O-Ring (or alternative) sealing solution for a pressurized casing 1

[Skatemav86]

Hi all,

I have a problem when trying to seal the pressurized casing in the image below.
Initially, I have thought about 2 axial o-rings and 2 longitudinal o-rings (as shown), but I don't know how to avoid possible leakage at the o-ring interfaces.
The fluid is cyclopentane at 260 °C.
Do you know how to overcome this issue?

Thank you in advance.

Fede

 

[zeusfaber]

That axial seal looks like quite a difficult thing to clamp evenly, even without the problem of leakage round the ends.

Can you tell us why the casing needs to be split? Would the application allow for a one-piece outer casing to provide the pressure boundary, with a slip-in split carrier to manage whatever it is that needs to be got in or out?

A.

 

[Skatemav86]

Zeusfaber thank you for you reply.
The casing in the middle has to be split in order to carry stator stages.
A split-in split carrier is an option that complicates the assembly phase too much, but it might be the only option in this case...

 

[Prometheus21]

"The fluid is cyclopentane at 260 °C." - Then O-rings are out of the question. Cyclopentane is a hydrocarbon that causes cracking and embrittlement, leading to leakage. FKM (Viton) is usually used in these applications, but have a maximum operating temperature of 200-205°C, so you are out of luck in this regard I'm afraid..

 

[TugboatEng]

There is a product called Copaltite that may work for this application. Use it instead of o-rings.

 

[Skatemav86]

@Prometheus21: we usually rely on Perfluoroelastomer(FFKM) o-rings for these applications.
@TugboatEng: I will check for more info on Copaltite, thank you.

 

[Prometheus21]

Hello again,

Even with FFKM you are usually restricted at 240 °C no?, unless you have a specialized compound(?).

I have to agree with TugboatEng on this one, I would look at Copaltite.

Looking at it now this reminds me of certain subsea connectors, they use silver coating on titanium spheres to create micro metal to metal seals that seals for pressures up to 2000 Bar. Quite neat, but expensive.

 

[EdStainless]

Crushing metal seals were my first thought. I have seen this done with both high purity Cu and Ag.
This really looks like a case for gaskets and not o-rings.
But getting a suitable material will be tough.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Skatemav86]

Prometheus21, according to this website FFKM reaches almost 320 °C.
Link
Thank you for your suggestions but I need to find a cheap solution (if really exists)

 

[TugboatEng]

Those maximum temperature ratings are usually for an inert atmosphere. You will likely have to derate significantly considering the chemical you are using.

I suggested Copaltite because it is the least expensive solution. Cost is around $200 per pint and requires only flat surfaces, no o-ring grooves. If you design for it, Copaltite can be pumped into grooves after assembly.

I don't normally recommend Copaltite because it is quite hazardous in it's uncured state, but so are the chemicals you're already working with.

 

[LittleInch]

What's the design pressure?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[MintJulep]

Hylomar has a range of non-curing sealants. I don't see any in their public-facing info that goes to 260°C, but it might be worth a call to ask about their confidence at higher temperatures, or if they have any secret sauce that's not advertised.

Molykote 112 is advertised as up to 232°C. Probably worth a call.

 

[Prometheus21]

Skatemav86: It seems like Parker and Trelleborg have two different temperature ratings for FFKM. 240 for Trelleborg, 320 for Parker. I can only assume it's due to differences in the compound they use. Europe has started to regulate and limit a lot of the basic compound ingredients that are still freely available in other continents...

You need to find a cheap solution; I assume redesign is out of the question(?). More details like design pressure (Like LittleInch asks) would be helpful.

Technical curiosity; what is the usage of cyclopentane at 260 °C? My experience has ever only been with cylinders and industrial gas mixtures containing it.

 

[Skatemav86]

@TugboatEng: thank you for your info on FFKM temperature resistance, I will investigate it.
@MintJulep: thank you
@LittleInch: design pressure is 32 barg
@Prometheus21: It is for a ORC
Link

 

[LittleInch]

The issue I see is that your split casing as shown has no real ability to seal itself.

If you added a set of bolts along the flat edge then maybe, but otherwise this would either bend due to pressure or simply move apart within the bolt holes of the two flange on the end.

Your axial seal (it's not an O ring) won't be able to energise or seal without a sealing force clamping the two halves together.
This whole thing looks unfeasible to me, especially at 32 bar and 260C. Far too much room for movement and leakage.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[TugboatEng]

If you do drill those flat edges for bolts do include two body fitted bolts or taper pins to ensure correct alignment of the round flanges after assembly.

 

[Skatemav86]

@LittleInch @TugboatEng You are right, I have simplified the representation too much.
The real configuration includes a bolted connection between the the halves:

 

[EdStainless]

And the longitudinal 'flanges' need to be heavier than you round flanges since they are taking more load (bending).
You might even consider stiffener ribs 90deg from your long flanges.
Even though the pressure isn't high you really need to prevent any flexing or it will never seal.
When you build it you will put a space in to represent the seal, slip the two halves together over the fixed pins, bolt up to full torque (using good bolting method), and then finish machine the end flanges.
When you install it, it goes the same way, apply gasket/sealant, place halves together and bolt up, then fasten ends.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[XL83NL]

There are various (I think at least 2) FFKM grades for use above 300 deg C.
We use one grade occasionally that goes up to 327 deg C.

Huub
- You never get what you expect, you only get what you inspect.