I am modeling a system and am having a sustained failure at a tee intersection. It is a 12" dia. tube with 0.25" wall thickness main, with a 6" sch80 (0.432" wall) pipe connected to it via a weldolet. I modeled this node and selected 'weldolet' as the Intersection/SIF type.

After analysis, it fails code (B31.1) at this point and in reviewing the results the SIF for this node was calculated by caesar as 14.582. I've been trying to figure out what's going on, and the strange thing is that by playing with the wall thicknesses of either the 12" or 6" pipe I can the SIF drop down to even as low as 3 (which then passes). For example if I change the 12" pipe to the same wall thickness as the 6" pipe.

Could someone explain to me what I am missing?

I don't have the Code in front of me, so I won't rattle off any equations. So ...

1) Go to Appendix D of B31.1 and compute the "h" value, then the "SIF" value. These will be based solely on your diameter and thickness.

2) Now read the notes for your fitting. Make sure there are no limits or other qualifications that could vary the values computed above.

Once you understand where your initial SIF value comes from, you'll be able to see how changing the parameters affects the answer. Note you can do a lot of playing around with the values in the SIF ScratchPad, built into the Piping Input Processor.

Thanks Richard -

I performed manual (hand) calculations using the equations from Appendix D in B31.1, and also entered the values into the SIF scratch pad in Caesar. In both cases I come up with an SIF of 3.3 for my configuration, yet when Caesar calculates the SIF when running my actual model, it is using an SIF of 14.582. There are no notes in B31.1 that I can see that would affect the result.

Any ideas what the problem could be?

Thanks again.

No ideas without looking at your file. Please e-mail it to techsupport@coade.com.

Dustin,

I have performed the calculations by hand (see attachment) and I match Caesar II's results perfectly (right down to the third decimal place). You are not using the branch equation which is listed plainly in B31.1 Table D-1 just below the Welded-in contour insert sketch.

You can reduce this SIF by switching to Pre-1980 version of B31.1 (see the Configure/Setup switch called Reduced Intersection). Or you can switch to B31.3. But for the modern B31.1 the results are correct for the input you gave. You could also beef up the run pipe wall thickness in this section to reduce this SIF considerably.

Thanks very much for your help. You are correct and I had overlooked the branch calculation.

The run pipe is actually already existing and it is just the branch that we are installing. I will decrease the wall thickness of the branch to SCH40 and see if that works.

Thanks again