Hi everybody,
If a line say 10" size has 2 branch connections
of say 2" i.e. one vertical and one horizontal, then does the SIF remain the same at that point ?
Regards,
John Wright

Correction : What I meant to ask was if the two no. of 2" tappings are at the same location, but in different directions i.e. horizontal and vertical, then does the SIF change as compared to a single tapping ?
Regards,
John Wright

Maybe....

If the two fittings are close enough to each other they will have an impact on the SIFs....

You need to give us information like what is the header diameter and thickness as well as the thickness of the branches...

And finally just for situtations like this B31.3 says YOU are responsible for coming up with an apropriate SIF...

Question why not simply move the branches apart on the header so this problem goes away?

Hi John,
Thanks for your reply. Yes the fittings are at the same location. One is vertical and one is horizontal. Due to space limitations, this was done.
The header is 10" Sch 100. The 2" branches are Sch 80S. The material for all is A790 S31803.
Regards,
John Wright

John,

I can see that conversation with piping design now:

"I don't know how to model this, can we move this branch over...."

Ouch Ed! You just struck a nerve.

Software limitations should not be an excuse for poor design as it too often is.

I'm still not convinced that abandoning pencil & velum has been a good thing.

My 2¢,

This is not a software limitation, you can model this. The problem is the SIF likely won't be correct because this fitting wasn't tested by Markl, and is therefore not addressed by the piping codes.

There are two alternatives here:

1) Move the branches apart (as John suggested) to avoid the issue, or

2) Bite the bullet and model this using FEA and then enter the analytically obtained SIFs.

Richard,

To clarify, I was refering to limitations in piping design software such as PDS or PDMS (as I thought Ed was) and NOT limitations in CAESAR.

Well Well ,

Apparently my questions were subject to misinterpretation as to their intent….

I wanted to know about the D/T ratios so as to get a feel for the sq. rt. of r/t distance for the header as well as what pressure range the system was running at. The other question concerning layout has nothing to do as far as any CAD or CAE software limitation is concerned but the fact is that simply because somebody can weld some lumps of metal together does not substantiate the structural integrity of the final product!

The Sch 100 implies high pressures so the first concern is pressure design…. Factually the two branches will have the minimum ligament space of ~ 5.4”. The transmission length or attenuation length of stress is for rule of thumb considered to be twice the sq. rt. of r/t which is ~3.2”. 3.2”x2 = 6.4” so this ill conceived layout may have problems from the pressure retention stand point alone… I point out to you ASME B31.3 para. 304.3.3

SIF wise… well the attenuation length for stress also seems to indicate there ,may be some affect there as well and para 319.3.6 in the next code should say something about if listed components are combined in a manner which they were never intended for the designer is responsible for coming up with I and h… (it took me a long time to get this paragraph revised but I did so when facing another questionable arrangement that some loud mouthed piping layout person said was fine because it could be built in a CAD package)


Change the layout or do some FEA if you want to do a proper job or ignore the situation and hope for the best.