Milestones (Reiersen)
Recent developments on coil design (Williamson)
03 October 2001
Presentations
|
Milestones (Reiersen) |
|
Recent developments on coil design (Williamson) |
Minutes
The weekly project meeting was held on October 3. Wayne Reiersen opened the meeting with a discussion of near term milestones. The first priority is developing a buildable, consistent stellarator core design. Four paths are being followed to develop a buildable design:
|
Put more weight on the engineering metrics in CoilOPt (Strickler) |
|
Relax the coil-vessel separation to see if this results in a quantum change. Physically, this corresponds to moving the integral structure inside the coils and eliminating the vacuum vessel. (Strickler) |
|
Manual smoothing the winding surface and moving the coils to solve the coil-to-coil overlap, coil-to-vessel separation, and minimum radius of curvature problems that we continue to encounter. (Williamson) |
|
Run the merged optimizer to see if freeing up the constraint of a fixed plasma boundary can relieve the difficulties encountered in coil design. (Zarnstorff, Ku) |
The second priority is developing an updated set of work plans and budget allocations to carry the project through the CDR. A series of coordination meetings to deal with requirements, scopes of work for the CDR, budgets, and work plans have been set up. The kick-off meeting is scheduled for October 5.
David Williamson provided an assessment of the 0918a14 design. This design had a low B-norm error (0.58% average, 2.2% max). Two areas where the windings overlapped were noted. Concerns were raised about supporting the coils off a shell with the winding orientations as shown.
Concerns were also about the coil-vessel separation. Subsequent to the meeting, Tom Brown clarified that the coil-vessel separation needed to be 3.3cm (1.3"). This is approximately twice the minimum separation on the previous modular coil design.
Dennis Strickler reported on a subsequent refinement to this design, 0918a17. In this option, Strickler pushed the engineering a little harder and got modestly improved engineering metrics with only a small impact on the B-norm error (up from 0.58% to 0.62% average, 2.2% to 2.4% max). The consensus was that this latest refinement would be the best point of departure for manual manipulations by Williamson and for merged optimizer runs by Zarnstorff and Ku.
Zarnstorff reported that he and Long-Poe Ku had succeeded in completing runs with the merged optimizer on the HECATE and SEABORG computers. However, these were still shakedown runs, not production runs.
Please forward any comments or corrections to reiersen@pppl.gov
(last edited on 10/04/2001 10:35 AM )