FEA Stress at Repad Edges

[pbc825]

I'm working on an API 620 tank and I'm using FEA with plate elements to design the tank. I'm hoping to solicit some information on classification of stress at discontinuities considering ASME Section VIII Division 2 Part 5.

I have modeled a nozzle, added a repad, and when total stress is shown (membrane plus bending) the shell is highly stressed at the edge of the repad from external nozzle loads. The stress is higher than 1.5*Sm but lower than 3*Sm. The repad isn't "discontinuous" with respect to the rest of the shell, but it does change thickness which could be considered a discontinuity (in my opinion).

So the question is, should one consider the shell at the edge of a repad as a "gross structural discontinuity" as referenced in ASME Section VIII Division 2 Part 5 subject to the acceptance criteria of Sps (typically 3*Sm), or should it not be considered a discontinuity and subject to acceptance criteria of PL+Pb < 1.5*Sm?

Thank you in advance.

 

[TGS4]

The bending stresses at the edge of a repad are secondary and not primary (primary bending is essentially restricted to flat plates).

Note that when evaluating Protection Against Failure from Cyclic Loading: Ratcheting, you are comparing Sps to the calculated stress RANGE from the operating load ranges (as distinct from those from the Design Load Case Combinations). Also, when calculating P+Q, make sure you are complying with 5-A.4.1.2 for calculating the equivalent stress range.

You will also be required to achieve comparing Pl to Spl at the edge of the repad.

 

[pbc825]

Thank you for your information TGS4. Quite helpful.

Do you care to elaborate on defining a stress range? For example, if external nozzle loads were reversed, would the stress range be considered to be reversed (i.e. would I have + 260 MPa and then -260 MPa for a range of 520 MPa)? Or are only absolute stress values relevant?

Maybe the better question is, am I being conservative comparing equivalent stress subject to design loads to Sps at discontinuities when displaying total stress with plate elements using linear-elastic material models? Or am I off base? I'm most interested in reducing complexity in favour of being conservative as the tanks are exceptionally complex.

I read 5-A.4.1.2 thoroughly. I understand it's relevant to three dimensional elements. I'm using two dimensional elements capable of displaying membrane, bending or total stress.

 

[TGS4]

For example, if external nozzle loads were reversed, would the stress range be considered to be reversed (i.e. would I have + 260 MPa and then -260 MPa for a range of 520 MPa)?

This is the correct way. You are interested in the algebraic difference.

Maybe the better question is, am I being conservative comparing equivalent stress subject to design loads to Sps at discontinuities when displaying total stress with plate elements using linear-elastic material models? Or am I off base? I'm most interested in reducing complexity in favour of being conservative as the tanks are exceptionally complex.

The proper way is to use the operating load ranges.