Elastic-Plastic Stress Analysis Method in ASME VIII.2

[IdanPV]

Hello All,

I am trying to understand the Assessment Procedure and the Acceptance Criteria for the Elastic-Plastic Stress Analysis Method.
I've already create the Stress-Strain curve, based on ASME VIII.2 Annex 3-D and I know how to define the material in the FEA software.

What I don't understand is how to know if my design meets the requirements of the Code.
5.2.4.4 stated that: "if convergence is achieved, the component is stable..." but how do I know if convergence is achieved?

PTB-3 Example for this case only shows the von-Mises stress plot and the Equivalent Plastic Strain and stated that "convergence was achieved" so, for me, it's not very helpful.

Thank you very much.

 

[TGS4]

With the factored loads applied, one performs the analysis with the specified materia stress-strain curve. The analysis either converges to a statically-permissible solution or it doesn't.

What software are you using?

 

[IdanPV]

I'm based on Appendix 46-4 for applying the Design By Analysis requirements of Section VIII, Division 2 Part 5 in Section VIII, Division 1 Vessel.
Therefore, as per 46-4(c)(1)(-c) the factor shall be 3.5.

I am using the Non-Linear simulation module of SolidWorks.

I think I didn't understand the meaning of: "The analysis either converges to a statically-permissible solution or it doesn't."

How do I know if this happen or not?

 

[TGS4]

The software will either give you an answer at that load factor, or it won't.

 

[IdanPV]

Do you have an example for it?
I have found this example: Link
but it only stated that: "The analysis converged for the given load and thereby satisfy the global acceptance criteria"

The example only shows Stress and Strain plots, nothing more.

 

[TGS4]

How familiar are you with the concepts of LRFD?

 

[IdanPV]

Not much, but I am not sure I can use it when using Section VIII.1 Appendix 46 which stated that the factor need to be 3.5.

Is there something like the Assessment Procedure of 5.2.2 in 5.2.4?
So I can tell if the stresses is above or below the allowable stress?

 

[TGS4]

No. The software/computer does all of the work for you (hence the brilliance of the E-P method).

Have you run the analysis yet?

 

[IdanPV]

Haven't run it yet. I wanted to make sure I understand what I need to do before I starts.
As I said, I started with define the stress-strain curve, based on Annex 3-D.
Than, I looked at the Assessment Procedure and the Acceptance Criteria for the Elastic-Plastic Stress Analysis Method, but I didn't understand it I guess.

From what I know, convergence mean that there will only small effect on the stress results when the mesh is refined.
But I'm assuming that it is not the meaning when talking about the Elastic-Plastic method..

 

[TGS4]

Do you have a mentor that you can connect with? We are reaching the end of what can be provided, for free, over the internet. I would also encourage you to look into taking a technical training course on the topic.

 

[IdanPV]

Unfortunately I don't have any mentor for this subject.

Do you, or anyone else in this forum have any example and/or FEA report he or she made using the Elastic-Plastic Stress Analysis Method which he or she can share?

PTB-3 and the examples which I found only say that The analysis converged for the given load, but does not provide any other explanation.

 

[TGS4]

Not with your software.

 

[IdanPV]

So maybe with other FEA software?
I just want to know what I need to look for.
Does "convergence" mean there is a solution and the analysis come to an end?

 

[TGS4]

Effectively, yes.

 

[IdanPV]

So, if I am running a FEA study with the correct Stress-Strain for the material under consideration and I get a solution, that's mean the analysis has convergence and the Code requirements have been meet?

 

[TGS4]

Correct

 

[IdanPV]

Thank you very much for your help.

So, let's say I run a FEA study for a shell made of SA-516 Gr. 70, stress has convergence but I get a maximum von-Mises stress of 450N/mm^2.
The stress is much higher then the allowable stress per Section II-D, is it OK?

 

[TGS4]

This would be at the fully-factored loads? Then yes, I would expect that the stresses would be well in excess of the allowable stresses. Potentially by a factor of at least 3.5 (you did mention that you were doing this for a VIII-1 vessel, right?)

 

[Paulettaa]

Hi
Nice discussion.

TGS4, what if I choose an elastic perfectly plastic material model? Is it possible that the results show a Von-Mises stress that exceeds the yield stress?
I understand that the strain can be higher than the yield strain but the stress should not go higher than yield. However, when I run FEM using elastic perfectly plastic model there are points at which the Von-Mises stress is higher than the yield stress that I defined for the material. Is it because of interpolation (or extrapolation) inside elements or is it something about the flow rule?
Actually I could not understand how to assign the flow rule to the software at all ( I use ANSYS).

 

[TGS4]

Yes, it's possible. But with adequate refinement of the mesh, such occurrences should disappear.

The Associated Flow Rule is the default in most major software, including ANSYS.

Are you able to attend my training class next month in Houston?