Modular coil winding scheme Layout and assembly Fogarty

Modular coil winding scheme • Layout and assembly– Fogarty / Williamson • Field errors – Brooks • Keystoning – Chrzanowski/Reiersen • Recommendation - Nelson NCSX WBS-1 Meeting July 30, 2003

Design Parameters Baseline design 3 -in-hand design Conventional pancake winding crossover at shell “jelly-roll” winding, 3 conductors wound in hand 2 x 14 2 x 3 x 9 . 665 x. 407 . 445 x. 400 Turn Insulation thickness . 054 in. . 035 in. No. of electrical turns 2 x 14 2 x 9 28 49 19. 7 k. A/cm^2 18. 8 k. A/cm^2 Winding scheme No. of cables Size of cables (after keystoning allowance) Current per turn Current density in Cu 2 2

Design Comparison Baseline design 3 -in-hand design Wind from middle Wind from end More layers Fewer layers Larger conductor - worse Smaller conductor - better adequate ok (~ 1%) ok? (~2%? ) No uncompensated leads Need extra restraint on uncompensated leads 19. 7 k. A/cm^2 (1. 3) 18. 8 k. A/cm^2 (1. 37) Current sharing No issue May be an issue for heating, but not field errors Current per turn 28 k. A 49 k. A 3 inch diameter 4. 5 in diameter Winding time Insulation application Keystoning Cooling Field errors (% flux in islands) Structure Current density at 1. 7 T (max field) Coax size in cryostat 3 3

Design Comparison – Cost W I N D I N G Baseline design (per J. Chrzanowski, April 03) 3 -in-hand design Winding facility $1810 k Same? Materials/supplies $1481 k Same? Area prep $146 k Same? Cryo/elect. Test setup $384 k Same? (higher current) $5 k Same? $1362 k (18400 hrs) 10% less? (- $136 k? ) n. a. (+ $50 k)? Potting cocoon and prep $639 k (8640 hrs) Same? VPI $597 k (8064 hrs) Same? $7147 k $7061 k $50 k cables $20 k for routing? ~ x 1. 5 (+$35 k)? $1289 k (to connect 30 4 modules? ) +10%? (+$129 k)? Receive conductor Winding Structure around leads, extra complexity subtotal: Other costs Local coax in cryostat Bus from D-site (Non-project cost) 4

Summary and Recommendation · · Either design is workable Baseline is more conventional, has lower current, lower field errors, smaller hole in shell for leads, smaller coax feed 3 -in-hand design is easier to wind, should have less keystoning and lower current density, but field errors may still need investigation Project cost should be less for 3 -in-hand, but program cost may be higher since we will need twice as much connecting bus from D-site Recommendation: · Use 3 -in-hand design for PDR pending confirmation of acceptable field errors for actual lead/crossover geometry · Look at multi-cable conductor again (is it really that bad? ) 5 5