I work at the National Superconducting Cyclotron Lab. at Michigan State University. The word vacuum is a relative term. For vacuum systems down to pressures of 10 mtorr, GMAW is completely acceptable. Down to this range it's also
okay to use steel for most vacuum vessels. Stainless only comes into play below this 10 mtorr pressure. 10-6 and
below usually requires stainless steel. There less out gassing. At pressures of 10-9, you can no longer use rubber
seals, and no virtual leaks are acceptable, so special designs are required GTAW or Electron beam. These pressures also require special pumps. You have to use all metal seals. Where I work, for transferring liquid helium at 4.7 K steel is fine. I welded 38, 3/4 inch steel vacuum vessels with GMAW and I only developed one leak (external) which was easily fixable. The old method here at the lab was welding these containers with GTAW root and 7018 all other passes. The welding time with this old method was about 12 hours per box. GTAW took one pass with .45 wire at high amperage 400-600 amps, time required per box was 2.2 hours. I was fast at rolling them.(on the floor) I don't remember the exact amperage. These boxes were about 5 feet long and 3 feet wide and about 2.5 feet tall. To this date they are all in operation with no leaks. The leak that I did develope was at an stop and start overlap. This area always takes precaution. By the way, I had to write a proposal to get them to accept this GMAW procedure.
I hope this is of help to you.
Dan Pendell, Skilled Trades Welder, Local 999 AFL-CIO