Design of heads with FEA (ASME VIII-2) 1

[julian89]

Good day gents,

I about to embark on something I have never done before, which is design of a semielliptic head (2:1) via FEAs (I have only used DBF before). The head is to be designed for internal pressure and external pressure. There are a number of failure modes to investigate in both the knuckle and crown; plastic buckling, elastic buckling and yielding.

The purpose of this post is simply to inquire if you guys (the experts) have any general advice for assessing these failure modes via FEAs. Simply follow the steps of Sec. 5? Is a limit-load analysis suitable? How to assess buckling in the 'best' way?

Thanks,

Julian

 

[TGS4]

Use the elastic plastic analysis method for demonstrating Protection Against Plastic Collapse. Use the Type 3 method for buckling, simulating imperfections with the first eigen buckling Mode scaled by the magnitude of the allowable fabrication imperfection.

 

[julian89]

@TGS4, I was planning on doing a limit-load analysis for plastic collapse and Type 1 for buckling. Why do you propose Elastic plastic and type 3 instead?

Thanks.

 

[TGS4]

If your thickness is low, then you will find yourself in a buckling situation, so limit load will not help you detect failure there. I pretty much never use limit load.

Type 1 has been shown to be unconservative.

 

[julian89]

@TGS4, thank you sir. Do you know of any resources where I can learn more about performing the FEA as you suggest?

Julian

 

[TGS4]

Unfortunately, not yet. The Code writers are developing white papers about this methodology. In the meantime, I will make myself available as a resource.

 

[DriveMeNuts]

TGS4,
I have been working with Ellipsoidal heads recently, but just using elastic analysis. My learning suggests that thinner heads under internal pressure will fail by buckling of the outer surface of the knuckle before any other failure mode.
I haven't worked with thick heads yet however I'm suspecting that they fail at the inner surface of the knuckle. For an elastic analysis, the bending stresses in this region are considered secondary. As a result, I would expect excessive plastic deformation of the inner knuckle surface to be the initial cause of failure followed by plastic collapse. Does this make sense?

 

[TGS4]

Buckling doesn't occur at an inner or outer surface - either the entire section buckles or it doesn't.

Classifying and categorizing the stresses in such a location may not be as clear-cut as you would think. I would steer away from purely elastic analysis in these situations. And always consider buckling.