Ahmad,

I can drop in a single miter 90 degree joint on a 6" std pipe in B31.1 and I get an SIF listed in the expansion stress report as 4.540. What are you doing that's different? (You may have to send me your input if this goes much further.)

On your other point - the SIF for the bend is not based on the angle (while it does play a role on the closely-spaced miters). The SIF will be the same for a 45, 38 and 90 degree bend. The bend angle will change the overall stiffness of the component through the changing arc length as stated earlier.

A revelation (to me) - I never worried about resetting the program's long radius default for single miter joints because the miter radius (R) doesn't appear in the Appendix D calculations (of h, k & i). But today, because of this thread, I ran a single cut with the default (long) radius and compared it to a single cut with the Appx. D calculated radius of r/2 (actually, CAESAR II will not allow me to run a radius less than OD/2 - I'll ask to get that fixed here). The SIFs were the same but the stresses were different. It turns out that, here too, the stiffness of the component is a function of the arc length. With different arc lengths you get different loads and stresses.

And still more... I focus on B31.3 but looking closer at Appx. D in B31.1 I see a few more complications. B31.1 limits the half-angle of the change in direction (theta) to 22.5 degrees. That means these data apply to single miters that sweep up to 45 degrees. Of course I see no guidance for greater angles. CAESAR II does not monitor this limit.