Nozzle Subject to Piping Loads - Allowable Stresses

[BenStewart]

All,

First post so apologies for any violations.

I have an ASME Sec. VIII Div. 1 pressure vessel for an FPSO application that is currently being designed by a subcontractor and an issue regarding the assessment of nozzle loads has come up. The project pipe stress group has calculated revised higher nozzle loads and whilst the vessel sub-contractor is on holiday I am reviewing the revised loads to determine if they are acceptable and in doing so may have identified an issue. The sub-contractor has evaluated the shell and nozzle stresses using WRC 107 and 297 but appear to have assessed nozzle stresses at the shell using the local membrane and local membrane + secondary bending allowable stresses.

My understanding is that it is seldom that nozzle loads result in stresses that are either only primary or secondary and therefore I use the lower primary membrane and membrane + bending allowable stresses. (Wrt to the piping loads, only results from un-factored load combinations have been provided i.e. there is currently no information for the loads which make up the load combination: dead weight, thermal, accelerations etc). This may be overly conservative but my approach was further reinforced when I read a paper that suggested in some conditions thermal piping loads were more like primary than secondary. I am from the UK and more familiar with PD5500 which applies a membrane + bending allowable stress of 2.25 times the basic allowable at nozzles irrespective of the stress categorisation. I believe this is for simplicity as is half way between the primary and secondary stress allowable - Might be wrong, just my interpretation.

Using the primary allowable stresses the assessment fails but increasing the allowable stresses in line with what the vendor used the assessment passes. Now i'm being asked to justify my choice of allowable stresses, why I think the sub-contractor is wrong (and other vessel sub-contractors, apparently) and provide evidence from within the code. I'm not sure if I can find such evidence so hoping some advice/experience can be shared and whether what I've said is broadly correct?

I think i'm going to push them down the route of doing elastic-plastic DBA to get rid of the stress categorisation problem and unless more details of the piping loads are provided assume all are as "D" in table 5.2 of Div 2. Would you also agree on this approach?

Regards,
Ben

 

[Trestala]

BenStewart,

I read a paper that suggested in some conditions thermal piping loads were more like primary than secondary.

Our pipe stress group describes 'thermal' loads as loads induced to the vessel from piping thermal expansion, thus are more like primary. It may be similar to your case? You could verify with your pipe stress engineer.

...with PD5500 which applies a membrane + bending allowable stress of 2.25 times the basic allowable at nozzles irrespective of the stress categorisation. I believe this is for simplicity as is half way between the primary and secondary stress allowable - Might be wrong, just my interpretation.

The calculation method of Annex G of PD5500 is different from WRC 107 & 297 (in considering pressure intensity), even though they have the same basis. Using allowable stress from the method of Annex G to WRC 107 & 297 would, in my opinion, not be correct.

WRC 107 and 297 are commonly performed with the allowable stresses on Figure 5.1 of ASME VIII Div. 2 as applied to Division 1 vessels, which have similarity with Figure A.1 of PD 5500.

Popular commercial calculation softwares perform WRC 107 & 297 calculations per Figure 5.1 of ASME VIII Div. 2.

(Pm) + Pl < 1.5f
(Pm) + Pl + Pb + Q < 3.0f

As a side note, for FPSO, you have additional considerations such as fatigue from wave motion, and probably an explosion scenario if applicable. Additional considerations for your nozzle and flanged connection design.

Just my 0.02.

 

[TGS4]

That would likely have been my paper that you read

Which stress category is failing which limit?

For the benefit of all, Table 5.6 provides an example that the shell membrane stress adjacent to a nozzle, resulting from the restrained free thermal expansion of the attached piping, should be classified as primary.

For the M+B check, that is for determining Protection Against Failure from Cyclic Loading: Ratcheting. Accordingly, you should be using the operating load RANGES and not just single load cases, as I described in my blog post:

http://becht.com/blog/asme-section-...ic-analysis-discussion-collapse-vs-ratcheting

 

[BenStewart]

RaymondN:

Thanks, that's helped with my understanding.

With these packages is setting the stresses in the nozzle adjacent to the opening to 1.5S and 3S by default always correct? I think that is true for the stresses in the shell adjacent to the nozzle and is comparable with Table 5.6 of Div 2, but I'm unclear about the stresses in the nozzle and why I think maybe the primary allowable stresses are more appropriate.

It seems that some of the project engineers and vessel sub-contractor have convinced themselves that a fatigue analysis is not required as Div 1 doesn't have the rules to cover it, despite an average wave cycle of approximately 5-10s. I've pointed them to UG-22. "Fortunately" this vessel has almost finished fabrication and has plenty fillet welded attachments...it will be interesting to see what is done here.

TGS4:

Thanks for the link, and yes it was your paper!

It is the stress in the nozzle that is failing on membrane + bending stress. If I believe this to be a primary stress I think this is for the Protection Against Plastic Collapse failure mechanism and if it were secondary definitely for Protection Against Cyclic Loads? In your opinion is it correct to categorise the stress in the nozzle as primary? I think this would be consistent with Table 5.6 i.e. nozzle, within the limits of reinforcement, pressure and external loads...

With commercial programs it appears many of them use the secondary allowable stresses for stresses in the nozzle, and seems to be an interpretation of Table 5.6 i.e. nozzle, outwith the limits of reinforcement, pressure and all external loads and moments. Is that incorrect or am interpreting it all wrong?

Thanks,
Ben

 

[TGS4]

I'm a little confused... WRC107 doesn't have the ability to calculate the stresses in the nozzle. Plus, there is little in the way of interpretation for classification of stresses in the nozzle itself; the rules in paragraph 5.6 are reasonably clear about that.

 

[BenStewart]

Sorry for the confusion, I used WRC 297 to calculate the stresses in the nozzle.

Yes, paragraph 5.6 is clear - should have read things more thoroughly beforehand.

Thanks for the help.

 

[TGS4]

Glad to help. Good first post. Please feel free to come back with more questions!