Stress analysis of thread

[Anonymous]

Hi everyone, I have 2 part, they're connected together by thread. I had model thread for 2 part (thread geometry), now I want to caculate stress for 2 part, with water internal pressure is 150bar. But I can't assembly thread for 2 part and how to caculate stress for thread of 2 part? Can you help me?

I have attached two part below,

 

 

 

  

[John_Holtz]

Hi @Anonymous 

There are several things that you did not mention.

1. What software are you using? (Just checking that you are using Inventor Nastran 😀.)
2. What version are you using?
3. What problem are you having with the analysis?
4. Why are you modeling the threads????

Threads require a very fine mesh to capture the stress concentrations. A fine mesh leads to a lot of contact. Both of those lead to a very long run time for the analysis. The usual method of analyzing threads is to leave them out of the analysis! Model the "threads" as a smooth pipe at some average diameter and bond them together. Based on the reaction (which you know already, but in case you want to see how it varies along the length), you can perform a hand calculation for the threads. Perhaps if you have a nonstandard thread, or have reason to believe the joint is behaving differently than the hand calculations, then modeling the actual threads would be justified.

I suggest that you model 1/4 of the full 360 degrees since everything is 99% axisymmetric. In fact, you should use this article to model a much smaller portion than 90 degrees, such as a 10 degree segment. See How to represent 2D axisymmetric elements in Nastran In-CAD.

[Anonymous]

Hi @John_Holtz 

I'm using Inventor sofware version 2019, and use Nastran In Cad sofware version 2019 to analysic. 

I want to analysic the thread in the file I sent above. I have a problem, I don't know contact for the thread, when I contact bonding for the thread, I run analysic, it's failed.

I'm modeling the thread because I want to analysic stress on the thread.

I read the way you reply, but how to contact the model ¼ of full the 360 degrees. And how to apply force on the model?

I will try reading your reply and the link you sent several time to understand better.

Thank you for your reply

[John_Holtz]

Hi @Anonymous 

Your Inventor files attached to your first post do not have any setup for Nastran. So we cannot run the analysis and see why it is failing. You can either upload the latest saved files (which should have the Nastran input), or you can tell us what the failure was during the analysis. (Ran out of disk space? Fatal error E5000, E5001, or E5004 because your model is not statically stable? and so on.) It is unlikely the analysis failed because the threads are bonded.

Whether the model is a full 360 degrees or 90 degrees or some other angle does not change how the load is applied. A 150 bar internal pressure is still applied as a pressure of 150 bar.

How the contact is defined should not change either, although there may be multiple surfaces when using a symmetric model compared to a full model with one thread. (This detail depends on how the CAD model is created.) For the contact, you should use "Contacts > Solver" and set a Max Activation Distance slightly larger than the gap between the faces in contact but smaller than the gap between the faces that are not in contact. Let the software determine what nodes are in contact with what element faces instead of trying to specify the faces directly.