We are involved in a piping project where some portion of an existing high pressure (1500 Psig) high temperature (1000ºF) steam piping in ASTM A335 P22 material will have to be removed and a heavy weight (20,000 Lbs) new steam control valve, along with new A335 P22 material piping and associated new pipe supports (hangers)and restraints will be welded to the existing remaining piping. The piping has a nominal diameter of 12", heavy wall thickness (1.3" minimum wall) and is in operation for last 15 years, at reasonably steady load.

As the piping is in operation for so many years in the creep zone, it must have self sprung and therefore a considerable portion of expansion stress, equipment terminal reactions and movements have reappeared in the cold condition. The pipe was erected initially without any cold pull up gap.

Can Caesar-II be used to simulate the configuration (that is movements), residual stress and moments and forces in the cold shutdown condition at all node points of the existing piping system? Please advise.

I had initiated this question on 02-21-2011 on an existing thread, but did not receive any reply. So I am re-issuing it as a new thread.

Regards (2011-02-25)
_________________________
kumar

maybe because many of us do not know a 100% right answer.

If i would be in your place, I will try to see if I can cut near an existing anchor point.

read these two, maybe there are other, do a search.

http://65.57.255.42/ubbthreads/ubbthreads.php?ubb=showflat&Number=4353
http://65.57.255.42/ubbthreads/ubbthreads.php?ubb=showflat&Number=34710

Good luck,

One more note, be careful when you cut piping that has self-sprung as large movements of pipe ends may result.

Regards,

If you want to simulate the configuration, displacements, forces / moments, stresses, etc. that may exist in the cold condition after elastic shakedown / yielding, you could model the system normally but for a 1000 F design temp, set the installation temp to 535 F, and run temperature cases such as T1 = 1000 F, T2 = 70 F. The T2 case would give an idea of the possible configuration in the cold condition. Full stress range would be given by T2 - T1 case. This would only be an approximation as Caesar is a linear program and will not model true yielding effects.

Dear Danb,

In this piping system, there are basically two anchors at two extreme ends - one at boiler superheater outlet and the other at steam turbine inlet. The old control valve is located nearer to the turbine, but still about 100 running feet from the turbine end. We are planning to put the inlet and outlet ends of the old control valve on temporary anchors (more than one at each of upstream and downstream sides)and then cut the piping at two cutting planes to remove the old valve. The new valve will be located much closer to the turbine, so we will keep the upstream anchors in position, slowly release the downstream anchors, remove most of the downstream piping, make another cut on the old pipe a few feet downstream from the turbine connection weld, connect new piping and new control valve between the existing pipe at the upstream anchors cutting plane and end of the turbine inlet piping short piece. But it will be necessary to calculate the the reactions at which the temporary anchors will have to be designed for. I am still looking for an answer, not necessarily an exact one, but I would rather like to err on the conservative side.

Regards

Kumar,

I'm afraid that in this case my advices can be randomly wright or wrong. Only someone that actually have done this can agree or disagree with your intentions.

One thing, your anchors need to be designed for full unbalanced loads as you will remove one side that now keep the system balanced.

Good luck,

Dave,

Your suggestion of 535ºF as the installation temperature tells me that you are simulating a 50% cold spring condition.

This piping was erected without any cold spring (no cut short). But after exposure to 1000ºF, the areas of piping which are subject to higher longitudinal stress, will yield permanently and release stress in the operating condition at 1000ºF. On shutdown, the stresses will return in the cold condition with opposite sign.

I am thinking to make a normal stress run with 70ºF as the stress-free installation temperature,and find out the node points where axial stresses at 1000ºF are 50% (an arbitrary number)or more of the allowable stress. These are the node points which will yield under creep condition after long exposure to operating temperature. Then I will make two more runs at 70ºF and 1000ºF, with cut gaps (with proper sign) at these high stress node points, to simulate self-springing scenario.

Regards,

Actually, I was assuming the piping was installed with no cold spring, and that after years of high temperature operation, the piping system may have partially yielded / experienced elastic shakedown, and now has a nuetral / zero expansion stress state at some temperature between initial installation temperature (assumed to be 70 F) and design temperature (1000 F). With no other data, using a straight average you get a new "install" or nuetral temp of 535 F. This gives some idea of the current state of the piping, but there's no real way of knowing it exactly.

Your method may give some ideas too.