ASME VIII stub-end calculation

[cornemo]

Hi all,

I have a question concerning the calculation of lap-joint flanges according to ASME VIII div. 1 appendix 2.

Concerning the loose flange the code is very clear and tells to calculate it as a loose flange with G being the center of the contact between loose ring and stub end (image 1 from appendix 2).

Concerning the stub-end the code doesn't give any recommendations, accept that the lap joint has to be checked for shear stress if shear stress can occur.

But now: How to calculate the stub-end?
In my opinion:
- calculate the stub-end as a fixed or loose flange type, just as a normal slip-on flange, but with the bolt circle being the center of the contact between loose ring and stub end.

This is also the method described in the European code EN-13445.
This however creates thick stub-ends and flanges like the Van-Stone flange can't be calculated this way.

Can somebody please give me some more insight in the calculation of stub-ends according to ASME VIII appendix 2?

Thanks in advance for your reply.
With regards,
Corné

 

[fegenbush]

cornemo,

You are correct that it is difficult to use a Van-Stone type flange with ASME work due to the thicknesses required. It is generally difficult to get sufficient lap thickness. To compound the problem, the surface of the lap joint will require a machine finish, and that further reduces the thickness.

A possible solution for you would be to weld an ID/OD ring to a pipe (fabricated stub end) where you are unable to find an acceptable pre-made fitting. This usually happens with very thick pipe or uncommon materials.

Let me know if you need a sketch.

 

[cornemo]

Thanks for your reply.
I understand what you mean, but the problem is that I need to design a lap joint flange. The stub-end will be very large if the method according to ASME VIII div 1 appendix 2 is followed.
Also the code doesn't mention how to calculate a stub end.

So my actual question is how do you calculate a stub-end according to ASME VIII?

The Van-Stone flange is just an example. The ring you mention will have the same problem, it will be too thin to be used as a stub-end.

 

[StoneCold]

As a point of reference, all the pipe stubends I deal with the "flange" of the stubend is the same thickness as the pipe wall. That is for cs or ss sch 40 1" to 6". Beyond that I am not sure. That is not a code method of calculating it but it may give you an idea if your answers are right.

Regards
StoneCold

 

[SnTMan]

cornemo, Apx 2, Figs (1) and (1a) tell you everything you need to know.

Regards,

Mike

 

[cornemo]

@StoneCold: I know that design, but that design can not be verified using the calculation methods given in appendix 2. Also, if I check ASME B16.5 lap joint flanges, the thickness of the stubend and the loose flange are almost the same.

@SnTMan: Only thing those images show is (in image 1a I believe) that tl >= t. Which means the stubend must be evenly thick or thicker as the pipe wall. However these images explicitely give a full face gasket underneath the stubend. If I have a smaller gasket this reference is not valid in my opinion because there will be a bending moment in the stub end flange.

I would like to here more / other opinions regarding this issue. Especially also from people who are also working with the EN-13445 code which gives this explicit calculation method for a stub end (comes down to use the fixed flange method).

Regards, Corné

 

[SnTMan]

cornemo, I read the words "independent of gasket location" as not limiting to full face gaskets.

True, no explicit calculation seems to exist for the stub end. I make them at least equal to the neck and go on.

After all, commercial stub ends are not restricted as to type of gasket.

Regards,

Mike

 

[cornemo]

The words "independent of gasket location" in my opinion concerns the loose flange gasket reaction diameter, not the design of the stub-end.

Commercial stub ends, as in ASME B16.5, have a thickness more or less equal to the loose ring thickness. And are thus much thicker than the pipe wall thickness.

 

[fegenbush]

cornemo,

ASME B16.5 (2009) 6.4.3.1 states: The finished thickness of the
lap shall be no less than nominal pipe wall thickness.

Generally, when fabricating from plate, I like to leave a healthy machining allowance, since I am allowed to be thicker.

 

[cornemo]

You are correct about B16.5, it was my mistake from table 8 to read the lapped joint thickness as the thickness of the hub, instead of the thickness of the loose flange.

Still it is strange that all European codes (EN-13445, DIN2505, AD-Merkblatt B5, RToD) calculate the stub-end the same as a fixed flange. ASME VIII seems to be the only one not explicitely telling anything about the thickness accept that it must be larger than the pipe wall thickness. Normally the German code gives thinner parts as the ASME code.

And I still can't find any justification to not calculate the stub-ends bending moment and flange thickness.

 

[SnTMan]

cornemo, you are free to do so

Regards,

Mike

 

[cornemo]

I understand that I'm free to do so. But I am trying to find some background information on why ASME doesn't have these rules and how the ASME rules are interpreted on this specific subject.

 

[SnTMan]

Understood, and good luck with the "why".

As for interpretation, I think the previous posts illustrate that.

Regards,

Mike

 

[cornemo]

Thanks for the answers so far. I would like to here from more people on this specific topic.

 

[Tomtation]

You need to check the shear due to the bolt load. The shear area is equal to the circumference at the point of contact of the flange to the stub end multiplied by the thickness of the lap.
Think of the flange as a punch die trying to punch a washer out of the stub end.
Hope this helps

 

[Tomtation]

You need to check the shear due to the bolt load. The shear area is equal to the circumference at the point of contact of the flange to the stub end multiplied by the thickness of the lap.
Think of the flange as a punch die trying to punch a washer out of the stub end.
Hope this helps.

 

[cornemo]

Tomtation, thanks for your reply. I know this check, which is also mentioned in ASME VIII. I'm looking for an explanation why ASME VIII doesn't do a bending stress calculation for the stub end, while all european codes (EN-13445, AD-merkblatt, DIN2505,...) do.

In my opinion the stub end will have to withstand bending forces too.

 

[Tomtation]

There are many instances where the Code doesn't require additional stress calculations, that is why they often limit certain configurations such as the lap to the same thickness as the neck and a minimum radius between the two.