Nozzle/shell junction; internal pressure; local regions

[tomislav.begic]

Hello all,

I have a trouble assessing the FESU results of a nozzle with internal projection.
The existing vessel's nozzle should be checked against secondary loads, but high stresses due to internal pressure lead me to further questions.

I do not understand how exactly the junction of a nozzle with a shell is defined in the FESU model.
In my specific case/example, I have a nozzle/jacket according to (Figure 1).

2022-05-17 20_20_36-Stutzen.png

Figure 1

Question 1
From the cross section (Figure 2) I will make an assumption that R2 FESU made a junction in the middle point of a shell wall. That is why defined pressure of element No. 4998 is 0 bar(g) (is this correct?). However, the nozzle element above, which is still located in the shell wall have pressure 40 bar(g).
Is this a mere simplification of ROHR2? What about welds on both sides of the jacket?
Can I dismiss this stress area even though in local analysis it is >1.5*f?
This is very confusing to me, part of it due to my lack of experience and theory knowledge, but I would be grateful if I someone could provide info or point me to a literature about how the software exactly works on this part.

2022-05-17 20_32_55-Local_stress.png

Figure 2

Question 2
While showing the SPL analysis in FESU mode, I don't understand why the stress area on the nozzle/jacket side is compared to the total width of the region - ls+ln, instead of individual region. This would mean that ln region is extended beyond (n2+n3) width, which is opposite to citation few lines below.

In my understanding it would make a lot more logic:
jacket local stress region <= ls;
nozzle local stress region <= ln.

2022-05-18 12_10_56-Localregion.png

Figure 3

Furthermore,
Citation from R2Manual: "For a simple stress analysis the shell elements around the intersection line are divided into
element groups. Normally the elements of range n4 and n5 are comparable to width ls and the
elements of the range n2 and n3 are like width ln. The total width is nearly ls+ln in this case
and can be defined as region of local membrane stresses."
If I look at the model this is not the case, as the region n4+n5 ≈ 62 mm (ls = 81.7 mm, acc. to EN 13445-3 equation (C.6-2)), and the region n2+n3 ≈ 27 mm (ln=5.83 mm).


I hope this wasn't too tiring.
Thank you for your time!

[rohr2admin]

Hello,
First of all in the attachment is only the substurcture itself included, not the ROHR2 system, I can’t say anything specific due to the missing load cases.
Sometimes it happens, that a load case without internal pressure leads to higher stresses, than the same load case considering internal pressure. Imagine a primary load, which is working in axial direction at the nozzle, towards the intersection. In this case it can happen, that internal pressure reduces stresses. Therefore ROHR2 always checks both cases and shows the worst case, like this it can happen, that the GUI shows “0bars”, as this leads to higher stresses.
Stress categorization should be done acc. to the code, in EN 13445 (ref. to annex C, Table C-2) eg: primary membrane + bending stresses can be evaluated against 1,5f, as well as local primary memrane stresses.
Regarding Question 2:
Have a look at C.5-2 of the EN 13445, there is written:

C5-2.png

For the determination of ln and ls ROHR2 calculates the length of the elements of the intersection, which have stresses > 1,1f, on both run and nozzle. This length is compared against ln + ls as shown above, ref. to C.7.2 of the EN 13445.
I hope this helps, if you need further information, please contact out support, please also note, that there we can answer much faster (within the same day, normally <4h ) than in the ROHR2board.

Best regards