What I meant in my post above was that you should not apply SIF values obtained from using the "standard Code equations", found in Appendix D of B31.1 and B31.3. These equations do not address trunnions.

If you're using FE/Pipe, then yes you can take the FEA generated SIFs and apply them to your CAESAR II model. As to the reduced flexibility of the elbow, if FE/Pipe gives you a value for "k", then you can use this as input to CAESAR II also.

Note, these values and assumptions must be properly documented, since you have just stepped outside the scope of the Code.