Summary of August 26 SIT Meeting

Development of new modular coil design. Dave Williamson provided a modified winding surface that moved the point of closest approach out by 3.8cm.  Art Brooks developed a Fourier representation of this winding surface in the same coordinate system as the reference design.  Art Brooks also developed the Fourier representation for a surface midway between the reference and modified (+3.8cm) surfaces in case the 3.8cm perturbation was too large.  Mike Zarnstorff will use the new surface representations in the optimization studies.

PF design modifications. The question arose of whether there were any design modifications to improve performance or reduce cost that should be made to the PF design at this time.  Mike Zarnstorff indicated that PF3 and PF4 could likely be combined into a single coil of the same size as either PF3 or PF4 in order to reduce cost.  The coil currents required would be on the order of 1MA for the shaping function along, but higher when the OH function was added.  Brad Nelson agreed to propose a revised coil geometry with the current centroid about midway between PF3 and PF4 and approximately the same cross-sectional area as PF3 (keeping the number of layers a multiple of 4 for field error mitigation).  Mike Zarnstorff will incorporate the revised PF coil set into his modular coil optimization studies.

Central Solenoid (CS) configuration. Trade studies have been performed to explore bucking the TF coils off the CS coils (PF1 and PF2) and off a separate structure in the CS assembly.  Reiersen asked for a recommendation on which way to go in order to incorporate the preferred option into the design change planned for the improved modular coil design.

Poloidal time constants.  Calculations by Nelson and Brooks indicate that the poloidal time constant of the modular coil structure is on the order of 50ms without a sprayed copper surface on the tee and 250ms with a sprayed copper surface on the tee.  The modular coil currents are changing over relatively short times (58ms in the 1.7T High Beta scenario).  Mike Zarnstorff judged that even a 50ms time constant was probably not acceptable.  10ms would be categorically acceptable.  Significant analysis would have to be done in order to determine the acceptability of an intermediate value.  Brad Nelson agreed to look into providing a poloidal break.  In parallel, Nelson will coordinate more definitive calculations of time constants, including configurations with cutouts, partial breaks, and complete breaks.

Miscellaneous.  Reiersen reminded folks that initial FY03 work plans were due Friday, August 23 and feedback on the GRD is due September 6.  He also requested suggestions for agenda items for the September 20 project-wide engineering meeting.

The next SIT meeting will be held next Tuesday, September 3 at 11am.

Please forward any questions or comments to mailto:reiersen@pppl.gov

Return to NCSX Engineering Home Page