Hi all,

I am being asked to estimate the hour required for some stress systems Packagewise. Among them some packages are for pump and AFC(Air Fin Cooler) system. Our client taking one system for pump suction and discharge lines.
But normal practice we analyse those system separately( to avoid complication) and check the nozzle load from both system by API-610(cen. pump). Once i breake the system and calculate the hour for that, client ask me to give the technical explanations for breaking those pump system...

Is there any idea to solve this issue?
I have made a write-up for that.pls make me correct if i am wrong....

To analyze a piping system we need to set a boundary condition at the terminal point by putting a physical/virtual support. When the terminal point is a Nozzle, we put an anchor at the nozzle point along with its thermal movements generated by the equipment itself.

We have two procedures for that:-

1. Calculate the thermal growth of the nozzle point from equipment geometry and design parameter and then impose the displacement data in the piping system (i.e. in CAESAR-II) on that point.
(Normally used where equipment geometry is complex in nature)

2. Model the equipment with rigid option (because equipment flexibility compare to piping is negligible) in CAESAR-II with supporting arrangement and design parameter per equipment drawing. Then thermal growth will be calculated by CAESAR-II itself per equipment geometry.

For (Pump) inlet (suction) and outlet (discharge) are connected through equipment. If we analyze it as a single system, then the load from discharge and suction will pass through the equipment. But in CAESAR-II software these are not properly simulated and it gives erratic results. But if the system is divided (suction & discharge) results are more practical and authentic.

On the other hand as exact modeling of the equipment is not possible in the software. So we need to simulate the logical and realistic condition and connectivity of Inlet/Outlet systems may become an issue.

Air Fin Cooler system philosophy is similar.

Thanks in advance,

If, when coding through the equipment, you specify an anchor for that piece of equipment, I see no reason for the combined systems to produce results that differ from your separate analyses.

Yes, it is difficult to exactly model the structural response of equipment. But, usually for thermal strain analyses, modelling these components as stiffer - rigid - elements is conservative.