Cryogenic Swivel Joint

[Chairs]

I'm looking into designing a swivel joint for cryogenic applications (33 K >). I'm assuming I'll use 316 SS for the body, PCTFE or something similar for the seals, and a low temp grease. Does anyone have any recommendation on what materials work best? Are there any special requirements for cryogenic fittings / Anything special that needs to be taken into consideration?

 

[Compositepro]

You say swivel joint as opposed to rotary joint. If your application only requires limited rotation of one or two revolutions each way, it is best to avoid seals altogether and use a spiral of tubing that can wind and un-wind.

 

[iainuts]

That's a very open ended question. Can you be more specific?
What type of fluid is the joint intended for?
What pressure?
What rotational speed?
What amount of leakage is acceptable?
What kind of lifetime are you trying to achieve?
Are the seals going to be at cryogenic temperature or can you get them away from that temperature?
Are there significant external radial and axial forces and have you considered how to support that load?
Is this for a product line you want to sell or is it strictly for internal use?

You can purchase seals directly from a manufacturer. The most common for this type of application are the U cup style and the face sealing style (ex: John Crane). Bearings/bushings for these are commonly plastic against metal but that wouldn't work well for high rotational speeds/loads.

 

[Chairs]

Compositepro can you elaborate on what you mean by using a spiral of tubing? And why is it better than a V-ring Style seal?

Media: Liquid Natural Gas
Pressure: I'm assuming less than 150 psi
This is not a rotary, it is not made for constant rotation.
Acceptable Leakage: This is one of my questions to anyone who is familiar with cryogenic regulations. Are there any criteria that need to be met?
Lifetime: About 100,000 cycles
I don't think I could get the seals far enough way to make a significant impact. The swivel should be as compact as possible.
This is for a product line.
I was thinking of a V-Ring seal packing, and sticking with 316 SS ball bearings. Are there any benefits to plastic besides price?

 

[Compositepro]

You can simply coil a few loops of tubing around the axis of rotation. With rotation the loops simply get a little tighter or looser. There are no dynamic seals required, just tubing.

 

[iainuts]

The use of swivel joints in cryogenic service is pretty rare from what I've seen. The ones I've come across are very expensive probably because so few are sold. The right way to do it is have a hose or other flexible joint such as suggested by Compositepro.

If you're going to make one, you should look at valve stem seals in cryogenic service. A V ring seal or U cup can be made to work but one problem is they shrink much more than the metal so don't try and seal on the OD of the seal, seal on the ID or face. The non-moving side of the seal should be clamped in some way to prevent it from shrinking and allowing leakage. Once leakage starts it often gets worse fairly quickly as the seal gets colder and shrinks more. Cheveron packings can also work well. There are a variety of decent seals available, just contact a seal manufacturer.

I don't think ball bearings are really needed as long as the seal can adjust for whatever tolerances there are and possibly support the loads. One problem with ball bearings is you can't use a lubricant. Any grease or oil is going to freeze solid. Note there are no ball bearings in cryogenic valve stems.

Other features you need to look for... make sure there is no way to trap liquid in an enclosed/sealed off area. Ball valves for example, can trap liquid inside the ball when closed so they have a vent hole to prevent dangerous overpressurization when they warm up. For flammable service, you should consider having some form of restriction to flow in case the primary seal is lost for whatever reason. One way is to have very close fitting metal parts that restrict flow. Also for flammable service, use SS rather than aluminum since a fire can melt aluminum.

In the US I would recommend using ASME code allowables for material strength and for general design guidance.