Modeling an already installed system

[NovaStark]

Hi all,

I have a some amount of experience using Caesar however I am not an expert by any means. I have a task to do which involves modeling an already installed pipe system and then seeing if the support system can handle a heavier component such as a valve.

Now my system is such that the piping is connected to a steam vessel (no thermal growths given by the vendor for this vessel) and then goes to a a 40 ft section on a pipe rack.

Now I have drawn out the system such that I anchored at the nozzle for now and went to the header however only to the first support along that header (otherwise I'd be modeling an entire piping run). I left this end with a +Y support as it's on a beam and has a guide support as well.

My main problem is that there is an already installed spring hanger. Using the program I let Caesar design the spring for me. Now the problem (if it is an actual problem) is that the spring hanger that was selected doesn't match what is on the installed spring's drawings. For example, CII selected a hanger with a spring constant of 1500 lb/in with a travel around -0.2 inches (cold load about 2500 lb) whilst what is actually on the drawing is a spring constant of 380 lb/in with a travel of about -0.7 inches (cold load around 4100 lb.). (This is modeling the system as actually installed). Now I am not sure as to if the original designer over designed or something.

Now I am pretty sure that my model is somewhat wrong as my springs don't match up. Can anyone assist me in my task or at least lend some advice on how to accomplish this ? I can upload the model if anyone is able to assist.

 

[DSB123]

NovaStark,
Either:
1. The original analysis is incorrect
or
2. Your analysis is incorrect.

If you have only modelled the pipe from the anchor to the first support then it sounds like your model is insufficient for obtaining accurate results.
Post a sketch to let people provide a better assessment

 

[NovaStark]

Hi,

Thanks for the reply. I have uploaded the drawings which I modelled from the anchor on the first sheet to the drawn in 10-20 ft length on the second sheet. I added in most of the details on the drawing as well for the supports.

One thing I believe I have done wrong from further research is that in C2 I put my spring variability as 25% whilst the spring has a 6% variability. So that may be the error there.

Another problem from the model is that the highest stressed points are where the trunnion supports join to the main pipe run (on the first page). The trunnions are made up of 30 inch long pipes. However since they aren't intersections, I am not sure if I need to model SIFs for them - One of the trunnions has a re-pad on them and a stiffener bar locaed within the trunnion. Not sure if I modelled this correctly or if I should manually find the second moment of area and then put a pipe which would have an equivalent second moment of area here.

 

[NovaStark]

I'll also upload the C2 model when I get back access to it just so that someone a bit more experience than me can review it.


EDIT: C2 file uploaded.

 

[NovaStark]

I guess the file didn't upload the last time.

 

[DSB123]

you need to apply SIF's at the Trunnions. All items welded to the pipe introduce Stress Intensification.

 

[NovaStark]

you need to apply SIF's at the Trunnions. All items welded to the pipe introduce Stress Intensification.

I did run with the additional SIF at the bottom trunnion however that just managed to fail the stress analysis while the spring being chosen was not the one installed. So C2 will suggest a spring with a constant of 761 lbf/in with a movement of -0.214 inches whilst the installed spring has a design of 380.5 lbf/in with a movement of -0.71 inches (i.e. long range spring).

 

[saplanti]

To go forward for the application you have I would suggest you to model the original piping with original valve to see the model matching the working piping for the verification of the model. Since the piping has worked so many years you may accept the original piping arrangement with supports is more or less correct unless you observed something different during operation (I would check this with operator). I would use the existing spring hanger (spring rate and cold installation load) in the model, and probably verify the spring hanger by checking the template on it and the displacement on the actual spring in hot condition. All the support should be verified at site against isometrics.
I would not forget to model the nozzle movement on the steam vessel that you described. There must be some kind of movement there and it has to be included. The nozzle allowable is another to check on the analyses.

In the new model I would add the the intentional new valve and re-analyse the piping with the existing spring hanger to see the system working without problem. If there are problems with the systems then look for solutions such as replacing the spring hanger with the new selected by the software. Selection of the location of the new valve ( if not the same as original or with different weight and different dynamic behaviour) should be done intelligently, it will have impact on the stress analysis and the piping behaviour. It should satisfy the process engineer as well as the stress analysis.

 

[NovaStark]

To go forward for the application you have I would suggest you to model the original piping with original valve to see the model matching the working piping for the verification of the model. Since the piping has worked so many years you may accept the original piping arrangement with supports is more or less correct unless you observed something different during operation (I would check this with operator). I would use the existing spring hanger (spring rate and cold installation load) in the model, and probably verify the spring hanger by checking the template on it and the displacement on the actual spring in hot condition. All the support should be verified at site against isometrics.
I would not forget to model the nozzle movement on the steam vessel that you described. There must be some kind of movement there and it has to be included. The nozzle allowable is another to check on the analyses.

In the new model I would add the the intentional new valve and re-analyse the piping with the existing spring hanger to see the system working without problem. If there are problems with the systems then look for solutions such as replacing the spring hanger with the new selected by the software. Selection of the location of the new valve ( if not the same as original or with different weight and different dynamic behaviour) should be done intelligently, it will have impact on the stress analysis and the piping behaviour. It should satisfy the process engineer as well as the stress analysis.

The system I modeled would be the original set-up with the original valve. I was doing this to check the model for accuracy in terms of finding the hot load as indicated on the original hanger. Like you said, what I may need to do is model the vessel in C2.

 

[saplanti]

It seems that you are at the beginning. Follow all the steps required above to make sure the model is adequate for the purpose. Unfortunately I do not have CAESAR II with me and cannot check your model. I suggest you to verify all the input against the isometric, and do not forget to add the insulation and cladding weight, it seems there is an involvement with high temperature.

 

[NovaStark]

It seems that you are at the beginning. Follow all the steps required above to make sure the model is adequate for the purpose. Unfortunately I do not have CAESAR II with me and cannot check your model. I suggest you to verify all the input against the isometric, and do not forget to add the insulation and cladding weight, it seems there is an involvement with high temperature.

Yeah I figured that I'd need to start over as I realized that I would need to model the vessel in C2 as the drawing for the vessel neither specified allowable nozzle loads or even the thermal growth associated with the vessel. However, I did initially model the line as given on all the isometrics.

 

[NovaStark]

Well I updated the model as best as I could, and while my loads did increase to select the correct spring stiffness, the hot/cold loads are still ways off.

 

[saplanti]

You need to use the spring stiffness and the cold installation load of the original spring hanger in your analysis, do not let CAESAR II to select the hanger. If the system complies with the code requirements, and the displacement are within the original design so your model is OK. Then you replace the valve and new routing , if available, in your model and run the analysis with the existing spring hanger. If everything seem OK you finish the job.

If there is a problem, and the source of the problem is the spring hanger only let CAESAR II select the spring hanger for you. Sometimes with addition of new mass into the system you may require to add supports around the mass. If the nearest support is the spring hanger so you may change the spring hanger or add another one. These sometimes do not work as suggested, you may try many combinations of the trials to make it works, the trials include the relocation of valve and the change in to piping route.

Unfortunately, we cannot provide the magic answer, you need to work on it systematically, and sometimes take notes on your trials not to get in to endless cycle.

Hope it helps.

 

[NovaStark]

Thanks. I'll keep that in mind.