Vacuum Box Engineering Meeting -- March 16, 2000


Present: Mark Ito, Dave Kashy, Arne Freyberg, Pete Hemler, Mahbub Khandakar, Dan Dale

Wish to be able to remove vacuum box without taking apart magnet.

Need horizontal and vertical wires on the target ladder to get absolute target position.

Want blank the same size as physics target to check backgrounds. So, we'll have 3 physics targets, one blank, one thin target, plus fiducial wires for positioning.

Arne suggests attaching target mover to front plate of vacuum box. Gears and shafts would give x y motion. Some discussion of backlash errors, but Arne says this is no problem.

They can put target arbitarily close to pole of magnet. The further donstream, the better the geometrical acceptance. However, we lose Bdl and Bdl is more sensitive to target position if it is too close. 10 cm upstream of the pole (but no more) seems optimal.

Front plate will have viton O-ring.

Some discussion of making box out of aluminum. If so, it would have to be thinner at the downstream edge of the pole and supported off the magnet. The geometry is quite tight, but pins (or ``steps'') mounted on the box could be made to mate with holes in the pole of the magnet to support the box.

Big deal to have target move in y direction -- about an order of magnitude more complicated (but still doable). Dave Kashy suggests machining targets down to uniform rho. Arne suggests mapping out relative thickness, and putting region of constant thicknesses on ladder at beam level so only one degree of movement is needed. Discuss with Rory.

Decided to put plumbing side of magnet downstream.

I presented my idea of design of vacuum box. There were a few suggestions, which are reflected in the attached sketches.

Arne requested that the part downstream of the coils be removable-- ``removable horn'' with O-ring seal.

Suggestion was to screw NMR probe permanently to the lower inside wall of the vacuum box at the center. We would then just need one electrical feedthru.

Still need to understand how Superharp fits into this.

-DSD