Non uniform wall thickness.

[CADCAMTech]

When calculating the approximate hydrostatic burst pressure of a part that has a non uniform wall thickness what is the proper way to go about doing the calculation? Can I just take the wall thickness at the thinnest section and apply the formulas for thin walled pressure vessels, or is this methodology too simple? I have attached a sample part, the thinnest wall is 0.082".

 

[JStephen]

First off, can't open the attachment, you might try saving as a pdf or something.
The code is set up to design vessels, not so much to evaluate them, so if you apply code provisions and it turns out the item in question is actually twice as strong as it needs to be, that's okay per code, but not much help in evaluation.
With that in mind, yes, taking the wall section at the thinnest point and applying the formulas is about all you can do for simple solutions.
If the application is appropriate, using finite element methods (with a very knowledgeable user!) would get you a lot closer, as would testing.
Also, there is some data available for evaluating local thin spots in heads in ASME, and perhaps similar info for evaluating corroded areas in existing vessels that might be of use if the geometry was close.

 

[CADCAMTech]

JStephen,

The attachment is a Autodesk Inventor cad file so you should be able to open it in most cad packages, but i will upload a 3D pdf or just a regular PDF later. This is for a piece of down hole oilfield equipment so the ASME code does not apply as far as I am aware. As for FEA, I have access to Autodesk Nastran In-CAD but I am far from a knowledgeable user and we are a small company so unfortunately I have the most experience with it and that is only based on what I took in school. I was planning on posting questions about this in the FEA section of these forums. I have done a hand calculation using the assumptions above and I get a burst pressure of approx 12000 psi. We have a pump setup that is capable of this but I wanted to make sure I had a reasonable burst pressure calculation before I did any testing.