Yes, it is a popular engineering practice to use the SIF of 2.1 at trunnion intersections.

However I do not agree to such a simplistic look into the problem when the trunnion attachment is at the bend.

We know that the flexibility of a bend is mainly due to its ovalization . This reduces the section modulus hence the flexibility. So "more flexible-more SIF ".

I think the presence of a trunnion can actually contribute to prevent the extent of ovalization , something in line by which a flange addition ( or two flanges) at the bend end(s) reduce SIF and flexibility. So you can have a reduced SIF at bend with trunnion.

However this problem does not have a simplistic solution as the stress distrubution can change significantly due to addition of the trunnion, something which you cannot make out from the simple concept of ovalization.

So there are contrasting views as whether SIF will increase or decrease by addition of trunnion at bends. So 2.1 may not be a conservative estimate.

There are lot many ASME papers in this field.CAESAR II ( at least till version 4.3) has the facility to compute SIF at bends with trnnion attachment based on ASME paper.However my personal opinion is that : "do not use this module " as it I had no. of interaction with CADE where they have recommended not to use this option.

At other locations ( non bends) I think 2.1 is a conservative estimate.

A.Bhattacharya

Stress Analyst

Bechtel Corporation