Reducing flange analysis using FEA or Div 1 1

[jt1234]

Dear all,
I have a Div. 1 vessel with top 36" nozzle, 600# B16.47 series A, connected to a forging flange with hub ( we call it reducing flange) and a 18" outlet pipe as shown.
Can this reducing flange be classified as custom made component, not fitting UG-34, UG-39, Appendix 2 or Div 2 Part 4, such that it meets Div. 1 46-4, "Design by Analysis", and using FEA only per Div. 2 Part 5 is acceptable ?

(Note, using UG-34, UG-39 or Appendix 2 will end up much thicker than standard 36" flange)

 

[SnTMan]

It looks to me that the reducing flange could be considered an Apx 2 integral hub flange, design would presumably be performed per Apx 2-5(2), flange pairs and subsequent rules. The operating and seating loads may not to be identical for each flange of the pair, due to the effect of Hp. Seems the B16.47 flange would govern.

Not clear to me whether you would be obligated to analyze the B 16.47 flange per Apx 2 if it is within its rating.

"Course this may not come out the way you want. Things don't always

Whether you can fit it into 46-4 I have no idea, might be a case of agreement among the parties.

Regards,

Mike

 

[TGS4]

Don't use the DBA rules to design flanges (unless you want to introduce the concept of "leakage" into your design - and you don't want to do that!). The DBA bolting rules tell you to go back to 4.16 (the VIII-2 equivalent of App. 2) anyway. How would this not qualify for App. 2 / 4.16?

 

[Si-bo]

As the others wrote, your reducing flange is an Appendix 2 integral flange and it needs to be calculated as per it. And there's not much you can do since most parametres are fixed by the 36" flange.

You can try making both custom designed flanges and calculate them as per appendix 2: maybe you can save some thickness and reduce the number and/or size of bolts.
Not sure it will be cheaper.

 

[r6155]

I agree with jt1234 about UG-39.
The thickness of the reducing flange must be at least equal to that of the 36" flange.

 

[SnTMan]

What Si-bo says makes a lot of sense. Why use a clunky Series A flange if no one is piping up to it?

 

[jt1234]

More clarifications to my post, to see if anyone can help:
This vessel was just fabricated and passed hydrotested and installed already. However, looking back in vendor's hundreds of pages of calc found no calc for this reducing flange.
The reducing flange has the thickness of 123.9 mm, the same as 36" Series A flange. By Appendix 2 or UG-39, it will fail due to such small 18" outlet hub and flange thickness. But it passed per Part 5 using elastic-plastic FEA analysis by vendor after the fact. Since this is custom made forging, will it meet 46-4 (a) such that FEA is acceptable to avoid a redesign and re-fabricate ?
(Note the hub in the sketch is considered as reinforcement. The hub taper ratio will not meet code as it is not considered as a typical flange design. It is considered as a blind and then drill a pipe hole with taper hub as reinforcement.)

 

[r6155]

I don't trust those who do FEA calculations. I'm sorry.

 

[r6155]

If an error in the calculations is detected after the pressure test, a full verification must be performed through an external audit before the pressure vessel is put into service.
My opinion.

 

[Trestala]

I don't trust those who do FEA calculations. I'm sorry.

If an error in the calculations is detected after the pressure test

Pointless responses.

 

[r6155]

@ Trestala
You have very little experience with pressure vessels.
You're not helping (OP)

 

[TGS4]

In this particular disagreement - I find myself siding with r6155. As a FEA practitioner myself, I may have a little more trust in "those who do FEA calculations", but I reiterate that FEA is not the right way to proceed in this situation. I also agree with r6155's response regarding an external audit being required since a defect in the calculations has been detected. I would also include in this audit the fact that this defective calculation was not found by the owner or their EPC during reviews. Things should never get to this point.

And don't try to use U-2(g) and Appendix 46 as an "out" for not doing the right thing in the first place.

 

[NovaStark]

I don't trust those who do FEA calculations. I'm sorry.

Are there any particular reasons for this ? I have my own reservations but that is just based on my specific experiences.

 

[r6155]

@NovaStark

I worked with nuclear pressure vessels and in the oil and gas sector, and those performing FEA lacked experience with traditional (including manual) calculations, fabrication, and inspection, and had never visited a fabrication shop. There were many errors, and it was tedious to review them one after another. I ultimately rejected them, and the work wasn't paid.
It wasn't a very happy endeavor.