Overview of the magnet winding system for the alpha experiment.
The HelCat device was originally called alpha for Auburn Large-area Plasma Helicon Array, and was constructed at Auburn University. The original machine include seven (7) helicon sources that worked together to make a large area plasma. Research on the large-area source has been suspended for now. Here is some background on its construction.
Construction of alpha was completed in 1981 at Auburn University by a team of graduate and undergraduate students. The device was moved to the University of New Mexico during the summer of 2005. The pictures below show the winding process. We are constructing 13 magnets for the facility. Each magnet consists of 10 pancake layers of 14 turns each. The total length of copper in each magnet is 360 meters (1000 ft) and it weight about 400 kg (800 lb). The total length of copper is over 3 miles!
Funding for alpha is provided by the U.S. Department of Energy.
The magnets were constructed in house. A single 74 m strand of copper tubing was used for each pair of layers. Winding was done onto a cylindrical template affixed to a rotating platform (see Figure 2). A crossover bend was made at the midpoint of the tubing using a jog press, a specially developed tool based on a design by Fred Skiff at the University of Iowa. The strand was clamped to the template at the midpoint and the “excess” half of the strand was placed on the rotating platform. The section of bare copper to be wound was placed on a second rotating platform. The bottom layer was wound outward from the kink at the midpoint, then the top layer was wound from the remaining copper. Turn to turn insulation is provided by fiberglass tape which was wrapped onto the bare copper during the winding process. The entire process was mechanized so that a one person could wind the layer alone. During the winding, bare copper was pulled through a straightener (courtesy of the HSX stellarator group) to remove twists in strand; the straigtener also provided tension for winding onto the template. The copper then passed through an in-house developed planetary tape winding machine that automatically wrapped the fiberglass tape around the copper. The speed of the winding platform was adjusted so that the tape was wrapped in a half-lap configuration. During the process the operator used both rubber mallets and a “twisting tool” to align the copper as it wound onto the template and insure that the copper lay flat. For the 13th magnet winding of each layer took about 3 hours.
All five double-layers were wound on to the same template. Successive double layers were wound with opposite helicity to minimize the field error due to the crossovers. Once all layers were wound, the completed magnet was removed from the template. Electrical and cooling connections were brazed on. The outside ends of the tubing are electrically connected in series, but the water connections are in parallel so that each double layer has a separate cooling connection. The magnet support base and side “ears” were attached at this point.
Tools of the trade: The automatic tape winding machine for wrapping fiberglass tape around the bare copper. The straightener for taking kinks out of the copper. The jog-press for making cross-over bends.
The entire magnet was then wrapped in a layer of fiberglass tape, and then wrapped in plastic “release” tape. It was then placed on a form for epoxying. During the epoxy process the magnet was put under vacuum. The liquid epoxy was poured into the form and allowed to settle, then the magnet was then brought up to atmospheric pressure. This process insures that the epoxy is forced into the voids in the fiberglass between coil turns. The magnet was heated to 80°C speed curing, which typically took two days. Once the magnet was cured, epoxy flashing was chipped away, facilitated by the release tape.
Potting the magnets in epoxy. The vacuum box. The fiberglass
wrapped copper is placed in a form and the epoxy is poured under vacuum.
Completed epoxy pour. Failure! The epoxy reaction is exothermic, so if
it gets too hot the reaction runs away, and you get a pot full of volcanic
glass.
Installation of the vacuum chamber. Set-up for
magnetic field calibration.
 Return
to Plasma & Fusion Sciences main page |
Return
to Plasma Sciences Laboratory |
