27 June 2001

Four topics were on the agenda for this meeting:

Brad Nelson addressed progress in the layout of the PFCs.  The steps are as follows.  First, Physics (Mioduszewski) needs to define a stay-out zone which only the divertor plates can intercept.  In addition, Physics needs to define the required pumping speed in each of the divertor regions.  This information will be used to size the divertor cryopumps and the divertor slot size.  Given the cryopump and slot size, an envelope for the pumped divertor can be established.  The vacuum vessel boundary can then be established.  The modular coils can then be offset from the vacuum vessel boundary.

For the engineering design, the stay-out zone will be defined by a set of points derived from following field lines.  The data set will be provided by Art Grossman and modeled in Pro/E.  In the concept layout shown by Nelson today, there were two cryopumps shown.  It appeared that there would be much more space outboard than in the tip of the bean-shaped cross-section for cryopumps.  Consideration should be given to moving both cryopumps into the outboard region and thereby minimize the offset required to the coils in the vicinity of the tips of the bean-shaped cross-sections.  

The cryopumps were copied from those used on DIII-D.  That appeared to be an excellent idea.  Guidance from Physics (Mioduszewski) was that 5,000 l/s of pumping capability should be provided in each of the six divertor regions.  

For developing coil designs using CoilOpt (and the newly coupled CoilOpt-StellOpt codes), Neilson suggested that the "bloat parameter" in VMEC be used to establish the stay-out zone.  The modular coils could then be offset (non-uniformly) from the stay-out zone.  Physics should determine what bloat parameter best approximates the stay-out zone.  ORNL would then provide the necessary coil offset information to Strickler and Ku.

Georgiyevskiy indicated that the capability to add a radial field has proven very useful for divertor operation on Uragan-3M and should be considered as a requirement for NCSX.  The thought here is that if there are any radial error fields, we could compensate for them if we had the capability to add a radial field (perhaps as an upgrade).  In addition, Georgiyevskiy reminded us to include divertor diagnostics in our diagnostics complement.

Mike Cole reported on progress in accommodating the inboard RF.  Cole presented a scheme for routing the coax feeds to the RF launcher.  Several concerns were discussed.  Brown noted that thermal insulation should be provided around the coax pipes to reduce the heat leakage (and trace heating that would need to be provided).  Cole was concerned about the feasibility of installation and maintenance of the coax feeds once the machine had been assembled.  Cole will discuss the RF launcher and coax feed concept with Majeski and complete his assessment of the feasibility and design impacts of adding a requirement to accommodate the inboard RF.

Dennis Strickler and David Williamson discussed a new coil design with 18 modular coils.  The coil design was generate using an algorithm to ensure adequate NB access.  The design had only three coil types (instead of four) and only 18 coils (instead of 21) which should translate into significant cost saving.  The design should have improved access because of fewer coils and the absence of any modular coils on the symmetry planes.  Dennis Strickler passed the coil data on to Art Brooks.  Brooks subsequently posted the coil data in the following location:

ftp://ftp.pppl.gov/pub/abrooks/NCSX_COIL_DATA/li383/modular/06_20_01/coils.li383_0620a7.

Long-Poe Ku agreed to check the physics properties for the new coil set.  Ku also agreed to test whether there was any benefit to powering the TF coils independently.  Ku indicated that the best case to date was the M12 coil set.  One of its drawbacks however, was a tighter radius of curvature.  It was not clear whether the tight radius of curvature in a few locations was fundamental to the improved performance or an incidental feature from using more modes in generating the coil geometry.  Williamson agreed to locally modify the radius of curvature in the regions where it was too tight and pass the modified coil set back to Ku for evaluation.

Reiersen presented a 5 coil, axisymmetric PF coil design.  The design appears to be able to provide much better nullapole (OH) and quadrapole fields.  Virtually all of the improvement was shown to be due to the addition of a PF coil above the plasma at the approximate location of the magnetic axis.  Physics was requested to test the performance of this PF coil set to see if benefits were apparent when using StellOpt.  If so, a change proposal would be drafted to revise the reference design.

Next Wednesday is July 4.  Happy Independence Day!  The next project telecon will be July 11 at the regular time and place.

Please forward any comments to reiersen@pppl.gov

(last edited on 09/25/2001 03:47 PM )