Stacked Repads?

[crosby84]

If the title invokes in you some kind of reaction and makes you say to yourself, "what an idiot - let me tell this guy how wrong he is", please know that that was my intent as I often invite criticism in order to keep me on the right track. So fire away.

Anyhow, we have a client with an inadequately reinforced manway on a F&D head (almost in the knuckle) on a Div. 1 vessel. I have spoken with the R-stamp holder who would be doing the work to see if they can physically even make the repad size I have in mind (3/8" x 3") that would form to the knuckle before I specify it. He said that it might be possible but would be tough, and that a more easily fabricated setup would be two thinner pads, 3/16" x 4.5, and 3/16" x 4 on top of that.

Now at first it seems that if the material is within the limits of reinforcement, there would be no issue with this. But something about it seems...illegitimate... though I cannot find anywhere in the code that prohibits it (if there is a section, please direct me).

I just don't want to be that engineer that tells a fabricator to make something that can't be made even though the calculations say it works.

Thank you for the help

 

[pdiculous963]

I've seen it done before. Its not an idea solution, but it isn't prohibited. The guy who performed the calculations was also an active member on the Code committee.

 

[JStephen]

The main potential issue is that you limit the strength of the outside fillet weld due to the plate thickness, but the weld would need to carry loads from both repads.
Also consider possibly putting the second repad on the inside.

 

[TGS4]

Although I don't have a fundamental issue with this, the proximity of the nozzle and repad to the knuckle makes me very uncomfortable. Whatever rules-of-thumb that people might use in cylinders or in the center of heads ought to be thrown out the window at this location - especially in an F&D head.

While I wouldn't prohibit it, I would ask for an analysis (per U-2(g) and VIII-2 Part 5, that would also consider the contact between all of the layers) before I signed off on it.

JStephen's idea of one repad inside and one outside definitely has merit.

AND, triple check that the vessel won't be in cyclic service - using the Method A screening in VIII-2 Part 5. Normal vessels would need at least 1000 cycles before they would be considered in cyclic service - fillet-welded repads in the knuckle of a head drops that number down to 60. So, even if you can do this, it might bump you into having to do a fatigue analysis (you might have to anyways, whether the repad is one or two layers).

 

[crosby84]

Thanks for the responses everybody. All useful information.

 

[DSB123]

Agree with TGS4. The proximity of the repad top the knuckle is not very good and in fact is not allowed if the vessel were designed to PD5500 or EN13445.