Hi,
I have done pipe stress analysis on an offshore Flare line which is connected to an existing KO drum and we have got different scenarios such as Max Des/Min Design/wind/wave and Blast.
Nozzle is overloaded in Operating case and Blast.
we have decided to do a FEA analysis on the Nozzle with Abaqus.

the procedure which we have considered is that to model the Nozzle(connected to the equipment)and put the boundary conditions and calculate the von mises stresses on the nozzle.
then compare these stresses with the allowable stress of the Nozzle's Material in Operating Case and Yield Stress in Blast.
my question is if the procedure is good enough or shall we consider other methods based on ASME section VIII Div 2?

The von-Mises stress calculation is appropriate. I question the comparison to the allowable stress in the Operating case. What allowable stress? B31.1/B31.3 don't define an allowable for the Operating case. von-Mises stresses are usually compared to some factored value of the Yield Stress.

Does the owner/operator have an opinion on this evaluation?

Thanks Richard,
this is an existing OLD KO drum and the existing flare line has been replaced with a new one.so we don't have access to the vendor.

I agree that Von mises stresses are usually compared to the Yield stress.but shall we consider the Yield stress in the operating temperature?(table Y-1 of ASME section II)
or we have to check them with maximum allowable stresses in B31.3?

for Blast I think there will be no doubt to check the Von mises stresses with Yield stress of the material at Room temperature.
Regards

Hi there,

The problem is a little more complex. For "Shell/Head-Nozzle Junction" area, the local stress analysis as per ASME VIII-2 Part 5 needs to evaluate and qualify three main type of stresses:

1) General Primary Membrane Stress -> Pm
1) Local Primary Membrane Stresses -> Pm+Pl (General Primary Membrane + Local Primary Membrane) stresses
2) Resultant Secondary Stresses -> Pm+Pl+Q (General Primary Membrane + Local Primary Membrane + Secondary Membrane&Bending) stresses.

Pm is calculated excluding gross-structural discontinuities (e.g. stress concentrations are excluded/ignored).

Pl and Q are calculated including gross-structural discontinuities effects
(e.g. stress concentrations are considered).

If you used Shell elements, then when you plot the Equivalent (Von Mises) Membrane stress map you'll get (Pm+Pl) effective stress distribution directly. When you plot the Outside and/or Inside Wall Surfaces' Equivalent (Von Mises) Membrane stress maps you'll get (Pm+Pl+Q) effective stress distribution directly.

Note that Pm+Pl should typically be checked for Pressure + Weight + Occasional Loads (Wind, Earthquake, BLAST, etc), and the common limit is YIELD LIMIT (1.5 times Allowable Stress).
However, as per ASME VIII-2 post-2007 editions, Pm+Pl need also to be qualified for Pressure + Weight + Occasional + Thermal-Expansion Loads (e.g. including external piping loads from restrained thermal expansion).

Pm+Pl+Q should typically be checked for Pressure + Weight + Occasional + Thermal-Expansion Loads, and the common limit is 3 times Allowable Stress or 2 times Yield Limit (see ASME VIII-2 Part 5 for details).

Now, regarding BLAST load-induced stress qualification: BLAST load is an Accidental Action, with a very low occurrence probability. Therefore, an increase of "Allowable Stress" and "1.5 times Allowable Stress" limits for Pm and (Pm+Pl) may be assumed.
In this regard, I suggest to have a look to ASME III-1/NB nuclear code.
A typical/conservative approach could be +20% increase (similar to Level C Service Limit / ASME III-1/NB). This would be also in line with the "old" (70's, 80's) ASME VIII BPVC provisions, which qualified "Occasional" stresses considering k = 1.20 increase factor for Allowable Stress.

For Pm+Pl+Q resultant secondary stress, an increase of 3xS or 2xSy limit is more difficult, because would mean to accept remanent/plastic strains development.
However, Blast might be considered a SINGLE CYCLE load, so that an increase of allowable stress limit based on low number-cycle fatigue stress limit might be a solution. You might consult the Owner/Client to agree an acceptance criterion.

Alternatively, an elastic-plastic analysis may be performed to quantify the effective strains and stresses. Abaqous is a high-performance software for this purpose.


These are some brief thoughts I tried to synthesize here. Maybe will help...

Regards,

...a short errata:

... When you plot the Outside and/or Inside Wall Surfaces' Equivalent (Von Mises) Resultant stress maps, you'll get (Pm+Pl+Q) effective stress distribution directly.

Sorry for inadvertence.

Dorin,
Thanks a lot for your explanations. I have some questions though:))

in ASME section VIII-2 part 5(2017) has been introduced 3 ways for elastic Plastic evaluation.in the first place in such a case which of these 3 methods are the most suitable for FEA analysis?

in 2017 edition of ASME section VIII div 2 I could not find 3 times allowable stress for Pm+Pl+Q.can you pls make a screen shot of this?

when we talk about the Yield limit shall we consider it in the Operating Temperature or in the room temperature for FEA analysis?


Thanks again for your concerns:)

<< in 2017 edition of ASME section VIII div 2 I could not find 3 times allowable stress for Pm+Pl+Q.can you pls make a screen shot of this? >>

See ASME VIII-2 Para. 5.5.6 / Clause 5.5.6.1 (Ratcheting Assessment - Elastic Stress Analysis):
S_PS limit for "Pm+Pb+Q/Pm+Pl+Q" is defined at point 5.5.6.1(d).

In addition, see Figure 5.1, which is very intuitive in relation with ASME VIII-2 stress categories and qualification.



<< when we talk about the Yield limit shall we consider it in the Operating Temperature or in the room temperature for FEA analysis? >>

Very good question. As I wrote above, Blast is an Accidental Load, with very low occurrence probability. Therefore, it would be unrealistic to assume the Design Conditions (typically being "upset" working conditions, and NOT quasi-permanent long-term operating ones) when we analyze Blast Load.
I would recommend to use the OPERATING PARAMETERS (for both Pressure and Temperature), because Blast would occur unexpectedly during operating service circumstances.
However, you should liaise with HSE Engineer on the Project, to investigate if there might be some fire scenarios previously to blast/explosion escalation. It may happen Vessel temperature to increase prior to explosion above the Operating level due to environment conditions (e.g. fire case).

Connected somehow with this idea, try also to use a Realistic Dynamic Drag Pressure and avoid conservative over-estimations. It is really HSE Discipline role to provide accurate input data in this regard...push them if they try to pass the responsibility...



<< in ASME section VIII-2 part 5(2017) has been introduced 3 ways for elastic Plastic evaluation.in the first place in such a case which of these 3 methods are the most suitable for FEA analysis? >>

It depends on the analysis requirements.
I would take as a minimum:
1) Protection against plastic collapse - see the Assessment Procedure described at Para. 5.2.4.4.
2) Ratcheting Assessment - Elastic Stress Analysis - see Para. 5.5.6.2 for the SIMPLIFIED Elastic-Plastic Analysis, which is based on Elastic Stress Analysis and does not necessarily require nonlinear FEA.
3) Ratcheting Assessment - Elastic-Plastic Stress Analysis - see the Assessment Procedure at Para. 5.5.7.2.

Please be aware that Abaqous is a GENERAL FEA package, although with high performances, but I am not aware if it has included the specific Pressure Vessel analysis capabilities as per ASME VIII-2.
On the contrary, PRG FEA Package (FE Pipe, Nozzle Pro modules) has been specifically developed for Pressure Vessel and Piping analysis as per ASME VIII-2 Part 5 requirements. The FEA modelling templates are included for the most typical "nozzle-shell/head" junctions, the loading cases are automatically selected to comply with ASME VIII-2 Part 5 Elastic approach, and, probably, the latest version includes also elastic-plastic analysis tools (I am not sure about this because I use with my Company an older version).

Well, these are my thoughts and personal opinions. Maybe will help.
Good luck!

3)

Dorin,

thank you very much for your very useful information.
I had to study a little bit since you made this post.

I was trying to read the help of Nozzlepro.there is something maybe you can advice also.in Nozzle pro our input loads are limited to Sus/Ope/Occ.how can we convert these loads to ASME Section VIII loads.I mean how should we define Pm,Pl,Pb based on the loads we derive from Ceasar II?Nozzle pro will do it Automatically.but is there anyway we can do it manually?

if we have a piping load on a nozzle,how can we devide it to local membrane or General membrane?

do you have any sample problem of a Nozzle assessment which you have done by Nozzlepro or FEpipe?I appreciate if you can send me a sample problem.
here is my email: khodayark@yahoo.com

many thanks again for your advices,
Khodayar