Dear Seniors and peers,

Request a clarification regarding the Modeling of vessels with rotations at the nozzle connection. It is a usual practice to model vessel as stick elements ( Rigid / flexible) in the piping model to account the thermal growth of vessel.
The question is : If the piping vessel interface connections undergo a rotation for a specific reason( Say differential thermal expansion on the side walls of vessel, eg: Coker drum )and we still wish to enter/ capture the effect of rotation at the piping vessel interface while modeling the vessel,How can it be modeled.

In the attched pic, if some rotations are to be captured at node 20 while modeling the vessel, how can it be done?

Ramkumar

ramkumar,

The thermal deflections are specified in 6 degrees of freedom. Usually the movements are described simply as DX, DY, and DZ, with zero rotations of RX, RY, and RZ = 0.
If you have rotations that you want to have considered in the analysis, then include them in the RX, RY, RZ terms. Be sure to use correct units.

If the rotations are specified as thermal displacments, the node is considered "anchor with displacments" in the analysis. My question is if the vessel piping junction experiences rotation, and the vessel is also modelled to capture effects of wind sway, How could it be done?

Thanks

Ram.

Your "stick element" approximation of the vessel is the best you can do with a program like CAESAR II. Your "element #2" should be designated as a "zero weight rigid" element. Additionally, for proper wind loading, you may need another vertical element above "element #1". (These vertical elements should be defined with the vessel diameter and wall thickness.)

Modeling the vessel in this manner will provide the proper thermal and wind displacements. You will not be accounting for any local effects or distortion of the vessel shell. Also note you are not properly analyzing the vessel shell. The vessel is being modeled as a "stick", with no awareness of the shell or cross section deformation.

At your "node 20", you should also specify this point as a WRC297 nozzle, so that the appropriate flexible connection between the pipe and the shell can be included in the analysis.

ramkumar,

Sorry that I have probably mis-understood your question. If you are seeking to solve for the rotations at a nozzle connection, then it is necessary to creater the model with components and loads that will generate the results that you need.

The 'stick element' model of the vessel will give you approximate results. Richard Ay has given instructions, and I would add that the vertical shell model should include several intermediate nodes for improved accuracy - the wind load over the length of element is split into two forces acting at the element end nodes. If only one element is used the model the height of a vessel, then half of the wind load is taken to the base anchor node. Include elements for upper shell courses, as recommmended by Richard Ay. The lateral element to locate the nozzle could be a 'zero weight rigid' but do not make the vertical shell elements as rigid - the stiffness would be greatly altered. If there is major temperature differences from vessel top to bottom, then yoou might want to consider defining the temperature gradient with the nodes of shell and any support skirt. If you have thermal gradient from side to side in a vessel, with deflection like bowing, then a simple bending element will not generate the results you aare looking for.