I apologize for scooting through this thread, but Twila and I will be driving home on Friday and things have busified. It would be fun (for me, maybe not for you :p ) to talk about other kinds of vacuum pumps, but I'll jump ahead to leak detection.
With luck, our vacuum chamber pumped quickly down to high vacuum. But luck? - hah! Suppose we can't pump to a low enough vacuum, or suppose the vacuum rises too quickly when we close the pump valve? We have a leak, and the fun begins.
Remember that we aren't talking about big globs of air in the chamber, we're talking about scattered molecules. It doesn't take many new ones to affect the pressure. Gas molecules are very small and there will always be a few getting in somehow, but not all vacuum leaks are leaks.
I'm at a machine that is not reaching the required vacuum. Once I've verified that the plug is in the wall, the pumps are running, and the ON/OFF switch is in the ON position, three possibilities pop into my cranial vacuum:
1) There is a leak. Gas molecules are getting into the chamber from the outside.
2) Outgassing is occurring. Some things, such as water vapor, are extremely difficult to pump. Or there might be a contaminant in the chamber that is releasing molecules.
3) There is a virtual leak. There might be, for example, a tiny crack somewhere inside the chamber. At atmospheric pressure, air eventually fills the crack. When the chamber is pumped, air leaks from that crack at a slow rate that mimics a leak from the outside.
There are ways to diagnose and deal with #2 & #3. For example, pump long enough and both problems will resolve. We'll assume first that there is a real leak.
It's interesting to have a large chamber with portholes and fittings and o-rings and welds, and to know that somewhere there is something wrong that could be invisible to the naked eye. How will we go about it? The first step is to check obvious things. Bolts on any portholes should be properly tightened so the o-rings between them and the chamber are not distorted. The action of the mechanical pump - which will become the backing pump - should be checked by seeing if the pumping time to crossover (from mechanical pump to high vacuum pump) is typical. Seals that open and close, such as the bottom of a bell jar, should be checked for the proper amount of vacuum grease, if appropriate. O-rings should not be twisted. If all of those things are okay, we bring out the big gun - a leak detector.
A leak detector is a self-contained device that contains a vacuum pump and a simple mass spectrometer tuned to helium. The pump is connected to a fitting so it can sample gas that has been pumped from the chamber, which is then sent to the mass spectrometer. If any helium is detected in that gas, an alarm sounds.
Once the leak detector is connected and the chamber is pumped, we take a wand and spray helium on the outside of the chamber. Helium is an ideal gas to use because its molecules are very small and will slip through any openings big enough to be considered a leak; and because it won't blow up like the Hindenburg if something goes haywire. If we spray helium where the leak is, it will enter the chamber and then the leak detector, which will alarm and let us know we've found the leak.
One problem is that if too much helium is sprayed, the leak detector could continue to wail for a long while while we aren't really sure when the helium got into the chamber. (I really wanted to write "while while".) So filling the room with helium wouldn't work, although it would be a lot of fun talking.
I guess I have to aside this. If you inhale helium and then talk, your voice will sound like Daisy Duck's. Try it with a big helium balloon sometime. There isn't any danger because helium is non-toxic. If your lungs are full of helium, though, you aren't getting any oxygen, so don't get so carried away with duck talk that you don't take a real breath now and then.
Finding a leak with a leak detector requires releasing a very small amount of helium at the exact location of the leak, which is where the art comes in. I once worked for a company that manufactured custom vacuum equipment, and I had to find a tiny leak in a chamber that was the size of a Hummer. MMMMMMMMM, that was a lot of fun!
4 comments:
Way back when B.T.(before Timex) I worked at Ford Aerospace building satelites. We had a REALLY BIG vacuum chamber there. Big enough that they could set up the satelites in their configuration that they would be operated in space, solar panels and antenna booms extended. I would guess 100 feet in diameter and 50 feet tall. When we worked in it we had to take a key out of the interlock box so they could not pump it down with us inside. One side of the chamber was lined with lights to simulate the sun, the other side was refrigerated to simulate the cold of space. They could power it up and talk to the satelite over the radio thru the quartz windows. All the pumps were in the basement, if I remember right there were 12 diffusion pumps the size of hot tubs down there. Glad I never had to find a leak.
Cool!
I once had a friend who was doing some testing on a space-bound instrument. He put it in a small vacuum tank, started pumping and the tank just wouldn't pump down. When he opened the tank to track down the problem, it was all too obvious. A mouse had set up housekeeping in the tank and had exploded all over the inside of the tank and his instrument! Like Durf said, it's hard to pump down a big bag of mostly water. It was quite a mess to clean up.
I know what he said when he opened the tank - "Rats!"
Post a Comment