ASME Sec VIII Div 2 5.2.2 Elastic Stress Analysis Method - ANSYS 1

[JJKi]

Hi,

Bear with me as I am an entry level engineer just starting to get into the woods. I am currently using ANSYS Workbench Mechanical, and I am currently running a static analysis.

In ASME Sec VIII Div 2 5.2.2. it does not specify whether to use small displacement theory (specified in 5.2.3.1 - Limit-Load Analysis) or to allow for nonlinear geometry (5.2.4.2 - Elastic-Plastic Stress Analysis).

1. If I am not completely incorrect, these are both referring to Large Deflection (NLGEOM)?

2. I have heard that I am not to use nonlinear analysis for elastic stress analysis method but does that refer to 'nonlinear geometry'?

3. Even if NLGEOM is off, there are other factors such as nonlinear materials, or nonlinear contacts that turns analysis to nonlinear. When this happens, the elastic stress method is completely off the table?

Thank you,

 

[TGS4]

5.2.2 uses a linear-elastic analysis. That means nlgeom,off. Your material will be linear. You are not restricted from using contact.

Anything else and it is not a liner-elastic analysis.

if you are new to this world, i highly recommend taking a training course on ASME Section VIII, Division 2, Part 5 Design-By-Analysis. of course, these training courses will be "virtual", but they are very effective.

 

[JJKi]

I was really looking forward to the Beckt course coming up locally in May, but it got cancelled due to current events.

Sorry to bug you, but when you're referring to material being linear, do you mean linear elements?

 

[TGS4]

That course didn't get cancelled (I know, I'm the instructor), it just transitioned to being delivered virtually. There is another coming up in November.

When I mean the material is linear, that means that the material has no plasticity. It has nothing to do with the order of the elements.

 

[JJKi]

I meant to say the in-person classes got cancelled, and yes I am aware of who you are! Not to be creepy or anything..

Thank you, I will give it a go.

 

[DriveMeNuts]

As far as I am aware, 5.2.2 is for "elastic" analysis, where elastic is referring to material properties.

It doesn't refer to "linear"-elastic anywhere. Where linear would refer to geometry behaviour (or nlgeom,off)

I have conducted an elastic analysis on some pressure vessels internal structures (i.e. beams and flat plates) and turned large deflections on.

It is a bit pointless applying large deflections to pressure vessel components which don't deflect, as it only reduces stress by about 0.1%.

The only component that I can think of that deflects enough for it to be useful would be a flat head.

Whereas limit load and elastic-plastic analysis sections specify whether to use large deflections, for elastic analysis the decision is left to the engineer to decide.

 

[TGS4]

DriveMeNuts - I wouldn't assume that because the Code is silent on that matter that it is permitted. I would recommend requesting an interpretation on that topic.

 

[mskds545]

When it comes to large displacements and elastic analysis, I've always enjoyed paragraph 5.2.3.2. It's under limit analysis, but applies "equally to... primary stress limits of 5.2.2." This paragraph seems to prohibit elastic analysis if the stiffness is reduced with deformation, a large displacement effect. The stiffness of cylinders and spheres under internal pressure decreases ever so slightly with deformation, since the wall thickness is reduced as the diameter increases. So one could interpret 5.2.3.2 as prohibiting elastic analysis on almost all pressure parts. I don't take this literally, of course, and I don't know anyone who does.

I've been trying to think of an edge case where I had reason to use large displacements on a geometrically stiffening part with an elastic calculation of a pressure part. I haven't been able to think of one. Things like springs, maybe just to avoid coming up with a nonlinear material model when it isn't necessary, but not pressure vessel parts. I might run the analysis both ways to see if there is any significant stiffening or weakening in the part, to help me determine if a more thorough calculation is needed. For the elastic code check, though, I agree that "turn it off" is a very solid rule of thumb even if the code does not explicitly prohibit it.

-mskds545

 

[DriveMeNuts]

msdkds545,
5.2.3.2 makes sense. And it addresses geometric non-linearities with respect to 5.2.2. And while addressing geometry which becomes more flexible as it deflects, places no restrictions on geometry which gets stiffer.

I had a case where the base of a vertical vessel had a cone (without a knuckle) welded to the base of the cylinder. A skirt was then welded to the cone immediately adjacent to the junction. There was a lot of rotation of the joint during the seismic case on the side of the vessel experiencing tensile uplift. I used "large deflections" because the geometry had an increase in stiffness as the cone-cylinder joint rotated. It reduced bending stresses and local failure of the skirt weld.

I suppose using large deflections may have implications on a fatigue analysis.

 

[mskds545]

I didn't say 5.2.3.2 doesn't make sense. I said that, if taken literally, it prohibits elastic analysis of cylinders, since their stiffness decreases with deformation.

For the cone-to-skirt junction, that's between you and the AI, but I probably would have included plasticity. I almost always avoid elastic analysis when checking collapse, though, so I would have been biased toward that from the start.

-mskds545

 

[DriveMeNuts]

mskds545
I hadn't read 5.2.3.2 and was thinking out loud that the clause makes sense.