Hellow to all!

I must simulate a nozzle at a flat header of a horizontal vessel. In order to take into account the nozzle flexibility (since as far as I know there aren't any guidelines for flat headers of vessels, like API650 or WRC107) I assumed a flexible anchor with 10^6 stiffness. My client argued that the flexible anchor has six degrees of freedom and this is not realistic for tanks and vessels nozzles because for these nozzles only three out of the six degrees of freedom are flexible. He allegated that only axial translational and bending in-plane/out-of-plane rotational degrees of freedom should be assumed flexible and the other should be rigid.

I am not quite sure about this. Can anyone help me about how many degrees of freedom can be assumed flexible for a nozzle according to API650 or WRC107?

According to my caclulations, the translations and rotations of the nozzles were very small for the flexible anchor modeled (less than 0.3mm and 0.1 degrees). Therefore I consider that the flexible anchor is a realistic approach of a tank vessel in case that API650 and WRC107 are not valid for this vessel (for instance for not cylindrical or spherical vessels, but flat planes). Im I right?

In many applications you will have three directions that may be flexible but the other three will be rigid - the shear terms. For your flat plate nozzle the two translation terms perpendicular to your nozzle are the shear directions as is the rotation about your nozzle. So the axial translation and the two bending terms (with respect to your nozzle) may be flexible.

Thank you for your advise. But if I consider a nozzle as flexible anchor (with stiffness 10^6) and the translation in shear terms is less than 0.3mm, do you believe that this would lead to an overstress of the vessel?

In my case, a translation of 0.25mm produces a signifficant relaxation of the stresses in the piping system and reduces signifficantly the forces on the vessel nozzle (from 2.5 tones to 0.6 tones in shear terms).

Since this negligible translation could be absorbed even by the nozzle flange, isn't it very conservative not to take it into account?

It makes me wonder what's on the other end of this line. Don't just focus on your vessel nozzle; that's a heck of a large load differential for your stiffness change. It appears you have an anchor on the other end as well. It would be improper modeling to fine tune one support while others remain coarse as this artificially changes the overall strain distribution.