Effective Gasket Width in ASME VIII-1 App2 Calcs

[antoniof]

Hello there. I am trying to do the calculations in ASME VIII-1 App 2, however, I do not understand how to calculate "N" (based upon the possible contact width of the gasket). On what aspect of the gasket do we base it : the sealing element of the gasket or the actual surface contact of the gasket on the raised face?

Thanks.

 

[SnTMan]

antoniof, Table 2-5.2 shows how to calculate N for various flange facings. Table 2-5.1 shows which column of Table 2-5.2 is used for calculating bo for various types of gaskets. Note the calculation for b from bo and Gcalc shown at the bottom of Table 2-5.2. Taking consideration of your particular flange facing and gasket type, these figures should get you there.

Regards,

Mike

 

[antoniof]

Thank you, however what I do not understand is why, lets say for a spiral wound, would you consider the outer ring as part of the sealing mechanism? If there is a defined sealing component in the gasket, why not use its dimension instead of the physical outside diameter of the gasket (or the end of the raised face, which ever comes first)?

Thanks
Antonio

 

[SnTMan]

antoniof, for a spiral wound, I would disregard the outer ring , and the inner , if one exists, and only use the "wound" part of the gasket. These rings are used as compression stops, and to keep the gasket from coming apart during handling. For a flat face as shown in sketch (1a), on the left hand side, N = 1/2(OD - ID) of the wound portion. For another example, double jacketed gasket using the same facing N is calculated the same, 1/2(OD - ID).

Having said that, I have seen a type of arrangement (not often) which uses a nubbin against the outer ring of a spiral wound. For this arrangement the calculaton used the loadings for spiral wound portion plus the loadings for a solid metal gasket with a nubbin. [A nubbin is as shown in sketch (1c).] It was not straight additive, each portion was "weighted". Very complicated, I don't recommend it.

With a flat face as shown in sketch (1a), left side, N will nearly always be just the nominal gasket width.

Regards,

Mike

 

[deanc]

Have you looked at SecVIII Div1,Appendix 2,Table 2-5.2?

May help.

 

[antoniof]

deanc, yes I have looked at that table, however, nowhere does it say if the dimensions they are talking about includes the outer/inner rings or not. My first guess was that no, since they aren’t actually sealing components, but I started to doubt myself because of the low bolt load I was ending up with, which I explain below…

And SnTMan, thank you very much. That was the answer I was looking for. But this leads me to my other question and the reason for asking the previous question. Lets use a spiral wound for a 24 inch ANSI 600 flange as an example. The sealing element width for such a gasket would be 1.125 inches. But the “effective gasket width” of it turns out to be 0.375 inches!

This is not a problem by itself, however, when calculating the bolt load required to seat the gasket, Wm2, in App 2, they use the “effective width” instead of the actual width to calculate the required bolt load. (Wm2 = pi*b*G*y) Doesn’t this basically replace the actual gasket by a smaller, narrower one (by quite a bit), which is easier to crush, which in turns leads to very low bolt load?

So why would you use the effective width when calculating the bolt load required to seat the gasket and Hp during the operating condition, instead of the actual width?


Thanks
Antonio

 

[SnTMan]

antoniof, now there's a question I don't know the answer to, the Code is what it is.

The methods of Appendix 2 were developed in the 1940's (?) and certain parts of it are empirical. You may be able to find a reference to the original work in something like the Taylor Forge Bulletin "Modern Flange Design", I suppose the original work itself is still available somewhere, although I've never read it.

Similar to your question, why is the flange design bolt load W calculated from the average of Ab(min) and Ab(actual)? Don't know, it just is.

As a designer, nothing prevents you from increasing the bolting if you think it necessary. I often do so if I think it is advisable, one thing I have learned is when you're out of bolting, you're out of bolting.

Regards,

Mike

 

[TGS4]

Before you think that the bolt stresses that you are calculating are the ones that you will actually use in service, please also read Appendix S, as well as ASME PCC-1. Note that Appendix 2 is a DESIGN procedure, that will help you design the flanges, not one that provides guidance on bolting these flanges in service.

 

[SnTMan]

TGS4, I second that.

 

[antoniof]

Ok, but what is the point of designing a flange with fictious bolt loads? Why not design the flange with the actual bolt load that will be applied?

And App S says “However, a distinction must be kept carefully in mind between the design value and the bolt stress that might actually exist of that might be needed for conditions other than the design pressure.” I thought the point of designing something was to try and predict the conditions the object will be in during operation so that it will be able to withstand them. However here, we clearly know that the stresses will be much greater, yet we still design it to lower stresses?

Thanks
Antonio

 

[eeromatti]

antoniof:

I have been thinking this same problem. I have been using higher seating stresses for gaskets to get higher bolt loads because I know that bolting loads will be higher than App 2 calculations predicts.

However, increase of bolt loads doesn't affect thickness of the flange too much. So App 2 gives you acceptable flange thickness although bolt loads and stresses are underestimated.