Experimental insights: Leak detection

Welcome to the latest installment of Experimental Insights! Your semi-regular guide to success in the Low Temperature Physics laboratory. Today we will look at leak detection.

Most low temperature experiments still involve the crude process of immersing dunking the experimental insert into a large vessel of liquid Helium, in order to get the majority of the metal parts to roughly 4.2K, before proceeding to lower temperatures. Newer types of fridge don’t suffer from this problem, but you do need to be rich to own one at the moment 🙂

Liquid helium can be pretty tricksy – unless your experiment is well sealed inside the Inner Vacuum Can (IVC), the liquid will find a way in through the seal. If you can’t remember what the IVC looks like from previous posts, here’s a picture:

leakdetector2

The seal is made at the join between the brass and the steel, with the bolts holding the two parts together.

Once the IVC is inserted into the cryogenic bath, the low pressure inside will lower the boiling point of any incoming Helium to below the Lambda point, where it becomes superfluid. Having zero viscosity, the helium then finds it even easier to get through a small hole or leak, and creeps all over your experiment. A helium leak into the IVC can be pretty disasterous – at the best you’ll lose your vacuum and the fridge won’t cool below 4.2K, at worst your IVC may fill up with liquid Helium, overpressurize and potentially form a bomb. So it’s a good idea to check that you have made a good seal at room temperature before you start dunking.

How do you make a seal which does not freeze up at Liquid Helium temperatures? The trick is to use the extremely malleable metal Indium – which remains slightly squishy at low temperatures. Indium is seriously fun to play with, it’s a bit like metallic plasticine.

To make the seal a ring of Indium wire is put around the vacuum can and squashed down firmly as the bolts are tightened. Once the Indium seal is nicely compressed, the leak detector is connected up to the experimental insert, through a pipe which links it to the IVC. So it looks like this:

leakdetector1

The leak detector is a piece of vacuum equipment, which ‘sniffs’ the incoming gas as the IVC is evacuated. The detector contains a mass spectrometer, which is sensitive to Helium. Any leak into the IVC from the surrounding air will be detected. The next step is to spray some helium gas around the seal, and other areas where you may suspect a leak. If the helium makes it through the seal and into the IVC, it will be detected, and the machine will beep at you. It it stays quiet, all is well. Here is a picture of me doing this:

leakdetector3

The machine measures the leak rate in mBarl/s. Between 1e-9 and 1e-10 is an acceptable leak rate, and shows that the seal is working properly. The photograph below was taken as the machine was pumping down, which is why the leak rate hasn’t yet reached the desired value:

leakdetector4

Of course this doesn’t guarantee that the seal will work at low temperatures, but it’s a start. A leak which only shows up at liquid helium temperatures and then disappears again once the insert is warmed is known as a superleak. It’s almost impossible to diagnose, and is one of the many things that will make low temperture physicists swear. A lot.

Advertisements

2 thoughts on “Experimental insights: Leak detection

  1. Ahh, the leak detector. I remember one occasion like it was yesterday. I was inside the shielded room, leak checking my > 50 epoxy joint experimental cell, using the “buzz” noise on the leak detector (outside the shielded room) as an indicator. For those who don’t know, you can set most leak detectors to emit a buzz or beep that rises in pitch proportional to the detected leak rate. I was very tense, and in the middle of my initial checks, the buzz pitch skyrocketed. I was just about to get upset when my thesis advisor stuck his head in the door and said “Kidding!”. He’d been walking by, saw me muttering to myself, and had decided to mess with my head by dialing around the zero offset of the detector. Somehow I didn’t kill him. Good times.

  2. Uncle Al says:

    Physicists should avoid any problem dealing with stuff (as opposed to things). Add a supported gasket compression ring of zirconium tungstate. As everything else cools and shrinks zirconium tungstate cools and expands, right down to 0 K. Your gasket seal will stay snugged.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s