Alcator C-Mod Quarterly Progress Report – October 2003

The primary activities at Alcator C-Mod during the third and fourth quarters of FY03 were operation of the device and inspection of the alternator.

Science Results

Core Confinement and Transport

Analysis of fast dynamics at the L-H transition, carried out in collaboration with Benjamin Carreras, ORNL, during his recent visit, has revealed some interesting results. The primary diagnostic was the high resolution, low noise FRCECE diagnostic (U. Texas). By using discharges with off-axis RF heating, the effects of sawtooth heat pulses were minimized. On several channels, there is a two-stage transient; an initial, near-linear 'jump' in Te over a period of order 1 ms, followed by a slower, apparently diffusive response at all radii. Somewhat unexpectedly, the biggest jump in Te is at about 87.9 cm, 1.5 cm inboard of the separatrix; the prompt response extends from 86-89 cm. A model by Carreras and del-Castillo-Negrete gives qualitatively similar behavior. The simulations indicate that the scaling of the size and rate of the initial 'jump' with P-P(thresh) would be the best test of the model; as power is raised above the threshold, the size of the jump should get larger, and the duration shorter. It is therefore planned to investigate the dynamics more systematically using a controlled power scan.
 
Anomalous momentum transport has been observed in C-Mod plasmas. The time evolution of core impurity toroidal rotation velocity profiles has been measured with a tangentially viewing crystal x-ray spectrometer array. Following the L-mode to EDA H-mode transition in both Ohmic and ICRF heated discharges, the ensuing co-current toroidal rotation velocity, which is generated in the absence of any external momentum source, is observed to propagate in from the edge plasma to the core with a time scale of order of the observed energy confinement time, but much less than the neo-classical momentum confinement time. The ensuing steady state toroidal rotation velocity profiles in EDA H-mode plasmas are relatively flat, with Vf = 50 km/s, and the momentum transport can be simulated with a simple diffusion model. (see Fig.1) Assuming the L-H transition produces an instantaneous edge source of toroidal torque (which disappears at the H- to L-mode transition), the momentum transport may be characterized by a diffusivity, with values of ~ 0.07 m2/s during EDA H-mode and ~ 0.2 m2/s in L-mode. These values are large compared to the calculated neo-classical momentum diffusivities, which are of order 0.003 m2/s. Velocity profiles of ELM-free H-mode plasmas are centrally peaked (with Vf(0) exceeding 100 km/s in some cases), which suggests the workings of an inward momentum pinch; the observed profiles are consistent with 
  Figure 1
 
 
simulations including an edge inward convection velocity of ~ 10 m/s. In EDA H-mode discharges that develop internal transport barriers, the velocity profiles become hollow in the center, indicating the presence of a negative radial electric field well in the vicinity of the barrier foot. Upper single null diverted and inner wall limited L-mode discharges exhibit strong counter-current rotation (with Vf(0) ~ -60 km/s in some cases), which may be related to the observed higher H-mode power threshold in these configurations. For plasmas with locked modes, the toroidal rotation is observed to stop (Vf<5 km/s).

MHD

The latest passive MHD spectroscopy results on C-Mod show that in addition to the usual Toroidal Alfven Eigenmodes, we now observe Alfven Cascades (ACs) in the current rise with strong ICRF heating (see Figure 2) as well as high frequency modes in flattop EDA H-mode discharges centered on the gap of the Ellipticity-induced Alfven Eigenmode (EAE) frequency. Such Alfven Cascades are observed in JT-60U and JET at the onset of ITBs when they have reversed shear q profiles. Modeling with the MISHKA code through our JET collaboration indicates that the q profile at the time of the Alfven Cascades has slightly reversed or very flat shear with a minimum q value of 3 at the minimum AC frequency. By fitting the frequency evolution of the ACs, it is possible to model the evolution of the q profile and show that the q profile remains flat or slightly reversed to a minimum q value of 2.5 as the next n=2 AC appears. Then, the modes evolve into TAEs indicating that the q profile becomes monotonic.
 

  Figure 2
 
In a few relatively low density (2 x 1020 m-3) EDA H-modes a series of modes are observed with frequencies from 700 kHz to 1100 kHz centered on the EAE frequency gap. The surprising feature of these modes is that they rotate in the electron diamagnetic drift direction, which is opposite to what would be expected for EAE modes with core ICRF heating. If the fast ion profile were hollow, it would be possible for EAEs to rotate in the electron direction.

Analysis of locked mode data with Ip = 1 MA, BT = 5.4 T indicates that the range of densities at which the mode appears is reduced from 1.7x1020/m3 with adverse (i.e. destabilizing) applied 2/1 field to less than 4x1019/m3 with stabilizing applied field. The factor of >4 in the accessible density indicates that we are able to compensate the intrinsic error very well in the "best" case. As previously noted, the C-Mod experiments appear to contradict earlier size scalings for the threshold error field for mode locking which predicted Block ~ R-9/5 by a factor>10. The threshold density inferred from 1MA, 5.4T C-Mod discharges is about a factor of 2 above predictions of dimensionally constrained scalings based on single-machine fits of DIII-D and JET data.

Divertor and Edge

A systematic analysis of plasma profiles in the edge plasma has been performed, drawing on high spatial resolution fast-scanning probe and edge Thomson data collected over a wide variety of discharges including Ohmic L-mode (forward and reversed magnetic field), and Ohmic H-mode (Elm-free and EDA regimes). The scanning probe and Thomson data are found to match reasonably well at the separatrix and show the same values of an increased electron pressure gradient scale length at increased plasma densities (in otherwise identical discharges). These observations lend confidence to the data set. Focusing on data collected in the Near scrape-off layer (i.e., ~2 mm beyond the last closed flux surface), two interesting and perhaps fundamentally important observations have been made:

(1) Despite the wide range of external control parameters in these discharges (plasma current, toroidal field, density) the local pressure gradient scale length in Ohmic L-mode discharges appears to depend solely on the local value of the 'diamagnetic parameter', a_d, a parameter that has been identified to regulate the character of drift-resistive ballooning turbulence in 3-D turbulence simulations.

(2) For fixed values of a_d in the Near SOL, the magnitude of the local electron pressure gradient is found to scale as plasma current squared, precisely the scaling required to maintain the MHD ballooning parameter, a, invariant. Thus the Near SOL plasma lies on a well defined 'trajectory' when plotted in (a,a_d) space. These observations offer very strong evidence that gradients in the Near SOL are set predominately by drift-resistive ballooning mode physics.

H-mode discharges appear as a jump to a new trajectory in (a,a_d) space, while reversed magnetic field discharges lie on a trajectory with reduced values of a compared to normal magnetic field discharges. The latter result suggests that other parameters influence the (a,a_d) state of the near SOL, such as poloidal velocity shear, which is found to be reduced in reversed magnetic field discharges. Further analysis is in progress.

Operations and Diagnostics

The 132 GHz reflectometer channel has become operational and initial data look very good. The auto power spectra show the expected specular reflection, MHD activity, and a broad background of density fluctuations. The coherent scattered power, which is an indication for the density fluctuation level varies when the plasma evolves and at times there is even a weak cross correlation visible with the 88 GHz channel.
 
 
The NSTX X-ray imaging crystal spectrometer was installed on C-Mod 5/24. Work continued on bringing up the NSTX imaging Xray Crystal Spectrometer. The crystal was rotated slightly to improve the centering of the argon spectrum on the detector. The crystal was also masked down to reduce the detector count rates to acceptable levels. Calculations indicate that the etendu of this spectrometer is much higher than that of the single chord C-Mod crystal spectrometers, consistent with the observation of symptoms of high count rate and detector/electronics saturation during C-Mod discharges. Measurements were made to better understand the apparent high deadtime fraction of the signal-processing electronics at high count rates, and to assess the individual contributions of the TDC and the KBSI interface box to this deadtime. Measurements on C-Mod plasmas with this instrument will continue.
 

Diagnostic Neutral Beam

Analysis of data from the deuterium DNB run last continues. The goal of this run was to investigate the slowing down and/or loss of fast beam particles by looking at their signature on the DD fusion neutron signal. This analysis is quite relevant to RF heating on C-Mod, since these beam deuterons presumably undergo the same physics as the H-minority tail generated by ICRF. Making the simplifying assumption that the deuteron slowing down can be represented as a single exponential decay time, ts, allows for the determination of ts by fitting to the temporal behavior of the 2.45 MeV neutron signal. Since the slowing down times are comparable to the timescale for variations in the beam current, the fit must actually use a convolution integral. The principal findings are that a single ts does indeed fit the neutron rise, fall, and flattop behavior for each shot, and that the slowing down times agree with classical values in magnitude and scaling. Further analysis of absolute neutron rate is continuing, in order to put further limits on the prompt loss fraction. Simplified calculations show that the measured neutron signal turns out to be 25%-75% of the calculated rate, and that this ratio has a dependence on density. The effect of Zeff on the deuterium plasma density has been included in these calculations, but profile effects and beam penetration effects have not, and they could, in principle, explain the deficit and its density dependence. Therefore the simplified calculations show that to within an order of magnitude, prompt losses are small or non-existent, but in order to get better than a factor of ~2 accuracy, TRANSP runs will be required, and are indeed planned.
 

 
Alternator Inspection
 
All UT tests of the alternator rotor have been completed. The rotor end rings have been replaced, and the rotor is ready for DC hi-pot tests early this week. DC hi-pot tests performed last month of the stator showed low leakage and polarization currents on all windings, which would usually be considered a very good result. However, we had questions about the shape of the hi-pot I-V curves that we asked GE, the vendor doing the inspection, and outside consultants, to help us understand. The interpretation and correctness of these curves weigh heavily on our decision about whether or not a rewind of the stator is required, and in any decision to put the alternator back into service now, as has been recommended by GE. At the recommendation of GE, and one of our consultants, we have repeated the DC hi-pot tests with results consistent with those found during several inspections ranging back to 1977. These tests indicate a stator in very good condition that is ready for a return-to-service GE repeated the DC stepped voltage test to 28 kV on the alternator stator. This test was followed by a hi-pot proof test to 36.7 kV. All stator windings passed these tests with excellent I/V characteristics. As a final test of the stator windings, an AC hi-pot to 21.6 kV RMS (a factor of 1.5 above the working voltage) was also performed. The results were again very positive and GE has recommended the alternator be returned to service. AC hi-pots are considered to be the most probative tests of the stator windings and an excellent predictor of safe in-service operation.
 

Computing

Lower Hybrid Project

A draft schedule covering all aspects of the launcher rework necessitated by the replating has been generated. This includes repair and replating of the plasma-facing coupler, disassembly, replating and reassembly of the forward waveguide, and comprehensive inspection and testing of all components before installation on C-Mod.
 
The "old" copper and underlying base have been stripped off the coupler. The stripping bath also etched away some of the braze alloy sealing the ceramic windows to the coupler frame, resulting in vacuum leaks. A coupler prototype is being used to duplicate this effect and to verify the rebrazing, cleanup and replating processes prior to repair and replating of the full coupler.
 
The forward waveguide assembly has been fully disassembled at PPPL. All of the stainless steel plates comprising the stacked assembly have been stripped of old copper and underlying base. They have been carefully cleaned and measured to assure dimensional tolerances after plating. All 50 plates have been replated and remeasured. 17 plates have gone through a 200 C vacuum bake at MIT as part of verifying the plating integrity.
 
Measurements of initial phase settings of the rear waveguide have been completed at PPPL, and the unit with all its components was delivered to MIT 7/2. The rear waveguide has been unpacked and placed inposition in the high power test area. The waveguide run to the coupler from the klystron as well as tapers needed to couple into the
waveguide are nearly fabricated at MIT.
 
A meeting between PPPL and MIT personnel was held at MIT on 9/3 for discussions on finite-element modeling of the lower hybrid coupler windows. Initial measurements of the coefficient of thermal expansion have been performed both for the titanium and the alumina that is used for the couplers, and the results are being evaluated and compared against published values. The vendor has started braze alloy selection tests and will subsequently braze coupler prototypes. The upper Forward Wave Guide has been reassembled, and is ready for shipment to MIT for high power RF testing. The lower FWG is in the process of reassembly bolting sequence optimization prior to low power RF testing at PPPL.
 

Meetings

Bruce Lipschultz co-chaired the ITPA meeting on SOL/divertor physics held in St. Petersburg Russia the week of July 14, following EPS. A number of studies on ELM cross-field transport were presented. They indicate that although the power delivered to the wall in an ELM may not be a concern, that the potential for sputtering of the wall was significant. A similar concern for the wall also came up with regard to disruption power deposition. JET either has poor energy accounting or a substantial fraction of the disruption energy is going to the wall (primarily during the current quench). The topic of the wall/blanket tile material was thus a hotly discussed topic. There is a concern that a Be wall tile will be melted during disruptions with the damaged surfaces leading to problems during 'normal' operation.
 

The C-Mod Ideas Forum was held on September 10, 11, and 12 to aid in planning the FY04 C-Mod operating campaign. The purpose of the Forum was to provide an opportunity for presentation of specific ideas appropriate for experiments during the campaign. The Forum was open to all interested parties, including current and potential future collaborators. The agenda of the Forum and links to the ninety-eight presentations can be found at
 
http://www.psfc.mit.edu/people/terry/Ideas_Forum_2003/forum_agenda.html
 
Ideas were submitted by personnel from PPPL, U. Texas, U. of Wisconsin, IPP-Garching, LANL, GA, RFX, LLNL, ORNL, CRPP, and MIT. Follow-on meetings of the Topical Science and Integrated Thrust Groups, in the coming few weeks, will determine initial prioritization as the next step leading to mini-proposal generation and run-time allocations.

Collaborations

Daniel Andelin, a National Undergraduate Fusion Fellow from Brigham Young University, completed his 8 week visit (June-Aug.) at C-Mod working under the supervision of John Rice, with help from Martin Greenwald. He worked on the creation of a database to study ICRF absorption under a variety of H-Mode conditions in C-Mod. The technique used to evaluate absorption involves examining the time derivative of stored energy in the plasma, derived using EFIT, shortly after fast transitions of ICRF power. The results show possible trends with plasma temperature and density, and the database should prove useful for guiding future studies.

Domestic Travel

Bob Childs and Rui Vieira were at PPPL on May 30 for ongoing evaluation of the LH launcher following the 150C bake cycle tests carried out last week.
 
Martin Greenwald attended a workshop June 3-5 in Reston, Virginia "DOE Science Networking". The workshop was charged by OASCR to draft a five year roadmap for networking for Office of Science programs.
 
Paul Bonoli visited PPPL on June 4-6, 2003 to attend a Program Advisory Committee Review of the OFES Sci-Dac Projects. Included in this meeting was a review of the RF Sci-Dac Initiative. 
 
Monty Grimes traveled to San Antonio, TX, the week of June 9 to inspect the waveguide needed for the high power launcher tests. He also traveled to San Jose, CA, to visit the vendor conducting the refurbishment and upgrades to two of our klystrons.
 
Miklos Porkolab attended the IPELS 2003 Conference (Interrelationship between Plasma Experiments in Laboratory and Space) on June 29 -July 3 in Whitefish, Montana, and gave the invited talk "Mode Conversion Processes at the Ion-Ion Hybrid Layer in Multi-Ion Species Plasmas in Sheared Magnetic Fields". The bi-annual international conference was attended by approximately 100 scientists from around the world, including space plasma physicists, laboratory plasma physicists, and fusion scientists. The theme is cross-fertilization of ideas among different branches of plasma physics. There was considerable interest among space plasma physicists in the C-Mod ion cyclotron wave mode conversion results for possible applications to explain energetic ion formation in space plasmas.
 
Tom Fredian and Joshua Stillerman attended the 4th IAEA Technical Meeting on Control, Data Acquisition and Remote Participation for Fusion Research, held at General Atomics in San Diego. Tom Fredian presented a paper: "Migration of Alcator C-Mod Computer Infrastructure to Linux." Tom Fredian and Joshua Stillerman also attended the 3rd MDSplus users group meeting, also held at General Atomics. Fredian chaired the meeting. There were approximately 40 attendees from the US, Europe, Japan and South Korea. A detailed discussion of "Long Pulse / Continuous extensions to MDSplus" was included in this meeting.
 
Miklos Porkolab and Martin Greenwald traveled to Gaithersburg, Md to attend a meeting of the Fusion Energy Sciences Advisory Committee on July 31-Aug.1.
 
Jim Irby was at PPPL Aug.5 to serve on the NSTX TF Flag review committee.
 
On Aug.19 Bob Childs visited PPPL to inspect the forward wave guide assembly and gather engineering info.
 
Martin Greenwald attended the 18th International Conference on Numerical Simulation in Plasmas the week of Sept.8 to deliver a review talk titled "Beyond Benchmarking - How Experiments and Simulations Can Work Together in Plasma Physics". This review talk may be found at
 
http://www.psfc.mit.edu/people/g/icnsp03-talk.pdf
 
On Sept.24, Earl Marmar, Martin Greenwald, Bruce Lipschultz and Miklos Porkolab travelled to Germantown to meet with OFES staff. Discussions centered on recent research highlights and plans for C-Mod. Attending from DoE were John Willis, Rostom Dagazian, Erol Oktay, Curt Bolton, T.V. George, Steve Eckstrand and Arnold Kritz.
 
Catherine Fiore, Dimitri Mossessian, Amanda Hubbard, Miklos Porkolab, Darren Ernst and Jerry Hughes (MIT), and Perry Phillips, Alan Lynn, and Matt Sampsell (UT-FRC) attended the 9th Annual IAEA TM on H-Mode Physics and Transport Barriers 9/24-9/26 in San Diego. Work was presented on control of ITBs in Alcator C-Mod (Fiore), and the role of TEMs and neoclassical transport in Alcator C-Mod ITBs (Ernst) . C-Mod/DIIID/ JET dimensionless scaling results for the edge pedestal were reported by D. Mossessian and the role of plasma and neutral transport in the edge pedestal in Alcator C-Mod was explored by J. Hughes. A. Hubbard presented her work on local threshold conditions and the fast transition dynamics of L-H mode transitions in Alcator C-Mod. 

Amanda Hubbard and John Rice were in San Diego Sept.29-Oct.2 at the ITPA meeting. Amanda attended the Pedestal sessions while John attended the Transport and ITB sessions.

Joe Snipes attended the IAEA Technical Meeting on Energetic Particles from 10/6 to 10/9 at General Atomics in San Diego where he presented a talk on Active and Passive MHD Spectroscopy on Alcator C-Mod. Paul Bonoli also traveled to General Atomics where he attended an ITPA meeting of the Steady-State and Energetic Particles Group.
 
Jim Irby was at PPPL from 10/7 to 10/9 to serve on the NCSX Preliminary Design Review committee. Along with Bob Childs, Ron Parker, Rui Vieira, and Steve Wukitch, he also attended the Lower Hybrid Coupler Brazing Review on 10/10.

International Travel

Joe Snipes, Dmitri Mossessian and Ian Hutchinson attended and presented papers
at the 30th European Physical Society Meeting on Controlled Fusion in St
Petersburg, Russia from July 7-11.
 
Bruce Lipschultz co-chaired the ITPA meeting on SOL/divertor physics held in St. Petersburg Russia the week of July 14, following EPS. 

Near Term Plans

Schedule

The short term schedule may be found at:

http://www.psfc.mit.edu/cmod/operations/schedule.txt