I modeled a large 10200 mm (401 in.) ID tower with pipe elements and the nozzle connection to the shell wall with rigid elements in order to obtain nozzle displacements due to temperature changes.
However, for the displacements due to pressure I can only include the axial effects (activate the Bourdon effect; translation only), but for the radial nozzle displacements I see no solution.
I cannot define displacements as the pipe node is connected to the nozzle node of the vessel.

Is the only solution to model the nozzle as an anchor with predefined displacements for the various pressure and temperature cases?
Any suggestions?
Greetings, Albert Riedstra

To compute radial displacements due to pressure simply use this relation :

Radial strain = (hoop stress/E)-(poisson's ratio*Longitudinal stress)/E.Then compute displacement from strain.

You can feed the computed displacement by modelling the anchor with C-Node and specifying displacements at the C-Node.

Else only specify displacements at the nozzle load.


Anindya Bhattacharya

Stress Analyst

Bechtel Corporation

This is what I figured out too, but the problem is that I want to include the horizontal movements of the tower due to earthquake and wind loads. And by modeling a Cnode with displacement in the pressurizing load case, this node will be fixed for all other cases and the tower displacements in the other cases will be absorbed by the (fixed) Cnode.

You can raise the temperature of the rigid element connecting the nozzle node to the centerline of vessel, by an amount that will give thermal expansion equal to radial deflection. This will simulate the radial deflection along with the others you want to include in the model. Obviously this trick is valid only for the load cases that include temperature or those which are derived from the cases that include temperature.

The same problem I experienced when I did the stress analysis for Tower Piping for a FPSO recently. For a FPSO, the displacement due to ship accelerations for various levels of the Tower will vary according to the height of the tower, in addition to thermal displacements. I found it difficult to include the tower in the same piping model for various combinations.

I made two caesar models one exclusively for the tower and another for the piping. In the tower model, I defined the nodes for all nozzles and pipe support clip locations. I could calculate the displacements for various load cases like wind, temperature, ship accelerations etc using that model.

In the piping model, I defined cnodes for nozzles, clip locations and fed the movements calculated from the tower model for each load case as D1,D2,D3 etc. I defined appropriate load cases in the piping model by suitably combining the above displacements.

I hope you can follow the above to overcome your problem.


K.R.srinivasan.