Estimating preload reduction in a joint between two threaded parts with a gasket for sealing.

[vaibhavahire11]

Here I have a threaded joint between a plate and a fluid adaptor. The joint has a gasket for sealing. I am trying to calculate the reduction in initial preload when fluid pressure is introduced. Specifically I want to know if I shall subtract the fluid separation load as a whole or some fraction of it?
I made a rough sketch in CAD which is attached here.
Thanks in advance!!

 

[LittleInch]

Hmmm. Not easy to see what you're doing here but I think if this seal is not exposed to the fluid directly, I can't see where pressure would make any difference.

This is something she you know what you're looking at but I don't from that sketch.

 

[GD2]

Op,
Not sure what you are trying to achieve. If it's about designing an O-ring, did you try the online sizing tools most O-ring manufacturer provides?

 

[keller529]

The sketch isn’t a good design, but I’ll entertain it - the answer is you need to account for some, but not all. You do need to estimate deflection due to gaps in the thread, and some deflection of the thread. (This gives you deflection - you calculate the loss in preload force by F=kx; your face seal is a spring in this case)

The better design is to allow some flange-to-flange contact where you can better calculate preload. Then you don’t have to worry about loss in preload. I’m also unsure about pressurizing threads, seems suspect. If you can switch to an id/od o-ring would probably be better, or if a face seal is required you could tweak the placement and design a bit to improve it. At a minimum you need a positive stop somewhere

 

[goutam_freelance]

Check with the screw formula if the threads are self-locking.

If self-locking, there will be no change in gasket preload as gasket compression is locked.

However, assuming that the internal pressure is also felt at the gasket seating surface, the axial force on adapter threads will reduce by an amount equal to the fluid pressure felt at the gasket seating surface multiplied by the projected horizontal annular area at the seating surface (either top or bottom).

 

[vaibhavahire11]

Hmmm. Not easy to see what you're doing here but I think if this seal is not exposed to the fluid directly, I can't see where pressure would make any difference.

This is something she you know what you're looking at but I don't from that sketch.

Thank you for your reply. The joint between two parts is threaded one! I could not show it properly because I couldn't use AutoCAD to draw it.
Pressure will act on an area equivalent to seal area which is bound by diameter of gasket in this case. The separation load will reduce compression in gasket. I want to know by how much.

 

[vaibhavahire11]

Op,
Not sure what you are trying to achieve. If it's about designing an O-ring, did you try the online sizing tools most O-ring manufacturer provides?

It's a cryogenic joint. So O-ring cannot be used. Design is already frozen. I am merely analyzing it.
Specifically, I am analyzing if the factor of safety on torque being given. To do that I should know how much of the fluid separation load shall be subtracted from initial preload.
Hope the scenario is clear.
Thank you.

 

[vaibhavahire11]

The sketch isn’t a good design, but I’ll entertain it - the answer is you need to account for some, but not all. You do need to estimate deflection due to gaps in the thread, and some deflection of the thread. (This gives you deflection - you calculate the loss in preload force by F=kx; your face seal is a spring in this case)

The better design is to allow some flange-to-flange contact where you can better calculate preload. Then you don’t have to worry about loss in preload. I’m also unsure about pressurizing threads, seems suspect. If you can switch to an id/od o-ring would probably be better, or if a face seal is required you could tweak the placement and design a bit to improve it. At a minimum you need a positive stop somewhere

The design is made compact because of space constraints. I want to know if there is any industry accepted number to estimate preload reduction. For example, if fluid separation load is Ff. I have to reduce (K)*Ff from initial preload.
So what would be the value of K in general? I have seen it being assumed as 0.2 in some cases.

 

[vaibhavahire11]

Check with the screw formula if the threads are self-locking.

If self-locking, there will be no change in gasket preload as gasket compression is locked.

However, assuming that the internal pressure is also felt at the gasket seating surface, the axial force on adapter threads will reduce by an amount equal to the fluid pressure felt at the gasket seating surface multiplied by the projected horizontal annular area at the seating surface (either top or bottom).

For calculating fluid separation load, we need to take circular area where pressure acts which is usually taken as area bound by gasket mean diameter. I am more interested in knowing its effect on the preload. Should I subtract the entirety of fluid separation load or only a fraction. If fraction, what is the generally accepted value.

 

[LittleInch]

My problem is that I don't understand how much pressure the seal is seeing past all the threads.

Seems to be anywhere from virtually none, to some to all of it.

 

[vaibhavahire11]

My problem is that I don't understand how much pressure the seal is seeing past all the threads.

Seems to be anywhere from virtually none, to some to all of it.

Since there's no sealing at threads, pressure will be communicated upto the gasket.

 

[goutam_freelance]

For calculating fluid separation load, we need to take circular area where pressure acts which is usually taken as area bound by gasket mean diameter

This is true for normally bolted flanges. However, for a threaded adapter like this, this is not true, in the sense that the internal pressure in the pipe is not trying to separate the flanges. This is the axial force developed in the thread when internal pressure is inside. As threads are presumed self-locking, this does not affect the gasket compression force.

Since there's no sealing at threads, pressure will be communicated upto the gasket.

As no gasket can seal 100%, there will be a leakage flow, however small, across the gasket. So there will be a pressure drop across the gasket seating face and atmosphere. This will be the pressure acting to separate the flanges, and in turn, trying to reduce the preload across the gasket.

This is very difficult to calculate theoretically, and the variables are, available preload, flow restriction across the thread, sealing property of the gasket etc.

If you are interested in reducing the leakage, you need to select a good gasket material and properly tighten the adapter.