We (AG and DSD) had a meeting with Jim Parkinson of the Accelerator group at the TJNAL to discuss the vacuum system. Our discussion started with the target and worked its way downstream.
Copper gaskets throughout the beamline are o.k.
Jim says we do need a bellows between the target and Q1.
Our design for a flange which is contained inside the pipe on the upstream end of the pipe which goes through Q1 (see vacuum system drawings) thus enabling the pipe to be slid through Q1 in the downstream direction is o.k.
We will go with conflat flanges everywhere (two knife edges with a gasket in between).
All pipes, flanges should be stainless steel. All their pipes through their quads and dipoles are 316L stainless steel.
Two methods of welding flanges to bellows:
Method 1:
Match wall thickness with overlay.
Method 2:
Butt joint for 100% penetration to avoid virtual leaks...
They use 0.065 inch wall thickness pipes for pipes up to 4 inches.
Jim says we need bellows between Q1 and Q2.
We should use stainless steel bolts with silicon bronze nuts.
They have 5/16 silicon bronze nuts.
For blind tapped holes, get silicon bronze bolts.
Viton o.k. for vacuum.
We discussed the virtual leak problem at the Moller/accelerator vacuum interface. The two solutions were to have the kapton (or whatever) held up from the metal by the thickness of the O-rings, or to mill out the downstream edge of the piece on the upstream side of the window everywhere except near bolt holes and O-ring groove. This would provide a clear space to pump.
Jim says milling is best and we want metal contact on Kapton rather than just O-ring.
Discussed using one square O-ring. Jim says go with two circular O-rings.
As for virtual leaks due to bolts, can buy bolts with thread relief or bolts with holes drilled down the middle.
We discussed the issue of providing vacuum inside the dipole using the pole faces as a wall of the vacuum box. He suggested we attach the plates which will cover the coils with high temperature RTV. He will, however, think about some other way.
Jim o.k.ed the technique of using an O-ring to make a corner vacuum seal for a Moller vacuum/atmosphere interface (see below):
Respectfully submitted, 5/28, DSD.
Kirsten DeNoyelles
Mon Jun 10 15:29:09 EDT 1996
Next: About this document