Hi all,
This is my first post here.

I want to know if there is anything available that compares the Appendix 2 calculations with an FEA stress analysis?

Thanks

Hi McKenna,

I am having similar problems with PD5500. Traditionally we have designed our vessels by formula, but we have recently started using FEA.

I am trying to work out whether they best way to look at an FEA is to simply look at the equivalent stress results then compare them to the yield stress (or design stress) of the materials.

This post probably was not much use to you - but you are not alone!

Cheers.

Andy.

If you are using FEA under U-2(g), read what U-2(g) says. "...subject to the acceptance ot the Inspector, .... will be as safe as those provided..."

Two conditions:
1. AI doesn't have any issues with it
2. Your FEA demonstrates that your design meets at least what you are required by the rules, i.e. all applicable requirements of 2-8 or other paragraph similar to your actual design (don't forget 2-14). Simple checking Von Mises stresses does not prove that the design is "as safe..."

_________________
Konrad Anikiel

You can find in literature / text books discussion on discontinuity analysis of shell, hub and ring (Process equipment design by Brownell and Young ). However I also experienced in a particular case, few years back in which FEM analysis resulted in very small stress (resulting in low thickness).

The interpretation at that time was the flanges are deflection controlled design and hence conclusion of FEM analysis is not correct. This means the ASME code rules has built in some additional margins so as to control the deformation (from leakage consideration).

I do not have any text book or published paper to confirm the above interpretation. But if it is confirmed, FEM can be used by giving a limit for deformation or a reduced allowable stress.

_________________
R.SRINIVASAN
Sr. VICE PRESIDENT
Head, Design and Engineering
PED,ISGEC,
YAMUNANAGAR
HARYANA -135001
INDIA

This is an interesting discusion indeed.

Let me contribute the Inspector's point of view, using the example of a pressure vessel to Section VIII Division 1, as I assume that the Appendix 2 McKenna is referencing, is from VIII-1.

The Code mandates the use of Appendix 2 calculations for bolted flanges under internal pressure, which are not any standard flanges like B16.5. You can make as many FEMs as you want, as an Inspector I do not have the authority to accept the design without the Appendix 2 calculation.

Only, and I mean ONLY, when the limits of Appendix 2 are exceeded, or the rules are not applicable, other methods can be used. And here comes U-2(g) into play, as Konrad has referenced very nicely.

For example, there is a WE flange made to the dimensions of the DIN flange standards DN100 from ASME material. The loads are internal pressure only. This flange shall be designed to Appendix 2, no doubt.

Another example: A rectangular shaped flange connection with some internal pressure, metal-to metal contact, external loads that dominate the design stresses and cyclic service. You will agree that the Appx 2 rules do not apply. in this case FEM can be used to prove that the stress limits from UG-23 are not exceeded. Do not forget to document the Computer program verification, and the Inspector can maye accept the design.

Clear?
DK

Dirk,
For your rectangular flange, I would still expect to see a similar stress breakdown and stiffness proof as Appendix 2 points out. Of course, the method of calculation would be different from Appendix 2, but if the Code requires the design engineer to check safety of the flange by calculating a defined number of stresses and applying specific acceptance criteria- any other justification of a design has to follow this rule as closely as possible. So, go ahead, do your FEA, but report the same stresses in the same locations as you would have to with usual flange and apply the same acceptance criteria from Appendix 2. These are the minimum safety requirements outlined in the Code and it would be a violation of U-2(g) to use a lower safety standard than this.
Recently I had to reject a design on basis of this argument. A U-certified supplier changed the Code formulas saying that their design slightly differs from what the Code provides for. Unfortunately, their modified formulas effectively decreased the safety level when compared to the original requirements. Which was a violation of the second requirement of the U-2(g). Very often forgotten by manufacturers, and the reviewers as well...

_________________
Konrad Anikiel