Dear All,

Basic question but needed to understand.
When we build a piping network which has fittings (obviously), the same material and thickness of Pipe will continue for the fittings also. Is there a provision to change the material and thickness (when necessary) for fittings alone, while in the loop? Or is it not necessary to change? Is this why SIFs' are calculated for fittings, to increase induced stresses with deduced factor?

Regards,
VGA.

I don't know of any automatic way to have fittings use a different thickness. I think you'd have to adjust the thickness/schedule at each fitting.

CAESAR II allows you to enter an elbow thickness as part of the elbow input; this will allow a different element thickness without explicitly adding nodes and changing basic thickness values.
If your fitting thickness is quite a bit larger than your pipe wall the stress in the fitting may be less significant than its effect on overall stiffness and weight. In these cases it might be better to call out the component as a rigid element.
If you are looking at socket welded components you might go through the drill of properly modeling cross sections and SIFs but you may soon learn that there may be little value gained from all this extra modeling detail.

Dear Dave Diehl,

Thank you for clearing my doubt. I thought the variable thickness would effect the model geometry.

Regards,
VGA.

Per the subject line, I don't think we addressed material changes.

Generally, material changes for fittings are unnecessary while the material properties for fitting and pipe are similar. Much of the time, the material properties are identical, but the callout for the material advises of different manufacturing methods.

If the fittings are wildly different materials, say fiberglass pipe and metallic tees, you will definitely want to encode those differences, especially if their lengths are comparable to each other. However, wildly different metallic materials is uncommon and impractical due to issues unrelated to stress.

With regards to whether or not you should model your fitting as a rigid element, you should ask yourself if the thickness of the fitting is approximately 10x that of the pipe, or if the stiffness of the fitting is approximately that of a pipe 10x the thickness. Enabling rigid element literally computes the displacement this way. However, if you do this, you must also be cognizant that you're also saying there's no way the tees themselves can fail. Calculated stresses for rigid elements will always be 0.

Michael,

Some of the enginner I met, changed fitting material/thk at the piping loop(not flexible enough) just to save money as not increase whole pipe line thk or higher material. For example changing material from A335 P22 to A335 P91 at fittings gives you fine result.(I'm not saying it's neither good enginnering nor safe)

I think what VGA asked is that sort of stuff.