FEA Nozzle analysis on a straight pipe "Boundary Conditions"

[jkal321]

You may see my question posted twice since I had it in the wrong subject category. So hopefully I will receive a much friendlier reply.

My co-worker and I have performed an FEA analysis on a nozzle on a straight pipe using different pressure and temperature load conditions. I also do a lot of Pips stress analysis with CAESAR II.

The boundary conditions that we applied were both ends of Run pipe are fixed, taking into account the length of the run pipe to be (4) times the diameter of the run pipe to assure the stress distribution near the nozzle is not affected or corrupted by the fixed boundary condition of ends of the run pipe. The approach is conservative and obviously the thermal stresses are high. Well we were told to free one end and let it free to expand and literally eliminate all thermal stresses.

Is his decision correct and was our approach in analyzing this problem off with fixing both ends?

My experience in pipes stress analysis although the piping system consists of numerous expansion joints to accommodate the thermal expansion but in this FEA analysis I would have thought the thermal expansion should be considered. It has been always a challenge for me to determine the boundary conditions in an FEA problem.

Thanks

 

[rb1957]

your boundary conditions need to reflect the real world. in the real world, will both ends of the beam be fixed ?
if not, then you are building alot of constraint into the model and you'll get unrealistic answers.

the pipes are under pressure ? then fixing, or pinning, every point on the pipe wall over constrains the pipe (it wants to expand under pressure (and your constraint is fighting this natural behaviour).

depending on your realistuation, you might constrain one end axially (Z) at all nodes, constrain one node in X and Y, and one node in X. this should be enough for the model to run, then look at the deflected shape ... is this real ?



Quando Omni Flunkus Moritati

 

[corus]

By assuming fixed ends you're assuming the vessel wall is infinitely stiff so as far as the pipe is concerned your assumptions are grossly conservative. However, if it satisfies design code limits then there's no problem. Obviously if there are thermal expansion joints in the piping system then you could assume free thermal expansion along its length, however that won't be so conservative as the thermal expansion joints will have some stiffness that should be included. Your assumption assumes the very worst.

 

[jkal321]

Understood. Thank you for the replies and clarification.