Appendix 46 or Fea for Nozzle 1

[Matrix_93]

Hello,
I have a nozzle with high loads that doesn't respect the ratio Rn/R and I need to follow U-2(g).
In this case I have two options:
Follow appendix 46 but in this case I need to use a very thick nozzle in order to verify primary stress;
Follow ASME VIII div.1 with a thinner nozzle neck thickness and to do a fea analysis with nozzle pro.
Which is the best solution?
With Appendix 46 I need a nozzle neck with 2" thickness; with FEA I need 1" only. We can't use pads or self reinforced nozzles.

 

[DriveMeNuts]

For both options, you need to follow Appendix 46.

 

[SnTMan]

Which is the best solution?

Which one is best for you?

We can't use pads or self reinforced nozzles.

What's left?

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[TGS4]

Ok - your options are not making sense to me. When you say Appendix 46, are you meaning that you are using Appendix 26 to go to VIII-2, Part 4? And when you say FEA, are you meaning that you are using Appendix 46 to go to VIII-2, Part 5? What methods are you using (from the various Part 5 methods) for satisfying ALL of the failure modes listed in 46-4(c)?

 

[Matrix_93]

I prefer to avoid appendix 46 and asme viii div 2. In the past, when we needed to follow U-2(g) I always used Fea analysis (nozzle pro for nozzles) to resolve the calculation. I checked the opening in div.1 and with fea analysis with allowables of div 1 I would check the nozzle.

I have seen with appendix 46 we would have a thicker nozzle, this is the reason of my doubt.

 

[TGS4]

You are not making any sense. With the current Edition, if you want to do FEA in VIII-1, you are obligated to follow the rules in 46-4. That is mandatory, even from U-2(g). This isn't being pedantic, you need to follow the rules as-written.

 

[SnTMan]

TGS4, I'm curious, and I understand you may not know, but, Nozzle Pro has been around for some time. Does it comply? If so it would seem the OP's path is clear...

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[saplanti]

Do not underestimate the given nozzle loads, it is a contract between pressure vessel supplier and piping designer. This is the only way for the EPC companies can manage and meet the schedule of the projects. Additionally the loads are not fictitious if you can see the entire project.

As the pressure vessel manufacturer you are obliged to comply with specification. Under the contract you are supposed to see what’s coming and tune up your quotation. Of course sometimes you need to comprimise on the vessel thickness to be able to take the nozzle piping loads. This was supposed to be known by yourself before quoting on the project. You probably did design the vessel for internal pressure only, assumed the nozzle load can be managed later. This was the big mistake and you are trying to get away from it. If you worked under the requirement of PD 5500 code you could see this was coming for the nozzle reinforcement pressure design not for external loads only.

Please note that those piping loads shall be taken by the foundation as well as the vessel supports with combinations to worst case scenarios.

Hope you can find a way not to increase the cost too much on your side.

 

[mechengineer]

FEA shall use only for a local stresses analysis due to the nozzle external loading and the internal pressure, rether than the opening reinforcement requirement by code.
For nozzle reinforement requirement you shall use Division 2, 4.5 (a pressure area method) which is without the ratio Rn/R limit whatever you use FEA or not.
The correct way is that use Division 2, 4.5 for opening reinforcement and NozzlePro for nozzles external loads. ASME VIII does not provide the nozzle external ananlysis method.
See the relevant threads below for nozzle reinforcement-nozzle load-thrust load,

https://www.eng-tips.com/viewthread.cfm?qid=477061

https://www.eng-tips.com/viewthread.cfm?qid=456579

Regards,

 

[DriveMeNuts]

SnTMan,
It is the users responibility to audit and accept any specific software.
That goes for ANSYS, NozzlePro or Microsoft Excel.

I'm regularly e-mailing Codeware Compress about errors.

 

[SnTMan]

DriveMeNuts, yeah I understand that. I am not a NozzPro user. I know it was in wide use since before what I understand are very specific requirements for FEA analysis under U-2(g), Sec VIII, Div 1 were implemented in the 2019 Ed.

I am simply curious to know if NozzPro was / is compliant with the same.

My suspicion is that the NozzPro people are on the Code committees and had input shall we say

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[DriveMeNuts]

I'm not aware of any specific requirements.
My view of Appendix 46 is that it hasn't added any requirements. It simply keeps the engineer on the correct path.
Previously, an engineer could use U-2(g) to go off and manipulate the code into dangerous territory that it was not intended for.

My experience of NozzlePro is that it has always produced good results when used "correctly" (in accordance with Appendix 46 and Div 2 part 5).

 

[SnTMan]

Was looking at TGS4's post: "With the current Edition, if you want to do FEA in VIII-1, you are obligated to follow the rules in 46-4. That is mandatory, even from U-2(g)."

I am totally unschooled the subject, and so I wondered if NozzPro does so (meets 46-4). Note it is not my intention to case aspersions regarding NozzPro...

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[TGS4]

Paulin Research Group (the makers of NozzlePro) claim that the software is in compliance with the Code. I have not independently assessed that claim, but I have no reason to doubt this claim. Nevertheless, it is the design engineer's responsibility (the software user's responsibility) that any software that they use is fit for purpose.

That said, it is my opinion that of an engineer is not sufficiently versed in the VIII-2, Part 5 methods and rules, then they are incapable of making such a determination for any software, even one that claims to do most of the work for them (black box). Unskilled software users are dangerous.

 

[SnTMan]

Thx

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[DriveMeNuts]

The weird thing I found abut NozzlePro is that where a nozzle has say a 10mm fillet weld, the FEA appears to analyse the nozzle without any fillet weld and then simply ignore the stresses within 10mm of the FEM mesh junction. So the location of the reported shell stress is 10mm away from the mesh junction.
Seems like a bit of a bodgy approximation to me. There is nowhere in the manual explaining what is going on here as well.
It also allows an opening consisting of a chunky set-in studded flange to be meshed like a regular nozzle. A set in flange consisting of phenomenally thick and stubbly shell elements just don't look right even if it does comply with the r/t > 4 rule.

 

[TGS4]

DriveMeNuts that approach with the fillet welds is pretty standard. I have no issues with that.

Chunky set-in studded flanges meshed with shell elements is indeed a concern. Follow-through with that concern.

 

[BJI]

It actually analyses the stress half the nozzle thickness plus the weld leg size away from the junction, the stresses within that region aren't really valid. DriveMeNuts, you should review the 3D Stress Criteria (WRC 429) to better understand how the stress classification locations relate to the failure modes being assessed in VIII-2 Part 5.

My verification of NozzlePro years ago showed that is wasn't compliant with the code but I have not used it in some time. From memory the allowable in the nozzle wall was not correct for longitudinal stresses caused by external loads and it mixed codes, using Section III in some cases..

 

[SnTMan]

What I like about FEA

The problem with sloppy work is that the supply FAR EXCEEDS the demand