Questions Regarding ASME VIII-2 Part 5 1

[IdanPV]

Hello All,

Few questions regarding the use of non-linear FEA and ASME VIII-2 Part 5.
1. Is there any mesh setings requirments for the elastic-plastic method?
I know that for the elastic analysis, you must have enough elements through the thickness in order to get an accurate SCLs.
But, what about the elastic-plastic method?

2. 5.5.4 Fatigue Assessment - Elastic-Plastic Stress Analysis.
From what I udnerstand, the "twice yield method" is the better way for the fatigue assemssment. But I didn't understand how to create the cyclic stress-strain curve for the analysis. I've read the PTB-3 example, but it wasn't so clear.
recommendation and\or examples will be great.

3. In PTB-3 Example E5.3.3, a full model contour plot of the strain limit is demonstrated.
I didn't understand the meaning of this strain plot.

Thank you very much for the help

 

[TGS4]

1. Your solution ought to be independent of the mesh. You're the engineer - you need to figure out what's adequate.

2. See the references to Kalnins in PTB-1.

3. I don't understand your question. Please elaborate.

 

[IdanPV]

Hi TGS4,
Thank you for your help and answers.

1. OK, thank you.

2. Which one:
[33] Kalnins, A., “Fatigue Analysis of Pressure Vessels with Twice-Yield Plastic FEA,” ASME PVP Vol.
419, Pressure Vessel and Piping Codes and Standards, 2001, pages 43-52.
OR
[34] Kalnins, A., “Fatigue Analysis in Pressure Vessel Design by Local Strain Approach: Methods and
Software Requirements”, ASME, Journal of Pressure Vessel Technology, Vol. 128, February,
2006, pages 2-7.

3.
A. Figure E5.3.3-1 show the contour plot of the strain limit, what is the meaning of the plot?
B. The calculated strain limit is 0.1515, and this strain limit is for the entire model (am I right?) so, why the plot in Figure E5.3.3-1 is capped at 8.04e+00?

 

[TGS4]

Ref 33

The limit is calculated at each point, because it is s function of the state of hydrostatic stress. That is why it is best to plot the ratio of the plastic strain to the strain limit, a quantity I call the strain damage limit ratio (SLDR). The instantaneous value of the equivalent plastic strain or strain limit are irrelevant without the other. PTB-3 plotted both to demonstrate the concept.

 

[mskds545]

IdanPV - Just last week on the FEA forum I explained to you that the strain limit should be evaluated at each point in the model and that this is best done with a custom contour plot. Now you suggest again that there is a single limit for the entire model. It says right in the text of PTB-3 that 0.1515 is for a single integration point and that each point is evaluated. The limit varies throughout the model from 0.0309 to 8.04. As I said in the other thread, I recommend you take a training course in design-by-analysis and find a good mentor.

-mskds545

 

[IdanPV]

mskds545-

I didn't mean to appeal your answer, I am sorry.
I just wanted to get a better understating of how things work in the Elastic-Plastic Method.
(Something I didn't manage to understand, yet...)

It is clear to me that the strain limit should be evaluated at each point in the model.
But I don't understand how to create the custom contour plot (using SolidWorks)
I didn't suggest that there is a single limit for the entire model, and I know that it could not happen.

PTB-3 stated that: "The following example calculation is for one integration point in the model.
The full model (all integration points) will be evaluated using the “user defined variable” option available in ABAQUS to create contour
plot variables of the strain limit and strain limit ratio."
1. Is that means that the author of example E5.3.3 checked each point in the model in order to create a user defined contour plot
2. Is the "user defined variable" to create contour plots is something that only available with ABAQUS?

As for taking a training course in design-by-analysis, I will do it in the future I guess.

TGS4-
Thanks, but it seems that this article is very old, anyway I can buy it over the internet?

 

[TGS4]

Try the ASME digital collection. If it's not there, there are other papers that are case studies of that method. Search for Kalnins Twice Yield.

 

[IdanPV]

TGS4,

I tried to find it in the ASME digital collection, but it isn't there.
I am trying to understand the background behind this method, do you have any reading recommendations?

Anyway, I have one more question regarding Table 3-D.2 "Cyclic Stress-Strain Curve Data"
The Table shows the values for 304 but values for others stainless steels are missing.
I am trying to create a curve for 316L stainless steel. what do you think will be the best approach?

 

[TGS4]

Test data

 

[IdanPV]

TGS4,

Thank you.
What is the procedure to determine these values for other materials?
What exactly a "Test Data" means?

 

[DriveMeNuts]

IdanPV, your knowledge on this subject seems good and you look like TGS4 doing a good job of filling in a few gaps.

I went on an advanced ASME pressure vessels FEM day course a couple of years ago and found that I knew 90% of the course content and that my questions to the teacher became discussions where both myself and the teacher learnt together.

It was very worthwhile even though I didn't learn anything completely new. If I was to go on another course I would research the syllabus thoroughly to make sure I know I am going to get the most out of it. Perhaps even e-mail the instructor to ask to include a specific topic.

Something to keep in mind.

 

[TGS4]

Unfortunately, I do not know the procedure for obtaining a stabilized cyclic stress-cyclic strain curve from test data. You would need to search the literature for that information.

 

[IdanPV]

DriveMeNuts,
Thank you for your comment, I will definitely search for the syllabus before the course.

TGS4
Thank you.

 

[IdanPV]

Hello All.

I did some digging into this topic, inculding reading some articles and do the PTB-3 example by myself.
Things are more clear to me now.

But still, I have few more questions:
1. When designing a pressure vessel for a Full Vacuum service using 5.2.4 (Elastic-Plastic) an atmospheric pressure is applied to the outside surfaces of the vessel.
Does the pressure need to multyplied by the ß factor in accordance with Table 5.5 for the global critirea?

2. As I've already wrote, Table 3-D.2 "Cyclic Stress-Strain Curve Data" contain data for some materials, but SA240-316L is missing.
As TGS4 said, the value for materials which are not include in the table need to be obtain from "test data" but I didn't found any thing regarding to this approach.
Can someone please share his knowledge regarding this issue?

In the meanwhile, I tried to create an approximation for the SA240-316L which is based on the yield strength ratio: Sy(304)/Sy(316L) (SA240-304L is include in Table 3-D.2)
Please, refer to the attached file and let me know what you think about this approach.
I know this is not a "test data" but the Code doesn't give any rule or procedure how to get the right data.

 

[DriveMeNuts]

1. Sounds like buckling. Don't you mean 5.4.1.2 Type 3?

 

[DMC05]

2. As I've already wrote, Table 3-D.2 "Cyclic Stress-Strain Curve Data" contain data for some materials, but SA240-316L is missing.
As TGS4 said, the value for materials which are not include in the table need to be obtain from "test data" but I didn't found any thing regarding to this approach.
Can someone please share his knowledge regarding this issue?

A very useful reference is Hales, R., Holdsworth, R., O'Donnell, M.P., Perrin, I.J., and Skelton, R.P., "A Code of Practice for Determination of Cyclic Stress-Strain Data", Materials at High Temperatures 19(4), pp. 165-185, 2002.

For an introductory overview of this topic, you should probably start with Chapter 3 of John Draper's "Modern Metal Fatigue Analysis" textbook.

The above references can be found with a simple google search. There is an abundance of information on this topic published online within the technical literature, a lot of which is freely available.

Good luck!

 

[IdanPV]

DriveMeNuts,
This is a non-circular pressure vessel with external reinforcements and an aspect ratio of somewhere around 2.
I don't think that buckling is a failure mode in this case.
But full vacuum is one of the opertaion mode of this pressure vessel, so in accordance UG-22(a) I shall consider it.

DMC05,
Thank you for the valuable information.