O-ring self acting gasket flange design

[bogu]

Dear All.

I have been asked by a friend to review some calculations on a vessel he is importing prior to submittal to the relevant authorities. The majority of the design is fairly straightforward but at one end there is a quick release lugged flange assembly with a self-energising o-ring gasket set up. I have designed many flanges but it has been some time since I have looked at this set up with the self-acting o-ring gasket and as I have changed companies all my previous designs are unavailable for reference. I have adopted our AS1210 section 3.21 simular to ASME VIII appendix 2 with dimension A being the outermost limit of the non broken perimeter of the flange i.e. the bottom of the lugs. Dimension B is the id of the flange with dimension C being midway of the lugs where they are held by the retaining mechanism. I have checked the flange assembly part that has the groove that holds the gasket (the locking mechanism welded to the vessel) using the moments MD & MT only as MG will be zero and I get a simular thickness to that being supplied. But when it comes to the lugged flange on the swing door, which does not have the o-ring groove, i.e. full face I get twice the thickness than that supplied. Does MT have to be considered for this scenario? I have also checked the swing door as per AS1210 3.14 simular to Appendix 1-6 ASME VIII and still get almost the same increased thickness what am I missing. I am sure that once my memory is jogged it will be obvious.

Thanks in advance.

Bogu

 

[johnnymist2003]

Hi Bogu,

for me to get a clearer picture of your problem, i'd need to see a little sketch or something, but from what you describe, I hope the following helps. For a flange pair working under similar design conditions, the moments to be used for the design of both flanges would be identical. If one of the pair has a groove for an o-ring, but the other is a machined surface that the o-ring seats against, then one would use the gasket factors and seating stress for the o-ring for the design of both flanges. I assume that in this instance, the o-ring groove is a little less deep than the o-ring diameter (so that the o-ring stands proud of the surface somewhat). However, if the flanges are on either side of a plate, and have different design conditions and/or a different gasket (like a heat exchanger trapped tubesheet with two girth flanges either side), then one would take the highest of the bolting loads and use them in the design of the other flange.

Sometimes, the bolt spacing criteria is governing for the design of a flange with a full-face gasket. More often than not, this is actually what governs the required thickness of the flange.

If this does not help enough, let me know.

John

 

[bogu]

Johnnymist

Believe or not what you have said has cleared up the problem.

Many Thanks

Bogu

 

[arto]

Also - if it's a wedge-type lug [rather than flat]be sure to include the equivalent effect to over-bolting/assembly moments [i.e., over-tightening the locking ring/over-pressure in the hydraulic cylinder]

 

[unclesyd]

If I'm reading you the post right there is should be no seating stress caused by the o-ring. A properly installed o-ring has very little to seat. When we used an o-ring we only accounted for hydrostatic forces using the O.D. of the o-ring groove as the effective gasket diameter.
We used the same approach on self-energizing Al gaskets.

 

[johnnymist2003]

Hi bogu,

glad I could help.
John