Metal to Metal seal

[Chel80]

Hi All,

I am a design engineer who is trying to overcome a sealing problem. We have been using orings for metal to metal seals and found it to be causing all sorts of problem. The seal itself is not good. Hence i was on the lookout for new type of metal to metal seals. the seal is for the equaliser blocks in a Blowoff preventer (oilfield equipment). The size of the seal can be in betn 20 and 35mm. Please advice.

 

[diamondjim]

Are you talking about a static or dynamic seal?
Is it a radial seal or an axial seal?

 

[MintJulep]

If you are using an O-ring, then you do not currently have a metal-to-metal seal.

What "all sorts of problem" are you having now?

Do you understand the reason for these problems?

Do you understand the loads on the existing joint?

Is the existing seal designed properly for the application?

 

[Philrock]

Pressures?

Temperatures?

Speeds?

Fluids?

Nature of current problems?

 

[Chel80]

The seal is static.

Yes you are right i am not getting a metal to metal seal. But since two metal blocks with a holes are to be connected with bolts, there can be liquid/gas leaks and hence this seal.

Pressure : 22500 psi
Temperature : 171 C
Speed : static
Fluids : hydrocarbon say .8 density fluid

The problem is we are facing is leaking of fluids and it is due to the failure of seals.

The cross section of the oring is circular and it sits in a pocket that has a cross section of square. The orring squeezes and gives a seal. I suppose the circular oring when pressed against the metal to metal become ellipse and seals opposite faces thus avoiding the flow of fluids.

Any more info please be my guest and i will do my best.

 

[ewh]

HAve you tried backup rings?

 

[btrueblood]

Do you ever rapidly drop the pressure? You could be getting explosive decompression effects due to absorbed gas within the O-ring material. To solve this problem, reduce the gas levels, or drop pressure more slowly (allow time for absorbed gas to diffuse back out of the seal).

 

[rorschach]

Is this an API 6BX flange? If so why in blazes are you using anything other than the metal ring seal that it was designed for? If it is not a 6BX flange, why not use similar geometry? when designed correctly, BX style flanges are capable of 20M service (20,000 psi).

O-rings can be used in this configuration to pressures approaching that, but a lot of attention must be paid to groove and gap design, o-ring crush, and backup ring design, it is not a conventional application for o-rings. also, 80-90 durometer o-rings, PPS or better yet, PEEK backup rings and at least a 20% crush are required.

 

[Chel80]

we have tried back up rings too.
It is not a flange. I have used metal rings in flanges and it works perfectly fine.
This one is for a manifold block in a bop which has a dim of 9 x 4 x 2 (all dim in inches) approx. and is placed on a body which is massive and hence the groove cannot be machined on the body. It has to be machined on the block which sits on the body.

I have thought about using metal seal ring. But have a circular dim of 25 to 35 mm and this is very very small. Would you recomment a metal seal ring for this dimension.

rorschack - "also, 80-90 durometer o-rings, PPS or better yet, PEEK backup rings" are these meant to be alternatives to the oring.

 

[IRstuff]

You still haven't really explained the nature of the failures.

Until you do, you can't expect a valid solution, since any proposed solution may wind up with the same problems.

TTFN

 

[Philrock]

Parker manufactures many kinds of seals. You might try contacting them.

 

[MintJulep]

How much bolt tension do you have holding the block against the manifold?

How much force from pressure do you have pushing the block away from the manifold?

The first needs to be bigger than the second.

 

[tomwalz]

A couple points before I start. We do some of this, mostly from the brazing and materials side. We operate under the assumption that almost anything is cheaper than failure so cost is minor design consideration. Finally, this is a very imprecise way to communicate so I may not understand fully what you are doing.

1. It is leaking because it is getting around the gasket. Therefore either the gasket fit isn’t good enough originally or the gasket fits well originally but deforms under pressure.

2. Use a flat gasket to fit the square pocket. Do the math on the “O” ring. With a round seal in a square pocket you are going to have a very thin seal on the top and bottom where the notch tangentially mates with the seal.

If you compress a round seal in a square seat you are disproportionately compressing it at the contact points. This changes the seal characteristics at the contact areas.

3. Check the tolerances on the mating metal surfaces.

4. We do lot of work where we will back the metal with tungsten carbide lapped very fine and smooth. This gives a very stiff, strong, inflexible backing to the seal on both sides. We “float” the tungsten carbide on braze alloy and the braze alloy compensates for incidental distortion in the metal housing due to thermal and other stresses.

If you want more help feel free to contact me.

tom


Thomas J. Walz
Carbide Processors, Inc.

http://www.carbideprocessors.com

 

[unclesyd]

Sounds like you need a self-energizing seal.

At your pressures the O-ring groove has to be very precise, cut specifically for you particular O-ring dimensions. There are other O-ring cross sections that may work better. Quad and so on.

One of the following companies will a sealing solution to satisfy your requirements. All three companies have very good technical assistance.

Static:
Metallic O-rings, coated and uncoated, and Spring Energized

http://www.helicoflex.com/products/

Static and Dynamic:
All types of energized seals

http://www.balseal.com/petrochem.jsp

Even more than all types

http://www.busakshamban.us/industries.htm?im=25

 

[rorschach]

Chel80, no, the PPS vs PEEK were alternate materials for the backup ring, but you'll need both the o-ring and the backup ring. and at that pressure, soft rubber o-rings will not seal well, hence the 80-90 Shore-D durometer callout. 60-75 durometer is not hard enough.

You also need to look at the seal face geometry, are the two surfaces flat (within .001 or less)? are they smooth? (16 RMS or less). variations in flatness or surface roughness will allow leakage.

consider instead of an o-ring, a thin soft copper washer/gasket instead. it will be soft enough to compressively flow into variations in flatness and surface roughness, and form a better metal to metal seal.