Bluedeath, I don't think that's possible - the big commercial simulators from the likes of CAE use active management systems. That isn't to say that sealing the vacuum plenum isn't a good idea - the better seal you have the easier the PID controller will be able to manage the mirror shape.

g.

Some seal for sure would allow to use a smaller pump so less noise. But is of course just a theory since my last studies on gas dynamics are back 18 years when i was in high school physics classes. If the the problem is just the noise you can try some kind of active noise reduction (sampling the vacuum noise and playng it countephased like the real acive noise reducer, but again is a complex issue to address)

Bluedeath, you're right that sealing the whole assembly would allow the use of a smaller pump. Our 60° prototype was poorly sealed, and worked well with the first valve box we built. When we built the full-size version, we sealed all the surfaces and joints, and ended up having to drill about 2 square inches of cheater holes in the valve box.

Simply drawing down and leaving it is not a real option, as the mylar can have a tendency to 'creep' when it's that close to it's stress limit. You can't even apply a constant suction and have it work. Basically, it's very close to the point where stretching further requires -less- force, so as it pulls down, it just wants to keep going. It's an unstable equilibrium, much like a ball at the top of a hill. Active management to keep the mirror in the right spot is a necessity.

Vacuum noise is something each user would have to deal with on their own. Gene's plan seems to be to build a soundproof box for the vacuum and valve box; my plan is to put the vacuum on the other side of my office wall, in my garage.