High Vac Out gassing & leak checking

[Dennis P Brown]

A few minor notes on out gassing and leak checking. My accelerator tube struggled to get into the mid 10-^5 torr even after a mild bake out (turbo running at full speed.) Using a spray can of "dust off" and an ionization gauge, I easily detected a number of tiny, slow leaks - one was the main feed-throughs for the metal pins into the ion gauge itself! Not exactly where I'd normally look for a leak and would never use alcohol (lol) there to detect for leaks! Used a special high vac liquid sealant to stop the leaks on non-KF seal locations. Alcohol some times works by itself but this HV liquid sealant does wonders.
Strange issue with the out gassing (but not true for real leaks), while the system still struggled to get to the mid 10^-5 torr; the next day (after sitting under vac all night) the turbo got to this level in five min even in standby mode. Then bottom out at 2 *10^-5 torr. The following day after this, the turbo reached 1.5 * 10 ^-5 torr in standby mode and after an hour, just got into 10^-6 torr. Then the day after this (again, always under vac), the turbo reached 8 * 10^-6 torr while still in just standby mode! Out gassing takes time, even after a mild bake out ...epoxy doesn't hold up to a real bake out (that is, near/over 100 C.) Out gassing, unlike real leaks, always gets better lower) as the system remains under vacuum. Leaks are constant.

[Jim Kovalchick]

Dennis,
You need one of these babies. I recently rescued this from a scrap dumpster that was emptied just a couple days after I found it. I did a preliminary test with party balloon helium and it works like a champ at finding tiny leaks. Of course, not everyone will get this lucky, so thanks for sharing your dust can trick.

Jim Kovalchick

[prestonbarrows]

Even if you have a perfect vacuum chamber with no leaks, real or virtual, the first few days under vacuum will be much higher than the 'true' baseline if you don't bake out the chamber. Even with a huge turbo pump and immaculate chamber, you will begin to hit a wall around 1E-6 or 1E-7 Torr with pumping alone.

Basically, water (and lots of other stuff) from the atmosphere sticks to every surface before you start pumping. Water is especially problematic since it take a relatively large amount of energy to pull back it off the surface. In addition, once you reach molecular flow levels things stop behaving like a gas and start acting more like individual billiard balls. Each time a water atom scavenges enough energy from the tail of the thermal distribution to overcome the surface potential and fly off into the vacuum, it only has a small chance of reaching your pump and actually being removed. Most of the time, it will just stick to the wall again in some other random location.

If you have gotten all your leaks sorted out, baking out the chamber or just running a plasma discharge within it will drastically speed up your conditioning times. The choke point of pulling most contaminates out of the chamber is the energy required to rip them off the surfaces they are stuck to; dumping in energy through plasma interactions or simple heating moves the process along much faster.

[Jerry Biehler]

I dont think you have much of a leak, you just need to let it run and heat things up, you just have a ton of water vapor in the system.

How big of turbo?

[Dennis P Brown]

Its a rather small turbo but pumping on very little. Not too concerned since I'm about in the range I need to be even in stdby mode for the turbo. Still, glade I did a bake out on the connecting hardware; foolish to have done any on the accelerator system (paid for that one with a major epoxy failure on the target holder feed thru.) Agree, water vapor is an issue. Time will work. While I use a drier system for intake air, that is still rather "wet" compared to 10^-6 torr. So, time and careful bake out of select sections is the correct "cure". Thanks for all the input. Good point about how water moves around in a system. It is rather amazing the difference in difficulty to go from 1.1*10^-5 torr (under 5 min from atm with the mech pump then turbo) to even reach just over 5 *10 ^-6 torr (almost two hours ....)