Rubber Pressure Seal Design

[Ussuri]

I am looking at designing a rubber (softhane) seal. The seal will be placed between a steel and concrete face and subject to a compressive force. The seal will need to retain around 6 bar pressure.

Does anyone know of any guidance, standards, etc for determining sizes, rates of compression, hardnesses, sealing force versus retained pressure, to come up with a suitable design?

 

[israelkk]

How are you dealing with the concrte porosity, cracks and face flatness?

 

[Ussuri]

Its an interesting point.

The concrete is existing. It was poured in 1975 and has been submerged in a marine environment ever since. At the time the concrete was placed it was a particularly high strength mix. Typically, the characteristics that make a high strength mix also reduce permeability. This will have been a high quality mix with good durability.

It will also have been designed as water retaining because it contains a mixture of crude oil and water so leaks to the environment are not acceptable. As a water retaining design the cracks will have been controlled by the provision of adequate reinforcment. It is possible that over the years chloride ions have permeated the cover to reinforcement and caused corrosion. However there is no visible spalling.

The upshot of this is that there is not much we can do about it and have to take it as existing, but to my way of thinking the primary leak path will be the contact interface between rubber and the concrete, not through the concrete medium itself.

Regarding variations in the surface profile. We were looking at a two layer compound seal using bonded sorbothane. The top layer at 30-40 IRHD the lower layer at 5-10 IRHD.

I am by no means an expert on rubber seal design, hence the request for guidance.

 

[hydtools]

Consider, also, an outer layer of the harder material to prevent extrusion of the softer material. The outer layer need not seal, only create smaller extrusion gaps which the softer material can not extrude through and provide stiffer backup material to retain the softer material. It would serve the same purpose as a backup ring for an o-ring, except in this case it would be a backup ring for the softer seal material.

Ted

 

[btrueblood]

Ussuri,

Are you talking about the foam-type sorbothane? It may be too permeable for sealing, you may want to test a piece at bench-top scale.

 

[MintJulep]

You are replacing an existing seal?

Since the original has apparently performed satisfactorily for 32 years, why not simply replace-in-kind?

 

[Ussuri]

The application is actually a deployment onto an existing structure. The steelwork is in the form of a large round plate (think of a jam jar lid). It will be lowered from the surface to rest on the top of the existing concrete. Once in place it will provide a platform for a number of operations.

Hydtools, we were considering providing a couple of sections of steel either side of the seal to retain it. Something like 50mm depth either side of say 75-100mm seal.

btrueblood, the supplier we spoke to advised us against a foam type seal. The product they were suggesting was a urethane (softhane ??) solid rubber. I apologise if the names are a little confused.

MintJulep, This is a new seal.

My main worries are that we need to get a handle on how much the rubber will compress under the weight of the 'jam jar lid' , and how much force (ie submerged weight) we need to exert onto the seal such that it will retain the 6 bar.

Thanks for the responses.

 

[hydtools]

You might consider o-ring cord stock. It is available in cross-sections up to 2" dia and a variety of materials. Make your custom seal by cutting the cord stock to length and gluing the cut ends together. There are kits available to help with the assembly.
Make your gland dovetail in shape to physically retain the seal. Allow some seal to extend beyond the gland to seat against the concrete. The steel gland material will provide the physical stop against the concrete. You do need to determine how much force will be needed to force the gland to contact the concrete. Determine balance of forces underwater.
Compression of the seal material should be on the order of 20%. Allow enough gland space to allow for the compression of the seal material without filling the gland 100%. Provide for the gland to fill to about 75% of the unfilled gland.

See the Parker Handbook for some guidance:

http://www.nedc.com/literature/nedc-parker-o-ring-catalog.pdf

Ted

 

[Denial]

I was recently involved in a project where we were investigating the behaviour of a rubber seal in an underwater structure. The seal was an extrusion, with a hollow, D-shaped profile. We removed a small length and tested it, up to and beyond the point where the D-section had closed on itself. We then set up an FE model, and analysed that under an increasing series of enforced displacements, again, up to and beyond the point where the D-section closed on itself.

I was surprised by the extent to which the results from the physical model and the FE model were in agreement across the full range of displacements, despite the gross deformations involved. FE material models for rubber have come a long way.

Based on this, admittedly limited, experience, I would suggest you could use FE to explore different seal shapes and materials. Presumably you think likewise, or you would not have placed you post in this forum (although I note you hedged your bets by posting in another forum as well).

As for your specific question of how much force would be required to retain a given water pressure, first principles suggest that water will not be able to force itself through a rubber-to-steel interface provided that the normal pressure across that interface exceeds the pressure of the water (by an appropriate factor of safety). The FE analysis will give you that interface pressure.

Minor irregularities in the smoothness of the mating surfaces will confound this a bit, but I have no idea how much. Perhaps some testing is needed? Or merely an increase in your factor of safety?