Yes, but it involves defining dummy header elements attached to the start node of your branch. Define two short elements, to the right and to the left, starting at your branch node. Leave the outside ends free, but make sure the diameter and thickness are correct.

Since these fictitious elements will be short, they shouldn't affect any system loads, but you could zero the densities and pressure, and set the temperatures to ambient. (Just verify the following elements so that this data isn't duplicated forward, erroneously, to any elements following these two header elements.)

Thanks Richard,

If I have understood your response then I have already attempted what you have suggested.

The problem is if I create a dummy header at the branch node (at the physical branch header intersection) I will have four pipes framing into a single node. CAESAR won’t correctly calculate the SIF if there are more than 3 elements. There will be the branch element, the rigid element from the branch to the centre of the header and the two dummy header elements?

Cheers,

I went through the Pipe Stress Analysis by Peng. What I found is below:

Equivalent Stress Intensification factors can be used to simply the process of evaluating the stress at the branch Leg. In this method, the stress at the branch leg is calculated with the unmodified original branch leg section modules.

I don’t have a copy of the book you are referring to, but I had thought about modifying the user SIFs to account for the effective branch modulus.

I was just hoping that there was a quick way of doing it in CAESAR. Does anyone have a modelling technique that works successful in CAESAR? I couldn’t get Richard’s technique to work.

Sorry to take so long to get back to this - millions of things to do.

I believe there are two options for you:

1) Instead of the dummy rigid element method, try using the "Offset" fields - define them on the branch element.

2) Alternatively, although more complicated, you can continue with what I proposed above, but use a CNODE to connect the end of the dummy rigid to the branch.

I feel the "offset" method would be the better option here.

What about the fact that "CAESAR II constructs a rigid offset from the centerline of the header pipe to its surface." (User's Guide para. Class 1 branch Flexibilities)
If the junction is modeled as shown above, Caesar will contruct one more rigid element starting at the outside of the surface. The flexibility of the system will not be accurate?

Second point, I've performed a quick test comparing Tee with and without rigid element between the centerline of the header and the surface of it. Caesar gives an almost zero displacement at the surface when a rigid element is modeled
When there is no rigid element there is a displacement even if Branch Class 1 Flexibility is set to FALSE (the displacement is bigger if set to TRUE).
It seems therefore that Caesar II doesn't construct a rigid element, isn't it?

The "Class 1 Branch" directive is an option, this does construct (or simulate) the rigid element from the center-line of the header to the surface. If the "Class 1" option is false, there is no (CAESAR II) constructed rigid element.