API 650, Figure 5.7a, detail b, vs. Figure 5.12, detail b

[Soln]

API 650, Figure 5.7a, detail b, indicates that the corner-corner weld of the flange to the neck shall be the thickness of the thinner member joined however, Figure 5.12, detail b, indicates that the corner-corner weld of the flange shall be the thickness of the thinner member joined, with a max weld of 1/2 inch.

Wouldn't the same apply to both? Is the fillet limiting note mistakenly absent from Figure 5.7a, detail b? Or, is the note mistakenly placed in Figure 5.12, detail b?

 

[Engdoitbetter]

Hi CSKCJK,

I know this standard but never designed a tank, anyway, pictures refer to different structures, the first a manhole (same weld bevel for the entire perimeter), the second a flush type cleanout fitting (different weld bevels for top and bottom).
Furthermore, the shapes and positions of the openings are different, so I would definitely expect differences in design.

Maybe you can find a clue in the related paragraphs?

Regards,

Stefano

 

[fegenbush]

I do not believe that the 1/2" limitation is missing from Figure 5-7A. Note 2 of Figure 5-7A requires that the gasketed surface shall be machined. This makes weld distortion of the flange face inconsequential, since it will be machined flat. The flush clean out internal fillet weld cannot be made with the cover plate bolted to the flange as the external fillet weld can, and could be distorted by too heavy a weld. The external weld distortion will be controlled by the presence of the cover plate while welding and the PWHT after fabrication.

 

[Soln]

Why doesn't the standard (API 650) recommend the face of the flush-clean-out (FCO) flange be machined? We machine ours. Do you? How is it different from a manway flange?

 

[fegenbush]

"How is it different from a manway flange?"

Manway flanges are round, flush-clean-outs are not. Manway flanges have ring type gaskets, flush-clean-outs have full face gaskets. Flush-clean-outs are subject to rotation due to interaction with the vessel floor (and possibly anchor bolts), manways are not.

How do you machine finish your flush-clean-outs? When do you perform the machining? Without a phonographic finish, how do you determine that you're not making the seal worse than a non-machined surface?

 

[Soln]

We use the same full face gasket with both manway (MH) and FCO, and we don't apply a phonographic finish- we apply a flat machine surface with 125 Ra. We have the FCO, and MH machined on a milling machine. We machine the FCO after PWHT. I take it you guys don't machine your FCO?

 

[fegenbush]

I will restate what I believe to be your question:

Is machining of FCO flange faces required by API 650? No.

 

[Soln]

Actually, what I'm questioning is- if 1/2 inch fillet is enough for the flg-neck weld of the FCO, why isn't it enough for the same weld of the manway? Additionally, based on the point you raised regarding weld distortion, I'm curious, does anyone actually sell FCO that isn't machined? And, what's the level of comfort/experience that the FCO does not leak?

 

[fegenbush]

Your question is valid, and I don't have a direct answer. I could bring it up at the next committee meeting if you would like. There are some things to consider, though.
1) Many manway necks are less than 1/2" thick, so the difference is only applicable when the thickness of the shell and reinforcing plate is greater than 1 1/8".
2) A FCO has more weld length per area of opening than a manway. I would have to calculate the percentage difference, but that could be a reason that the limitation does not exist for manways.

 

[IFRs]

I don't see FCO leaks and with all the bolts I can't see it leaking. I recently saw a 36" manway whose flange I measured and was warped by one-eight of an inch and it did not leak. It is curious however, and I have send API to remind them that the new figure 5.12 is missing all its notes.