A Stainless Steel Baby Bottle

Let's say you have some liquid that you want to contain without leaks, say, milk for a baby. What do you do?

Well, you put it in something like a baby bottle, the components of which are shown here:

i-61c47c7ed2e2aab1659937dcdef6753a-sm_baby_bottle.jpg

You have a hard plastic bottle, a soft silicone nipple, and a hard plastic ring that screws onto the bottle. When you put it together and screw the cap down tight, it compresses the silicone between the two plastic bits, squeezing it into the small gaps, and plugging any leaks. Done properly, this will ensure that milk doesn't leak out of the bottle except through the whole in the nipple.

Now, imagine that you want to contain some extremely dilute gas, in an ultra-high-vacuum chamber. What do you do there?

Well, the principle is exactly the same as with the baby bottle: you take two hard pieces, and clamp a soft gasket between them. The only difference between a UHV chamber and a baby bottle is the materials. When you're working with vacuum hardware, the "bottle" is stainless steel, and the soft gasket is copper:

i-e48cdec1749cf25eb68bc62e80f02bb2-flange_w_gasket.jpg

This picture shows a partially assembled vacuum system, with a fresh copper gasket on a flange, waiting for the top piece to be put on. Below that, you can see a bit of the copper gasket on the flange below-- the top flange in this case is a reducing flange to connect a rather large vacuum pump to a smaller tube coming off the main chamber. You can see the copper gasket pinched between the reducing flange and the main chamber flange, around the inside of the tube.

When the two flanges are put in place, the gasket is squeezed between two knife edges, one on each flange. They're hard to get a good picture of, but you can see their effect by comparing used and new gaskets:

i-1811182036c77e4adc67e744a43f9bc4-sm_gaskets.jpg

The gasket on the right is a fresh one, prior to being put on the flange. The gasket on the left is the used one that I took out when I took the chamber apart. You can see a small line going around the middle of the copper ring, which is the impression made by the knife edge when the copper was squeezed down.

How does this thing go together? Well, in the picture of the flange, you can see a bunch of holes around the outer edge. Those are through holes for 1/4"-28 bolts. You put a bolt through each hole, with a nut on the top, and tighten them down to crush the copper gasket into place.

Here's a picture of a just-assembled flange, showing the bolts, nuts, and a tiny gap between the stainless flanges with just a hint of copper showing:

i-aaeec1c8d2eaf087a831be30fabc6059-loose.jpg

Here's the same flange, after all the bolts have been tightened down:

i-c3047e7ee38447ba3e455256bf02ac86-tight.jpg

You can see that the gap has gone away, and no more copper is visible. That's the sign of a properly sealed system.

If this sounds like kind of a pain in the ass, you're right. If you look at the picture of the flange above, you can see that there are 20 bolt holes on the flange, each of which must be cranked down hard enough to cut into the copper. Better yet, you need to make sure that the compression is uniform, which means tightening the bolts in a star pattern-- tighten one bolt, then the bolt on the opposite side, then a bolt 90 degrees from the initial bolt, then the bolt opposite that one, and so one. It's tons of fun. Better still, each bolt (on this particular flange, anyway) is coated with a grey anti-seize grease that is next to impossible to get off your hands.

Fun, fun, fun. And that's how I spent the day Saturday.

More like this

Over in the reader request thread, Richard asks for experimental details: I'd be interested in (probably a series) of posts on how people practically actually do cold atoms experiments because I don't really know. I needed to take some new publicity photos of the lab anyway, so this is a good…
I spent most of Saturday in the lab, swapping out a turbopump that was starting to die. How could I tell? Well, for one thing, it made an awful noise, even more than usual for a noisy pump. But after it was stopped and unmounted from the chamber, there was a simple test: comparing the rotation to a…
We've continued plugging away at the optical excitation experiment discussed in the Week in the Lab series last year, and have finally managed to get a decent metastable signal out of the thing. The signals are at pressures that are considerably higher than I would like (and quite a bit higher than…
The other day, I made a suggestion to one of my research students of an experiment to try. When I checked back a day later, she told me it hadn't worked, and I immediately realized that what I had told her to do was very stupid. As penance, then, I'll explain the underlying physics, which…

We went up to Mount Palomar this summer. The telescopes were very interesting, but the engineering was fascinating. The mechanism for moving the biggest telescope in all three dimensions rests on a millimeter-thin layer of grease. The volunteer giving the talk said that getting the grease off of his hands was harder than greasing the entire assembly.

So, how do you get the grease off?

By Lauren Uroff (not verified) on 04 Jan 2010 #permalink

I use this nasty orange-scented shop cleaner that I got at Lowe's. It gets most of the grease off, but at the price of having my hands stink of fake orange scent for the rest of the day.

You don't have to torque the bolts down twice? I'd have thought you'd have to do an initial low torque tightening and follow with the final torque (and a third check since as you tighten down, the system is still slightly dynamic).

And when you've turned a mole of bolts, you're ready to receive your PhD.... That was the running joke in a friend's electron diffraction lab when we were working on our degrees.

By Grant Goodyear (not verified) on 04 Jan 2010 #permalink

I do two passes of bolt-tightening in the star pattern, and then another one or two just going around the flange in order.

The right way to do this is using a torque wrench, but I don't have one, and am too lazy to dig through the shop looking for one. I just tighten them all until the flanges are right together, and call it done.

Ah, bolt tightening patterns--one of the areas in which lab superstitions get developed and passed down to new group members, probably derived from whether what someone did a long time ago randomly worked or didn't.

Bolt patterns also give you a chance to demonstrate a practical application of group theory--I forget the precise statement of the theorems, but if there are n bolts all total, and you move along the bolt circle m bolts with each (partial) tightening, then if n and m are relatively prime, you know that you'll tighten all of the bolts before coming back to the first one you tightened, thus ensuring that you tighten all the bolts evenly. So with 12 bolts, go along tightening every 5th bolt.

UHV systems use these metallic sealing systems (the Kleenex-like trade name is Conflat) for two main reasons:

- Materials exposed to vacuum have very low vapor pressure at room temperature

- Materials exposed to vacuum are relatively easy to clean (ie, to remove high vapor pressure contaminants like grease)

The joke is that UHV guys hate using their equipment for experiments; their precious chamber will get dirty.

I've had poor luck with that particular (nameless here) brand of baby bottles. The "o-ring" doesn't seal well enough, leading to leaks one doesn't need a helium detector to find. ;)

We do see occasional leaks with the baby bottles, but it's not too much of a problem when SteelyKid is her usual hungry self. The milk doesn't stick around long enough for much of it to leak.

Ah! Orange goop cleaner. That's the smell of macho scientists and technologists the world over. We have a can under the sink, as well as Boraxo (for the sandpaper effect) and Lava (when only pumice will do.) If one of those three doesn't work, we dig out the acetone.

I wonder if Steelykid's brain associates you with oranges. My kid has me pegged as smelling of menthol (bad back) and to me, my own mom smelled of Paquin's hand lotion. My dad always smelled of that stuff used to duplicate blueprints.

M. Proust lives!

By Lauren Uroff (not verified) on 04 Jan 2010 #permalink

Perhaps the finest anti-seize, anti-gall treatment between ambient and 1000 C is (off-brand - Phillips is pricey) milk of magnesia. We had a terrible time with 316SS bolts soaked at 850 C overnight. Blue Goop was silly, Silver Goop failed, an exotic nickel-based grease was inadequate, h-BN didn't like the temp. Small brush, milk of magnesia; coat, dry, coat, dry; slightly bake, cool, assemble. One imagines a micronized thickened liquid suspension of dead-burned MgO would be even better.

The solution to leaks on that brand appears to be to wet the nipple on the surface facing the bottle before screwing together. We had the same issue, but this has proven to be a reliable fix.

[Why, Chad? What property is altered? :)]

Next time you see a pipe fitter give him a knowing wink.

On second thought, winking at pipe fitters might be a bad idea...

Aside from bottle jokes and mechanical engineering reminiscences, I want to say that I really appreciate the hardware/equipment posts you've been doing, both here and on youtube. Please keep it up!

(now why do I feel the need to visit the scientific equipment and materials surplus/salvage store?)

By featheredfrog (not verified) on 05 Jan 2010 #permalink

let me guess--moly disulfide. i always use gloves when i am messing with bolts on my chamber. i can throw them away. plus, what with the arsenic in the chamber, it is nice to have the gloves on anyway...

You need to make friends with a mechanical engineer.

He/she will design a special tool that will turn all the nuts at the exact same time with some fairly straightforward gearing.

Be sure to post a video of the result.

This comment is about the n bolts and skipping every m while you tighten. Its an unusual case to have the n number of bolts be a prime number. 6, 12 8 are the most common. For higher numbers just pick a number smaller than half the total number of bolts probably (n/2-1) and make sure that n % (n/2-1) is non zero. % is the mod operation. As a matter of fact 6 is the only number where

By Don Robison (not verified) on 27 Jan 2010 #permalink