Cryogenic Valve Design

[g1a1]

I am developing a cryogenic Ball valve and have had limited success in achieving Zero leakage @ Temp. Does anybody know what the static sealing suface finish sould be? My current research has indicated 4-8 Ra.

Thanks,
g1a1

 

[jonralph]

you do have the d/s side of the ball drilled don't ya?

 

[g1a1]

No, the down stream side is not drilled. We have alternative means to relieve cavity pressure. What about surface sealing finish ??

 

[msmedesign]

there are many parameters that effects your sealing other than surface surface roughness such as form tolerances and also strrength of the material that you use. I f you give us more info about your design and how you have it ttested and failure .... I might be able to help you out.

 

[g1a1]

The material is a custom PTFE, engineering in our Lab. Mechanical Properties match that of Std. Teflon. The O.D. of the Seat is the Primary Datum. All other diameters must be with 0.005" true position. Valve is purged with helium prior to cool down (since all oxygen must be removed ). During cool down there is 50 psi helium within valve cavity (Closed position). Once -320F is obtain for 1 hour, valve is cycled 5 times and Helium is applied to the closed valve @ 50 psi, 100psi etc. Failure occurs if leakage greater than 0.08 scfh or blow thru results.

 

[g1a1]

Almost forgot msmedesign, thanks for your assistance!!

 

[Fog1]

Ah what the heck Serck Audco has been bought striped and gone to America so I don’t feel I’m spilling the beans.

Surface finish does play a roll in sealing certainly when your hovering around –198 C but the seat design is the key.

You do not indicate if you are looking at a so called floating ball, the old Jamesbury Serck design, or a trunnion. I guess your looking a floating as they are easier to get to seal.

The surface finish on the ball is around Ra 8 (ish) if your measuring stick is in units of inches. The seat can be as high as Ra 16 fresh off the machine or out of the mould but carries a specification of how any remaining machining marks should be arranged. Try spiral. The Ra 16 wont be Ra 16 for long as the high spots will be smoothed out by the first operation of the ball.

The seat, PTFE virgin or glass filled doesn’t matter, needs to have a certain volumetric compression imposed on it by the ball, somewhere around 11% would be a good start although this will depend very much on your seat shape. I’ve already provided the clue to that one.

The stuffing box or stem seal will be the next problem and that’s all to do with your bonnet extension. That bits another story.

Now if you’re dealing with super fluids then you have a different problem all together. A lot of people give up with ball valves at this point and go to a globe valve because they can get much higher seat loads.

I’m puzzled, why are you re inventing the wheel but good luck anyway.

Fog Jones

 

[iainuts]

Seats:
- Seating finish you're quoting sounds fine to me.
- Seating contact stress must be significantly higher than the delta P across the seat. For Teflon, that number is around 1.5 to 2 times the dP. Make sure contact stress isn't higher than compressive yield strength of material.
- Seat width must be at least 0.030". Seats thinner than this are likely to leak due to what I've heard described as "moats" but this difficult to explain here. I'd suggest a slightly wider seat, around 0.050" (ie: the outer radius of the seat MINUS the inner radius of the seat should be about 0.050".