cryogenic valve seat craking

[jabonet]

we have a problem in our plant, with the valve seats of valves, after a lot of checking seems that at least one of the causes is cryogenic thermal shock to the seat material.

kel-F, PTFE, or anything else just can´t keep it up, specially on the weekends the valve gets to ambient temperature or even more as it gets warm with the sun at maybe 50 C on the summer, and then the valve gets open all the way, in a couple of minutes (I have to measure this) it get down to -200 C (its used in both oxygen and nitrogen)

the original OEM part costs around 1500 €, we have found a non original for 500€ that lasts around the same time (not much).

any ideas of the cause, and posible solutions, or changes in the material?

Thanks.
jabonet

 

[EdStainless]

What is the current material?
These valves are supposed to be designed so that the seal remains at cryo temps all of the time. At least the ones that I have used.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[iainuts]

Actually, it's not unusual for a cryogenic valve to have to cool down in 30 seconds and not have any problem with cracking seats. Valve seats are typically Neoflon (was Kel-F before product was sold to Daikin) or Teflon of some sort. Virgin Teflon is rarely used, but glass filled isn't uncommon. Neoflon is the better material. If the seat is cracking, typical problems are how it is constrained and cross sections that are too thick.

Most filled and all unfilled Teflons shrink more than metals, so if they are constrained in such a way that they can't move, they may crack (not always true, I've seen small valve seats that are fully constrained and don't crack).

If the cross section is very thick, there could also be some thermal gradients within the material that cause differential contraction and that might cause cracking.

Please clarify how large a seat this is, how it's constrained (photos or drawings help), how thick it is, and any other details.

 

[bogiedreamer]

PEEK is a good material with thermal coefficient that is closer to metals. However, it is more expensive.

 

[Compositepro]

The CTE of PEEK is similar to most polymers (greater than metal).

 

[dfly9891]

Hi,

I've got an applications where I need an insulating washer/disk in order to insulate stainless from stainless at 50 degrees Kelvin. I need help in selecting a material to do the job. I.e. low thermal conductivity, durable enough at 50 K to take a compressive static load of approx 40 lbf. Performance, rather than cost, is is the most important characteristic in this case. Any Ideas??

Parameters: OD of washer ~2" Thickness ~1/4"(I can vary on the thickness if necessary but not on the OD). OD of footprint of 40lb load (compression spring) ~1.5".




Thanks in Advance.

 

[iainuts]

Micarta is commonly used for inner vessel supports on cryogenic tanks. Also, crygenic G10 is another excellent choice. Both have relatively low thermal conductivity.

Take a look for properties here:

http://cryogenics.nist.gov/MPropsMAY/material properties.htm

 

[belowzero]

In response to jabonet's original question, can I suggest a basic increase in good thick insulation to stop the valve reaching higher temperatures and as the valve gets down to -196 the material will be kept colder and may even stay chilled for the next days operation, if it does you will then get an accumalation of chilling and the valve and pipework will get colder each op.

 

[itascot]

I'm not sure what type of valve is being used, snd is exteriencing problems, but I have used metal seated plug valves in cryogenic test circuits without having experienced any functional problems. The circuit was assembled extarnally (due to health & safety reasons) where during summer, valves reached ambient temperature of up-to 40°C then went down to -196°C in a very short space of time (1 1/2" line).

 

[belowzero]

Metal seats do not tend to be bubble tight, as I would assume thats why soft seats are being used.
That said triple offset valves are a 'metal seated' but comply with soft seat leak rates.
Soft seats do not like uneven temperatures, if the valve is cold and one side warms up the disparity in temperature will inevitably cause a leak. same goes for the other way around.Virgin PTFE definately won't work, Just Glass filled PTFE won't work. A combination of glass and carbon, typically no more than 12 1/2% of either (although your not going to find this on the shelf).
The design of the seat is also critical to stop any extreme stresses on the PTFE seat. I would assume this is the major problem as the PTFE is cracking it is happening at lower temperatures, as it will just flow at higher temperatures. If you could explain the cracking, radial etc it would be easier to tell you what is going on. The seat must be allowed to move rather than wedge itself and cause stress fractures.

 

[itascot]

If the temperature is -196°C the valve would not have to be bubble tight, as it would be dealing with liquid.
However, best to be advised on currently used valve type before any real conclusions can be given as to failures & reasons for these.

 

[belowzero]

jabonet states that the temperatures goes up and down above -196 Deg. Therefore its probably quite critical and thats why soft seats have been chosen.

 

[itascot]

At this point what we need to know from jabonet is the following:
Are the valves in service at ambient temperature in open position allowing both liquid oxygen and nitrogen to flow through valve hence cooling line & valves before closing ?
...OR are valves in closed position during this cooling process.
The way I understand it, you only use these valves to "fill something" OR to "cool something" and valves are used to Firstly allow flow (and all components cool to -196°C), then at weekends, the system is isolated, resulting in all components returning to ambient temperature.
If this is not correct, jabonet, please correct me; IF it is correct, the problems you have with soft seats are due to thermal contraction of soft seats Vs the thermal contraction of remaining valve components, leading to valve failure (this can be when cooling in closed position, as seat tries to contract more than valve disc allows, or during closing stroke where the seat is smaller than disc, and is damaged during this operation).
The above reasoning is what leads me to the opinion that metal seated valve could well-be the answer.
Ciao,