Threaded Cylinder With End Caps Subjected To Internal Pressure - Thread Failure

[AFG03082015]

Hello,

I've got a question as to how the threads would fail when considering a cylinder with threaded end caps when subjected to internal pressure (see top drawing on attached pdf)

Would they fail as the second drawing, completely shearing the internal threads of the cylinder (shaded in black)

or as seen in the bottom drawing due to flaring stress (which admittedly I do not have great knowledge of) where the cylinder will splay and as a result a crack propagates from the internal thread root (although stupidly i drew it from the thread crest).

Just trying to get a picture of how such a vessel would fail, all thoughts and discussion are welcome

Thanks in advance

 

[TGS4]

Just to clarify, your title says external pressure but your post says internal pressure (and your drawing implies internal pressure). Which is it?

If it is internal pressure, please red-flag the post and ask the site to change the title.

 

[CoryPad]

Either are possible depending on wall thickness, materials, sealing methods, etc.

 

[EdStainless]

It is nearly always be a hybrid. The end starts expanding, and at some point the remaining threads are too weak and the tips shear off.
On HP systems you see square threads (Acme or modified) used to minimize this.
Some designs also use heavier walls at the ends of the cylinders to reinforce them.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[r6155]

See ASME VIII Div 1 figure UG-34 (q) and table UG-43

Regards
r6155

 

[AFG03082015]

Firstly I'd like to thank all for their input,

TGS4 Thanks for pointing out my mistake, I did mean Internal Pressure and I have subsequently changed the title

EdStainless & CoryPad The cylinders I'm interested in are all specifically metric screw threads, to investigate this further would you say non-linear FEA is the way to go (TGS4 you may be able to enlighten me here) because physical burst testing is not really an option!

r6155 I'll investigate figure UG-34 (q) and table UG-43 and get back to you when I've had a look

Thanks again

 

[TGS4]

I would be cautious with applying non-linear FEA in the context of LRFD. You need to fully understand the physical processes happening when you run those types of simulations.

Also with respect to running an FEA, you should note the caution in 5.5.6.5 in VIII-2 regarding ratcheting in non-integral connections. Even if you can assure yourself that the connection won't burst, you have significantly lower limits with regards to ratcheting.

 

[AFG03082015]

TGS4 Thanks for your reply.

I am thinking about the FEA in support of a ASME VIII Div. 1 vessel (over 3000 PSI). So all of the wall thicknesses, end cap lengths and thread lengths are already calculated as per Div. 1.

When thinking about the supporting FEA I concluded that the suitability of the cylinder (namely the wall thickness) can be assessed using Div. 2 Part 5, 5.2 Protection Against Plastic Collapse.

But I've when thinking about the assessing the suitability of the thread (external thread on end cap and internal thread on cylinder) I've hit a bit of a stumbling block, I cannot determine which methodology in Div. 2 Part 5 is most suitable. I conclude that the most likely form of failure is due to flaring stress and resulting in a combination of a splayed cylinder and sheared thread teeth. Personally I see this as a plastic failure mode, but looking at it deeper the the failure mode is also dependent on cyclic loading (yielding/crack propagation may only occur at maximum working pressure).