In theory, when the stress of the pipeline exceeds the yield strength, and then removing any loading , back to the initial loading conditions, the pipe system would happen self-cold-spring.

However, the B31.3 code allows the stress in operation codition exceeds the yield strength ,when the termperature go back to the amibent condition,does the pipe system happen self cold spring?

If so, the spring selection and the nozzle checking would be different from what the program does.
If not, why?
can anyone give me a clear interpreting?

hi,everybody.
can anyone help me?

In most practical cases, I don't think this is worth considering.

You use the term "self cold spring" which is a bit different than the terms I'm used to. You can investigate for yourself the effect of self spring, a.k.a. relaxation, a.k.a. "self cold spring" by adding cut shorts to your model.

Most pipe stress books I've read also cover the question you're asking.

The Codes recognize shakedown or this self-springing as you call it. It's built into the the allowable stress range.
Loads and position (strucural response rather than stress) are not that definitive in the Codes which work with linear response (these Codes are written for slide rules!).
Your initial loads (on the hot side) will be conservative with that linear assumption (no yield) but you will not be able to identify the current loads because of this self-springing. Same with spring positions.
Here's a thought exercise...
If you would open up a flange of an aged and "sprung" system, the faces would take a new postion. This distance between the flange faces is the current "cold spring". If you want, you could model this as a new pipe element, a cut short (or cut long) in your CAESAR II model to estimate the current load and postion of your pipe when you heat it up again.

Dear Dave Diehl.
Thank you for your reply!
IF I understood you correctly, you agree with that when the stress exceeds the yield strength of pipe , the pipe system will occur self-cold-spring after the removal of the temperature load.
As you said, when the pipeline's stress exceeds the yield strength , the forces is nonlinear ,and codes did not say how to evaluate the force and displacement, in order to facilitate the calculation, the software do a linear asumpting .
But here I have a opinion:
I do not deny this treatment is conservative according to codes, However, spring selection is different, If the direction of displacement(causing by the temperature) is a positive displacement, linear treatment will lead to selection of spring's movement is larger than the actual displacement of pipe, but when removing the temperature load,the pipe will have a reverse displacement which is not in the consideration of the selection of the spring, so the spring from the software doesn't work normally in sus condition.
same in the nozzle checking.
we can't know how much does the self cold spring occur,
so we can't know calculate the force for the nozzle in sus condition when the pipeline occur self cold spring.

in the actual calculation we have never considered such a situation ,so i think the self cold spring maybe doesn't occur when the pipeline stress exceeds the yield strength,because only a few points on the pipeline occur yield,it can't not take the whole pipe system to yield.
so if the code stress compliance with the codes, I think There is no need to consider the influence of the self cold spring.
is it right?
Can you give me a more detailed explanation?
thanks.

I believe the Codes do a good job in assuring a safe system that prevents pipe failure. Your point concerns the structural response of the system where yield may occur. That's not the main focus of the piping Codes.
I believe the tolerance in your CAESAR II input and tolerance in the materials and construction are too open to give any confidence to an accurate caluclation of actual pipe performance.
That's why you may run a hanger audit after the system is in operation and that's why you may have to realign troublied equipment.
Analysis can only go so far.
If you want, you can go to a complete, nonlinear finite element analysis to take that next step. But such work, in my opinion, is suited more for forensic rather than design engineering - where you have an existing systemm to match rather than a proposed layout to evaluate.