Considered Local thickness of shell at trunnion pipe OD in WRC

[ting123]

Dear's,

The trunnion sketch as shown in the trunnion calculation page depicts the tunnion pipe is directly connected to the shell and in the WRC calculation the local thickness at the pipe OD (RM/t)is considered as repad+shell thickness.
My query is shall the trunnion pipe be directly welded to the shell or shall it be welded on the repad. In contrast PVelite uses a different approach. It does not considers the pad thickness in the WRC calculation and shows the sketch in the same way as compress ie the trunnion pipe is directly connected to the shell.

Could someone please provide some clarification on which approach to follow. Till today we have been welding the trunnion pipe on the repad and not directly to shell.

Thanks in advance to all the members who provide their valuable view on the subject matter.

 

[TomBarsh]

Consider the load path. By attaching the trunnion to the reinforcing plate the entire lifting load must transfer through the reinforcing pad. What is the purpose of the reinforcing pad? It is to reduce the local stresses in the shell as analyzed by WRC-107 or other analysis and so is not especially thick. The reinforcing pad is generally not intended for direct application of loads unless it is suitably designed for this. By applying the full lifting load to the reinforcing plate the plate must be designed for these loads, which may result in a thicker plate than is expected. The attachment welds of the plate to the shell must also consider these loads. For smaller/lighter vessels this construction may be acceptable, but I don't believe it is acceptable for refinery-type vessels.

 

[TomBarsh]

Should the local stress analysis include the thickness or the "pad" plate? or not?

COMPRESS does both!

COMPRESS checks the stresses in the shell at two locations. Each location is checked using the WRC bulletin 107. The bulletin determines stresses in the shell at the outside radius ("rO&quot of the circular attachment.

1- COMPRESS checks stresses in the shell at the outside radius of the trunnion barrel. At this location the combined thickness of the shell and reinforcing plate is used.​

2- COMPRESS also checks stresses in the shell at the outside radius of the reinforcing plate (thus the reinforcing plated is assumed to be the "attachment&quot. At this location only the thickness of the shell is considered.​

This method is consistent with what I have seen done in industry generally, using non-COMPRESS methods (hand calculations, in-house fortran programs, etc.). Which is not to say that everyone in industry has done it this way.

Edited to add: Remember, COMPRESS assumes that the trunnion barrel is directly welded to the shell.

 

[TomBarsh]

It would be interesting for you to post the question (whether the trunnion barrel should be welded directly to the shell or need not be) to the main pressure vessel forum. It would be interesting to see the views on this and what actual practice may be.

There is a wealth of discussion regarding design of trunnions and reinforcing pads on eng-tips if you use the "search" function. It seems there is no consensus on whether the reinforcing pad should be a "donut" ("bagel"?) or a "pita". But certainly if the pad is a "pita" then it must be fully designed for the lifting loads that must be transferred through it (and how is this done without FEA?).

 

[saplanti]

The trunnion needs to be welded onto the shell and as well as to the reinforcement plates around the trunnion. Therefore you are able to use the (shell + reinforcement) thickness in the WRC analysis. Reinforcement plate needs to be welded all around as well at external circumference.

Welding the trunnion onto the reinforcement only gives you the reinforcement thickness to use in WRC calculation and this way of the calculation is not reliable, however using the reinforcement thickness only is conservative in compression loads only. Tension and/or bending stresses on shell are greatly underestimated if you use the total thickness in the calculation without welding the trunnion on shell.

Hope it helps.

 

[ting123]

Thank you Sir, for your valuable replies.