Nozzle on Flat Head - FEA (Stress concentrations)

[Sjqlund]

The pressure vessel i'm designing is basically a box, i.e. flat plates. And i've understood that WRC107/297 are not applicable for nozzles attached to flat plates (correct me if i'm wrong), soo FEA it is.

The nozzle is welded to the plate. The weld is simply modelled as a straight line. This causes a stress concentration in the top-weld / nozzle junction. In an earlier report they assesed the stress values a distance away from this stress concentration (they simply put the path a distance of ~10mm above the stress concentration) - but this was however according to another code. Is this allowable according to ASME code?

 

[TGS4]

Nozzle attached to flat plates can indeed be evaluated using WRC 107 (now 537).

You say that the weld is modelled as a "straight line". Are you using shell elements of 3D-solid elements? That will make a huge difference in how you post-process the results for compliance to ASME Section VIII, Division 2, Part 5 (I assume that when you say "ASME", this is the Code that you mean...).

 

[Sjqlund]

Oh really? I guess i'll have to order WRC 107 then. Thanks!

The FE model is made of solid 3D elements. What i mean by a 'straight line' is that there is no fillets where weld meets nozzle / vessel.

Regarding which ASME code i'm following, i should have stated that. However i think the reason that i didnt, is that i'm not really quite sure. I've only been told to follow ASME VIII div 1, and also "Stress analysis by FEM". I guess i should clarify with the customer which code to follow when looking at FE results.

 

[TGS4]

For starters, please check out my post on qualifying a vessel component in ASME Section VIII, Division 1 using Article U-2(g) and FEA here.

If you are not including the fillet weld cap, then I would not place the SCL any distance from the junction. I would suggest that the stress linearization be done appropriately. If there are actual issues, you can either add the fillet weld and place the SCL at the toe of the fillet, or modify the loads.

 

[Sjqlund]

Very helpful blog - thank you.

About the WRC bulletins being applicable for flat plates.. When i search around different forums, i find that the bulletins hold some limitations regarding thickness ratios etc.

If i have a flat plate of thickness 80mm, with an attached nozzle of wall thickness 8,2 mm (NPS8 - placed on top of the plate) - would WRC107/297 still be applicable?


PS. i'm sorry i keep asking about WRC, but it'll take a while before i can acces the bulletins. And it would be extremely helpful if i could get confident whether i need them or not. And i tried emailing WRC already (no answer so far though)

 

[SnTMan]

Sjqlund, are you aware that WRC-107, etc, analyzes EXTERNAL loads on the nozzle, rather than reinforcement of the opening?

Regards,

Mike

 

[Sjqlund]

Yes i'm aware of this. I found WRC bulletins online now.

If i take D as infinite (due to flat plate) i have the following ratios:

d/t = 26, OK
D/T = inf OK
d/T = 2.7 NOT OK (should be larger than 5)
and (d/D)*sqrt(D/T) = 0 (i guess d/D goes faster to 0 than sqrt(D/T) goes towards inf, but not quite sure.

the bulletins doesnt mention flat plates. What you you guys in here think about using WRC with these ratios - would it be useless or useful?



d=219, D=inf, T=80, t=8,2 (in mm)

 

[TGS4]

Further to what SnTMan said, the purpose of WRC537 is to calculate the stresses in the shell (in your case flat plate) caused by the external loads on the nozzle. In you case, I am quite confident in saying that the flat plate is most definitely not the weak point due to external nozzle loads. The T/t ratio of almost 10 is the give-away.

Some other failure mechanism will be at play. Since WRC537 doesn't address the nozzle itself, I'm going I lean towards useless.