Hi everyone,

Could anybody tell me and the new learners too perhaps how to determine stiffness of nozzles (pumps, compressors, pressure vessels, etc) that I could plug-in directly into Caesar while modeling a flexible anchor. So, they are the x,y,z directions as well as the rotational stifness if required say rx, ry and rz. Thanks a lot.

Gimini55

No code/standard addresses the issue of flexibilities of pumps/compressor nozzles. WRC 297 which can be applied for vessels provided they meet the geometry requirements.API 650 can be used for computing nozzle flexibilities for storage tanks ( > 120ft in diamtemer).Finite element analysis is a general purpose tool and can be used in any application in computing flexibilities.

I feel that you spend some time going through through these documents, understand them and go through basics of stress analysis ( including basic mechanics ) before attempting to use them.

Regards

Anindya

As suggested by Anindya a FEM study will answer your queries.

For your information:

Normally the Pumps nozzle stiffnesses are not considered in the analysis as they are assumed to be rigid.( debatable of course) Rigidity is relative in nature.
The pumps allowables are normally governed by the deflection of the shaft and the clearances between the Impeller and Casing. Please keep this information as a go-by while including the stiffness of the rotating equipment nozzle stiffness into your analysis.

Thanks Anindya and Neelam,

Thanks for the quick reply Anindya and Neelam. I studied FEA and actually could model a nozzle and the vessel and obtain rotational stiffness as well. As I learned from seniors five years ago, I normally used 1,000,000 for pump nozzle or a bit bigger as the pump sizes go bigger and noting that using stiffness (a measure of rigidity I believe-please correct me) a couple of thousand differences in these values does not have remarkable effect on forces, moments and stresses. So, I also understand that a real rigid a default figure from Caesar is more conservative. But I don’t know with your experiences on handling allowable loads but I experienced these as not an easy one specially on SUS + OCC (normally seismic)(moreover when I realized your load cases is that SUS+U1+U2). In seismic regions where your seismic load acceleratring factor is high say 0.5, a massive pipe and valves could have a great impact on the nozzle loads and moment. And, so many times I have to figure in a flexible anchor on pump’s nozzle. My main focus is the fact that I realized some folks out there are using Caesar on heavy steels pressure vessel to obtain stiffness value. For example, I was informed that they model the piping and vessel and twist it around the x, y and z direction to get the stiffness figure – and probably the vertical and horizontal direction as well. Is this possible? (Mr. R. Ay....etc?) Thanks and I really enjoy and having this forum very helpful.

Gimini55

Well you can build the model as you describe, and displace it to determine a stiffness. However, the resulting stiffness will only be as good as your model. CAESAR II only uses a 3D beam element, which is nothing but an infinitely thin stick. So if you have a vessel (tower), on a conical skirt with anchor chairs, with conical sections between shells, I would think your results, while in the ball park, would not be all that accurate.

Thanks a lot for the clarification Rick. I do appreciate your time for this.

Gimini55

Richard,

I saw this subject a while ago and how do you actually do it in Caesar. I mean what would you do in the model when you say displace it and determine the stiffness. I understand the theory but I am not certain how I would do it in the model so that a specific nozzle in the pump or in a vessel will give a stiffness that is very much representative of the conservative (approximate?) value. I would say that once I get this figure on the procedure that you have mentioned and has been mentioned by Gimini55, I would use this back in the model to run the analysis. Thanks for explaining a detail procedure.

Ed-Lamigo

Assuming we are talking about a vessel nozzel, I'd just build the vessel as part of the piping model, and call the nozzle node a "WRC 297" nozzle, with the vessel node as the CNODE.

I would leave the vessel model in as a permanent part of the piping model. This way, if the temperature or wind speed changes, CAESAR II can figure everything out - and most importantly, you don't have to remember to do this.

=======================

But, to answer your direct question, you model a flexible displaced anchor as follows:

1) Call the node in question an anchor.
2) For that anchor, specify the stiffness you have, if applicable. (Note, all 6 DOFs acquire the specified stiffness.)
3) Specify a CNODE at the anchor.
4) Specify displacements for the CNODE, in all 6 DOFs.

Richard,

I knew how to perform what are you explaining 1 to 4 and I've been doing it. My question is on item number 2)-specifying a stiffness where some suggests to use an FEA. And as you noted it out above you could model it in Caesar, apply loads and twist it to get rotation stiffness or apply horizontal and vertical loads to get stiffness in x,y and z axis. How should I actually do this in Caesar? This stiffness, is the one suppose to appear on spreadsheet (in.-lb/deg or lb/in.). I normally make an approximation on these based what my seniours were using. Thank you.

Regards

Assuming we're dealing with a cantilever with no intermediate restraints:

1) Apply a tip deflection of 1".
2) Run the analysis.
3) Get the restraint load.

4) Stiffness = load / 1.

I'll try it. Thanks Richard.

I'm back Rick. Now I know what you mean when you say that by displacing a nozzle location to get a stiffness, the stiffness value obtained is only as good the model. I think the fact is you simply get the stiffnes of the pipe next to your nozzle load but not the the stiffnes of the nozzle itself which supposedly could be much higher. Am I right? Sorry I am trying to make sense out of this. Thank you.

To obtain the nozzle stiffness, the easiest course is to use the WRC-297 option as mentioned above. This procedure considers the inherent flexibility of the vessel shell, which will locally deform as moments are applied to the nozzle.

You can't arrive at these values using a CAESAR II analysis of the vessel-nozzle area, because CAESAR II uses 3D beam elements, and there is no concept of a shell.

All CAESAR II will give you is the overall stiffness of the vessel-nozzle system, based on the gross geometry of the vessel.

Okay, I think finally your recent reply woke me up and I'm back to my senses actually while learning. I would even forget this stiffness thru Caesar. I know basics of FEA at least and WRC 297 which I have not done a lot yet but not a big deal to do. Thank you so much Rick. This is a valuable thread for me so far. Have a nice weekend all.