Alcator C-Mod Weekly Highlights Dec 27, 1999 The winter maintenance period continued at Alcator C-Mod last week. Engineering: ------------ Good progress was made on DNB development last week. The mod/reg was pulsed at 50kV and about 7.4A into dummy load. Response time of mod/reg rollback was measured by injecting a pulse into the dV/dt sense circuitry in the source area during a mod/reg pulse into the dummy load; the time was ~3usec from pulse injection to the start of rollback. Master control dV/dt interrupt counter operation and fault reaction was also verified. A test setup with water load, adjustable spark gap and wire tester was built and will be installed Monday. This unit will be used to test response of the mod/reg to HV breakdown in the source area; successful wire tests must be performed before the HV is connected to the beam, to verify that the amount of energy available to an arcdown is limited to < 5J. Work continued on maintenance and upgrades of the cryo systems. All of the LN2 filters in the cryostat were replaced. An improved feedthrough design for cryostat penetrations is being implemented. In the power systems area, inspection and maintenance of the HV breakers and switchgear continued; this task should be completed this week. In-vessel work included work on the F-Port ECE mirror and measurements in preparation for installation of neutral gauges. Work also continued on the installation of components of the J-port ICRF antenna. ICRF Systems: ------------- In the J-port antenna refurbishment, the pacing item is the re-plating of the large and small leads. We believe the plating problems were caused by the stripping solution attacking the joint material around the brazed joints and leaving a porous surface compound. The braze material is Nicrobraz and has a composition of 80% Ni and small amounts of C and Fe. The stripping solution is attacking Nicrobraz more aggressively compared to the inconel. The plan now is to strip and surface polish before replating. The faraday screen assembly is complete except for three units (48 total). The washer assemblies for these three units needed to have their resistive washers remade. The washers were curing over the weekend and testing will begin Monday. The shorting straps have been made up and a second fit up is required. The port cover and septum have been installed in the machine. The conductor covers and two current strap assemblies will be installed this week. A review of the BN tile replacement of E-port's front protection tiles was held. The BN tile design and fastener were approved with some modifications: spot weld the threadded post, wavy washer should be placed on top of the BN, improve compliance via thread removal, determine optimum number of fasteners based upon an upper bound for disruption induced acceleration, radius the face, and 10-15 degree slope on the tile side. A drawing is ready and the pieces can be manufactured up to the point of drilling the holes for the fasteners. This will await the calculation and evaluation of the disruption forces. The pacing item is the arrival of the 0.5" BN material from the supplier. Work on the upgrade of the water cooling systems for the FMIT transmitters continued on schedule. FMIT#1 and FMIT#2 water cooling system has been successfully upgraded and tested. The basic improvement was to increase the pressure drop across the FPA tube by reducing the output pressure and increase the overall flow. A heat exchanger and new pump were required to achieve this. The original flow was 150-160 gpm which would allow 2 MW dissipation power and the improved system is 190-200 gpm which allows 3 MW dissipation power. This will permit the transmitter to provide higher power output and withstand higher plate dissipation operation. Physics: -------- On Alcator C-Mod, two arrays of photodiodes measure the radial profiles of the soft x-ray emissivity at the outboard edge above the midplane, and at the top of the plasma. The x-ray emissivity is approximately proportional to the product of impurity and electron density. This allows us to study the timing of the onset of the collapse of the x-ray pedestal at the H-L transition. For cases with very high signal/noise ratios we find that the pedestal starts disappearing at the outboard edge before it starts disappearing at the top of the plasma. The time delay is consistent with the time it takes a plasma perturbation to travel along a field line from the detector view at the outboard edge to the top of the plasma, if one assumes that the plasma perturbation propagates at the ion sound speed; however, not enough instances have been examined so far to establish a scaling with plasma parameters. A commonly-used EFIT MDSplus tree structure would allow fusion labs to more easily share analysis codes that access EFIT data, as each code would need only one MDSplus interface to the EFIT tree. During Jeff Schachter's visit to C-Mod last week, he and Steve Wolfe worked on a plan to unify the C-Mod and DIII-D EFIT tree structures. Inconsistencies between the two implementations were catalogued, and a scheme for reconciling them agreed upon. The plan is for the trees at each site to contain the superset of the output nodes presently in each experiment's structure (with the exception of a few site-specific outputs); redundant nodes will contain pointers (TDI expressions) to avoid duplicating I/O and storage requirements. This plan allows users at either site to read data directly from the tree using the nodenames with which they are familiar. User codes with embedded tree paths would need only minimal modification to achieve full portability. The plan also will allow DIII-D to adopt the C-Mod routines for direct MDSplus I/O in the EFIT code, when this capability is implemented at GA. A prototype of the unified tree has been produced at the C-Mod site. Travel and Visitors: ------------------- Miklos Porkolab attended the VLT Program Advisory Committee Meeting in Albuquerque on Dec. 15,16. Jeff Schachter from DIII-D visited C-Mod last week. He worked with Steve Wolfe on unification of the MDSplus interface to the EFIT code, and with Martin Greenwald on database schemas and graphical interfaces.