems to me that there are several areas for experiments that would be mutually beneficial to both CMOD and DIII-D in areas that I am interested in. ... As far as the broadcast goes, given the limitations of MBONE, you guys did a good job. The audio seemed to be well set up and I was able to hear most of the speakers quite well. Even most (not quite all) of the audience questions could be heard along with discussions. I didn't see anyone passing around a microphone nor did I see a professional crew so you guys must have done a good job at wiring the sound in the room. ... The only negatives were a low frequency background hum and an occasional high amplitude burst of noise between talks, probably at the microphone exchange which I could not see coming. I learned to mute the sound when this happened. ...''

Topical focus groups were formed to follow up on the prioritization of run time for experiments proposed at the Forum. A total of fourteen groups refined and developed the relevant ideas into a conceptual list of experiments, and assigned relative priority to each experiment according to its importance to the development of a particular physics topic. On average, each idea presented at the Forum was considered by two focus groups, due to overlaps in the physics or techniques involved. The groups reported their results at a meeting of the C-Mod Experimental Program Committee (EPC), and the EPC adopted a plan for run allocations consistent with constraints on operating time. These allocations are based on broad programmatic themes and include small reductions to the run-time requested by some of the topical groups. Appropriate combinations of the original focus groups are meeting to prepare revised experiment lists corresponding to the allocations. Mini-proposals will then be written in support of the high-priority experiments. (Approved MP's can be found at http://www.psfc.mit.edu/server-java/MiniProposals.)

Collaborations and Participation in the Fusion Science Community

Professor Cy MacLatchy from Acadia University, Nova Scotia arrived in July for his one-year sabbatical. Professor MacLatchy will be participating in divertor/edge experiments on C-Mod. His initial research will concentrate on video image-capture and 3-D tomographic reconstructions of impurity injection ``plumes'' produced by the gas-injecting scanning probe system.

Our collaboration with Glen Wurden and Richardo Maqueda of LANL has been continued into FY 99. The collaboration will provide a IR camera, viewing the divertor, and a fast-framing, gated visible camera for turbulence studies.

A significant ESnet connectivity milestone was reached in August when a number of long-awaited events occurred which will contribute to a vast improvement in network performance for communication between C-Mod/PSFC computers and our off-site collaborators. Along with this new address space came the ability to start moving nodes from the pfc.mit.edu domain (or namespace) into the psfc.mit.edu domain.

A quarterly review of the C-Mod program was held at MIT on October 6. Rostom Dagazian represented DOE, Bill Rowan from U Texas discussed the DNB and related diagnostics, Randy Wilson from PPPL discussed the new antenna and other PPPL/MIT collaborations. Stefano Bernabei also from PPPL, discussed plans for the C-Mod lower hybrid installation. Steve Wolfe discussed FY99 operating plans, Jim Irby discussed machine reassembly, and Miklos Porkolab and Paul Bonoli discussed lower hybrid performance under various operating scenarios. Earl Marmar reviewed budget issues.

Selected Domestic Travel

Miklos Porkolab participated in the MFE/ICF program leaders' meeting at General Atomics, San Diego, on July 20,21. This is the first time that the leaders of the two communities have met to discuss improved communication in their common quest. Agreement was reached for a joint workshop, to be held at PPPL in mid-September, at which information on technical issues will be shared.

Steve Wolfe took part in the Conceptual Design Review for the NSTX Control System, held in late July at PPPL.

Earl Marmar attended the July 29th FESAC meeting in Gaithersburg, MD., and the DoE Energy Research Strategy Plan Workshop, Chantilly, VA. on Aug 7 and 8.

Amanda Hubbard, Miklos Porkolab, Spencer Pitcher and Joe Snipes attended the DIII-D Brainstorming meeting at GA September 20-22, presenting a total of 9 ideas for the 1999 DIII-D experimental campaign.

Following his visit to GA, Spencer Pitcher attended the NSTX PAC meeting at PPPL.

Bob Granetz gave a seminar at Dartmouth October 7 on ``H-modes and the Edge Pedestal in Alcator C-Mod''.

International Travel

Steve Wukitch returned from Garching, Germany at the end of July. During his 3 week visit, he participated in ICRF experiments on Asdex-UG. Most of the experiments focused on measuring the power deposition profiles from mode converted ion Bernstein waves using a power modulation technique. Successful measurements of the deposition profile were made, and the data will be compared with simulations using the TORIC code. Poloidal flow measurements were also attempted; the data have yet to be analyzed. Piggyback experiments were performed, looking at 2nd harmonic damping on deuterium beam ions and preliminary analysis indicates the presence of a deuterium tail; detailed analysis of these results is ongoing.

Ian Hutchinson attended the Editors meeting of the New Journal of Physics, to be launched in October. The journal is targeted as a high-calibre, all-electronic, fully refereed physics journal, freely available on the web and taking advantage of its new media opportunities. Ian hopes that plasma physics will be strongly represented in the journal's papers.

In early September Martin Greenwald attended a meeting of the European-Transport- Task force in Goteborg, Sweden. The meeting was attended by 75-80 researchers. Major topics covered were: core transport scalings, stellarator, RFP, tokamak comparisons, fluctuations, transient transports, improved confinement regimes, and edge/core interactions.

Presently a large fraction of the physics staff is attending the IAEA meeting in Yokohama, Japan. Paul Bonoli, Bob Granetz, Martin Greenwald, Ian Hutchinson, Bruce Lipschultz, Earl Marmar, John Rice, and Steve Wukitch are participating. Six C-Mod papers are being presented:

Overview of recent results for the Alcator C-Mod Tokamak, presented by E. Marmar

H-mode regimes and Rotation in Alcator C-Mod, presented by M. Greenwald

Detached divertor plasmas in Alcator C-Mod: a study of the role of atomic physics,

presented by B. Lipschultz

Scaling of H-mode pedestal characteristics in DIII-D and C-Mod, presented by R. Granetz

ICRF heating experiments in Alcator C-Mod, presented by S. Wukitch

Modelling of advanced tokamak physics scenarios in C-Mod, presented by P. Bonoli

In addition, Miklos Porkolab will present ``Advanced tokamak burning plasma experiments''.

Near Term Plans

Our near term plans are focussed on the up-coming run campaign, scheduled to begin in December. This entails reassembly of the machine, installation of the in-vessel components, including the new controllable, divertor by-pass ``flappers''. The new RF antenna must be installed, as well, along with the DNB and its associated diagnostics. The standard set of diagnostics is also being re-assembled. Proposals developed from the Ideas Forum presentations are being written.

Forty-two C-Mod poster and ten oral presentations will be given at the APS Division of Plasma Physics Meeting, Nov. 15-20 in New Orleans. This total includes thirteen presentations from our non-MIT collaborators. In addition, three of the C-Mod staff were selected to give invited talks at this meeting: John Goetz, on the C-Mod achievement of high core confinement with a dissipative divertor, Martin Greenwald, on C-Mod H-mode confinement regimes, and Bruce Lipschultz, on particle sources and sinks in the divertor.