The primary activities at Alcator C-Mod during the first quarter of FY98 were: execution of a run campaign which began October 30, 1997, presentation of Alcator C-Mod results at the meeting of the American Physical Society - Division of Plasma Physics, continued work on major systems upgrades (the RF transmitters and the Diagnostic Neutral Beam), final repair of the alternator which powers most of the machine magnets, and continued participation in the critical assessment of fusion science. A more detailed account of these activities follows.
Scientific Results
The Alcator C-Mod team is presently involved in a productive run campaign which has yielded a number of new results.
Core Confinement Results
Measurements of H-mode pedestal characteristics are currently of interest, primarily because of the strong influence the edge temperatures have on the core temperatures. Because of C-Mod's small size and high field relative to other major tokamaks, its pedestal width and characteristics have significant leverage on multi-machine scaling relations. A number of diagnostics, including electron cyclotron emission and the edge x-ray and XUV arrays, exist to study the pedestal region. In recent experiments the toroidal field was swept over about a 5% range to allow the ECE to map the pedestal shape. Data were obtained at 1MA and at 0.6 MA. 0.6 MA H-mode discharges have rarely been run on C-Mod; these had indications of enhanced Da (EDA) behavior, i.e. a steady state density and radiated power is achieved late in the H-mode. (Typically in ELM-free H-mode the density and radiated power increase monotonically.) It will be interesting to explore whether this is a reproducible effect at lower Ip. Preliminary analysis of ECE data shows similar widths to the EDA H-modes measured in 1996, ( ~ 10-12 mm raw, 6-9 mm deconvolved) but the pedestals are apparently further in from the LCFS by ~ 6 mm. It is still under investigation whether this is a real difference or indicates a systematic error in the calculation of Grating Polychromator (GPC) radii and/or EFIT flux surfaces. Pedestal heights, and confinement generally, were not as good as in our best 1996 discharges. Comparison of different diagnostics and scalings with conditions will be carried out in the present run campaign.
C-Mod and JET are collaborating in a study of dimensionless scaling through a set of ``dimensionless identity'' experiments. This widely used concept assumes that, if the dimensionless core parameters (r*, b, n*, q, shape, etc.) are the same, then the scalings of the dimensional parameters are determined. The collaboration is attempting to test the validity of this concept by making plasmas in each machine which are similar in dimensionless parameters. Two sets of engineering parameters which should do this have been identified, and we have run one of the sets on C-Mod. Up to 2MW of RF power was coupled into 1.2 MA JET shaped plasmas. These discharges produced ELM-free H-modes which generally have more impurity accumulation. RF coupling to these plasmas is also more difficult. The same shape at 1MA is also ELM-free, whereas the normal, higher triangularity shape at 1MA are EDA. This is strongly suggestive that it is the triangularity that controls the EDA, although there are of course simultaneous changes in the divertor strike-point geometry that also could have an effect.
Initial attempts to create Reversed Shear plasmas in C-Mod have been made. In these plasmas hollow profiles were obtained and various MHD modes, including TAE modes, were observed. Also observed was a delay in the onset of sawteeth (due to localized ICRF heating). However no obvious evidence of the ERS confinement regime was seen. In these plasmas Te was greater than Ti, and no internal fueling was used. In future experiments pellet fueling and He3 minority-heating will be used, which should heat ions. Both changes should have the effect of stabilizing the eta-i mode with Ti greater than Te.
``Cold pulse propagation'' experiments were performed under the direction of Ken Gentle (U. Texas- FRC). These experiments sought to produce, by the laser ablation injection of impurities, temperature perturbations which were suitable for analysis of thermal transport. The specific goal was to find discharges similar to those in TEXT and TFTR in which edge cooling induced core temperature rises. These were circular limited ohmic discharges, generally at low density. The core temperature rises observed in TEXT and TFTR are particularly difficult to explain and offer a unique set of conditions for testing transport theories. C-Mod target plasmas were inner-wall limited circular equilibria at low density. Plasma current was varied between 0.3MA and 0.5MA, at a fixed toroidal field of 5.4T. Various impurities - carbon, scandium, niobium, etc. - were tested for their ability to induce suitable increases in radiation and reduction in edge temperature. Neither carbon nor scandium produced significant perturbations, but both iron and niobium did. Core temperature increases were not found under any conditions. Comparison of the injections on these circular, limited discharges with standard 1 MA diverted reference shots showed that injection efficiency is equivalent for the two configurations. The niobium and iron injections are suitable for electron thermal transport analysis, but the analysis will be complicated by the need to include impurity transport for the space-time profiles of radiation losses.
Disruptive density limits and the density limits for the H-to-L transition have been investigated. The approach was to establish a high density H-mode and slowly ramp down the plasma current thereby lowering the density limit nGreenwald µ (I/a2). A strong deterioration in H-mode confinement at the higher densities was found. Both global and local parameters deteriorated toward those of L-mode, however clear H-to-L transitions could still be seen on the edge Te and x-ray measurements. H-to-L transitions occurred at n/nGreenwald = 0.75-0.8. Temperatures and stored energy continued to fall in the subsequent L-mode phase, which were terminated by disruptions at n/nGreenwald = 0.9-1.1. On one shot, only 1 RF transmitter operated and the plasma stayed in L-mode. The density apparently ran up to 5.7×1020 m-3 which is the highest density gas fueled discharge seen to date. Strong marfing was detected so the interferometer density measurement might be suspect. This shot will be examined in more detail.
Divertor and Edge Results
Experiments continued in order to quantify the effects of neutrals on edge dynamics. These experiments include the investigation of the potential role of neutrals in the formation and sustainment of H modes. Measurements of ``radiated'' power through escaping neutral particles are now being made. Typically two measurements are compared. One is made by normal foil bolometers, while the other is made using AXUV diodes, which measure photon radiation only. These observations indicate that a significant source of ``radiation'' loss is carried by neutrals just inside the separatrix. This effect is seen clearly during experiments where nearly pure helium plasmas were compared with plasmas having an approximately 50-50 deuterium-helium mix. The results are summarized in Fig. 1, where the measured total ``radiated power'' (in MW) for photons and neutrals (solid line) and photons only are compared (dashed line). In the first case, only helium was present (for which charge-exchange neutrals are virtually non-existent). In the second case an approximately 50-50 mix of deuterium-helium was used. The power measured including the neutral component is clearly greater than the radiation-only power in the 50-50 mix case. It is also found that it is more difficult to obtain H-modes in the pure He plasma, which leads us to hypothesize that edge neutrals are playing a role in the H-mode formation.
Deep Tracer Gas Injection System
A new intensified, fast-gating camera and a new coherent fiber bundle (both on loan from Princeton, courtesy of Stewart Zweben) were installed to record images of impurity line-emission `plumes' generated by the fast-scanning gas injection probe. With exposures below 0.5 msec, the new system is capable of `freezing' the motion of the fast-scanning probe. The intensified camera can detect impurity line emission at reduced levels of impurity injection, allowing the system to operate more like a true `trace-gas' injection diagnostic. A system that allows the camera exposure times and durations to be synchronized with the scanning probe injection was also put into place. New impurity gases of Ethylene (C2H2) and Freon-12 (CF2Cl2) were acquired to perform impurity dispersal experiments (via CII, CIII, CIV plumes) and scrape-off layer screening experiments (monitoring levels of Cl in the core). However, just prior to performing this experiments, it was discovered that the inertial valve in the scanning probe head had become frozen in the closed position. Experiments with this system are therefore on hold until the valve (which is under torus vacuum) can be repaired.
Scrape-off Layer Transport Analysis
The newly developed EDGEFIT code (work of graduate student Maxim Umansky) has been used to analyze over 500 time-slices in the edge database. Values of effective c^ at the separatrix scale in a similar way to those obtained with the 1-D `onion-skin' analysis method. Regression analysis shows that c^ near the separatrix is independent of magnetic field strength (2.7 < BT < 8 T) and has a weak inverse dependence with local density (ne-.5). Similar to previous analysis, EDGEFIT results show that effective c^ values appear to depend on the magnitude of the divertor bypass leak. An implication of these results is that neutrals play an important role in determining the scrape-off layer heat transport. This modelling result is qualitatively consistent with the experimental results mentioned above concerning the important of neutrals as a significant power loss channel in the edge. Direct comparisons between transport coefficients derived from EDGEFIT and UEDGE are planned.
Using the UEDGE code in conjunction with simple estimates, graduate student Maxim Umansky has found that in order to fit the measured edge density profiles, some combination of a spatially varying diffusion coefficient or an inward particle pinch is required. Work is currently under way to model the SOL density profiles consistent with the measurements of midplane neutral pressures. The effect of divertor bypass leakage is also under investigation.
Impurity Compression
Investigations of impurity compression in the divertor (MiniProposal #167) using the divertor RGA and core impurity diagnostics. Impurity compression is defined as the impurity density in the divertor divided by the impurity density in the core. In 1MA ohmic plasmas, data were obtained with Ar at 3 different densities. The compression of puffed Kr was also studied. An analysis of the effect of impurity mass on scaling awaits MIST analysis and further calibration of the RGA with Kr. Divertor detachment is observed to reduce the compression ratio.
Recombination and Opacity of D0 Lyman Line Emission
The ionization/recombination balance is affected by trapping of Lyman series emission. The primary effect on the ionization/recombination balance is to lower the temperature at which the ionization rate equals the recombination rate for ne = N0. Trapping of Lya,b has been observed in C-Mod and has been described in previous quarterly reports. In collaboration with A. Pigarov and S. Krasheninnikov the effects of Lyman series trapping on the recombination rate and on the number of recombinations per D0 Balmer series photon has been quantified. A diffusive model for the transport of the Lya,b photons out of a uniform volume of absorbers (D0 atoms) was used, but diffusion of photon wavelength away from line center was not included. The effect on the number of recombinations per Dg photon is shown in Figure 2. The top solid curve is for the optically thin case, applicable for N0 DL £ 4×1017 m-2, where DL is the length of the volume along the line-of-sight. The other curves show the effects of trapping on the recombination rates. The trapping depends on the number of absorbers along the line of sight, N0 DL. Each curve is labeled by this quantity (1.0(18) means 1.0×1018). In the opaque case the 3-body recombination rate is reduced by a factor of about 5, as is the number of recombinations per Dg photon.
RF Research
The ICRF systems have performed well this run period. In general the RF system coupled 2.5 MW reliably (without faults) and < 3.0 MW with some faults during RF injection. All experiments employed H minority in deuterium plasmas. The H/[H+D] was measured to be about 1-2% spectroscopically (J. Weaver, U. Maryland) and 2-4% according to PCX (R. Boivin) for typical discharges. In order to ensure the ECE is not cutoff at high density, it is sometimes desirable to operate a a toroidal field of 5.7 T, which moves the H resonance about 0.1 m toward the low field side. The absorption was not significantly different compared to typical Bt=5.4 T H-modes from this campaign.
MHD Results
Through very fast sampling (2 MHz) of poloidal field pick-up coil signals, high frequency (580 kHz - 700 kHz) modes were observed during the Enhanced Da (EDA) H-mode. The modes appear well into the H-mode in discharges that start with relatively low target density (ne < 1.5 ×1020 m-3) but reach normalized beta values exceeding 0.7. In EDA discharges with much higher target density and in ELM-free discharges, the high frequency modes were not observed. In several cases, the modes persisted for about 10 msec after the ICRF was turned off though still in EDA H-mode. This time-scale is about equal to the slowing down time of a fast ion tail driven by the ICRF. These modes are probably driven by the fast ions and may be Toroidal Alfven Eigenmodes or Energetic Particle Modes.
Further Results of the Discharge Cleaning Studies
During the last quarter we performed controlled experiments to quantify the plasma conditions of our discharge cleaning (ECDC) plasmas, and to quantify effectiveness in removing carbon which had been deposited on a stainless steel probe. These studies are of interest to ITER for a number of reasons, perhaps the most important of which is the problem of tritium trapping in the codeposited graphite layers. The graphite and the buried tritium must be periodically removed, and one proposed method for doing this is by electron cyclotron discharge cleaning. In C-Mod a radial profile of the carbon removal rate by ECDC was obtained by moving the exposed part of the sample relative to the resonance location. The thickness of the carbon layer remaining after exposure has been now measured using proton backscattering by our collaborator Bill Wampler at Sandia Nat. Lab. The major results were:
Operations and Diagnostics
Tokamak plasma operations resumed on October 30, 1997. Good plasma discharges were obtained almost immediately, due primarily to extensive discharge cleaning and a long period of time with the vessel under vacuum. Three days of tokamak operation were used for RF conditioning and H-mode pedestal investigation before a fault in the newly rewound rotor occurred. This forced a 3 week delay in the run campaign, while the alternator was repaired (under warranty). Operations