I am planning to use CII for analysis of large diameter piping ( 3500 mm Dia). As CII uses beam formulations for stress calculation, how accurate is it to use CII for the intended purpose?. As a large diameter pipe behaves more like a shell, is it acceptable? Request any user of CII to help me in this.

S. RAGHAVA CHARY

If your d/t is >100 then you are outside the Code limits for stresses, so results could not berelied upon. Consider that most Code data for sifs comes from fatigue tests on 4" pipe. Extrapolating to 138in pipe is a bit of a stretch.

If you do have a relatively thin wall, local flexibility at branches becomes far more important and will not be 'seen' under a beam element analysis.

You may find FE-Pipe is more appropriate to your work in this instance.

-In case of D/t > 100, you may calculate SIF thru' FE and enter that values in CAESAR II, at both header and branch.

-Model rigid elements in CAESAR II, at support location, to take into account radial growth.

-In case of sadddle support, gravity type, you may model two +Y restraints, on the either side of the shell, instead of just defining one in the centre.

-Moreover, due to thin shell (if D/t >100), local crushing strength should be done at support location.

So Pradeepd, Chary will need FE (presumably FE-pipe) in any case to get his sifs. Then you propose to return to Carsar and fiddle around with support modelling, where FE-pipe has standard templates, as it does for all the usual fittings and components.

Local shell crushing would be evident in any FE-pipe calc too.

As I pointed out above, in addition to this, local flexibility will tend to govern this kind of analysis assuming it is a thin wall.

I am a strong advocate of Caesar II, but IMHO this particular analysis appears to be much better suited to FE-pipe.