Tangential Nozzle Question 1

[PipingEquipment]

I know its poor engineering practice to put a nozzle on the outside of a head going through the knuckle radius, but my question is does the ASME Section VIII div 1 Code allow (or prohibit) a nozzle that is truly tangential with the OD of the Head (or Shell), that is to say can a nozzle with the same thickness as the head be pushed so far out that the OD of the vessel is the same as the OD of the nozzle (ie a nozzle that takes up side wall)? This is for a flush on a horizontal tank, and the nozzle is positioned on the extreme lowest point possible so that the vessel can be flushed of all liquid (ID of straight flange = ID of nozzle), and the nozzle is inline with the longitudinal direction of the shell.

I wasn't sure if code barred this because the philosophy of replacing what you take out seems like it may be violated if you are essentially butt-welding a nozzle (on the outermost portion) to the head but no way to RT it to bring the joint efficiency to 100%. I think I've seen this done but would like a more experienced vessel designer to confirm or deny.

Thank you,
-Kyle

 

[gr2vessels]

I am replying to this post, assuming that I might be a bit more experienced designer. The code does not prohibit the installation of a nozzle in the knuckle area, including the tangential one you mention. It's only matter of appropriate calculation, because the reinforcement could be e bit tricky. The easiest I would suggest for you is using the FEA for design. Depending on the design conditions, size of the nozzle, wall thickness, external nozzle loads, etc., the materials and fabrication method, you might need to consider PWHT, subsequent deformations, UT in lieu of RT and DP/MP tests. Also, you might want to consider an alternative design with a radial drain nozzle on the shell instead of this relative difficult nozzle on the dished end. Also, you could use an eccentric conical end in lieu of the dished end.
Cheers,
gr2vessels

 

[PipingEquipment]

Thank you very much for your response, I could do an FEA stress simulation using Autodesk Inventor Pro, but my ASME Code calculation program (Design Calcs formerly Advanced Pressure Vessel) does this calculation and I can get it to pass, but my understanding is that the ASME Code (and thus the calculation program) uses a membrane stress aproach on calculating stress/strength in heads which can be more accurate if done by FEA, is this why you recommended this approach as thin out and geometry of the knuckle cause a rise in stress more precisely represented by FEA?

 

[PipingEquipment]

Also about using an eccentric head; due to head diameter and multiple nozzles on the head means this aproach is unpractical unfortunately.