API 620 roof joint tension / bending stress 1

[FoxRox]

I am working my way through a cone roof design for a 12' diameter silo rated to 14.9 psi. I am using API 620 as a design basis, although the silo will not be stamped / monogrammed to any standard. The rest of it is designed per ASME Section VIII, but I've been asked to explore a shallow cone roof option instead of a F&D head.

I have a design that I believe passes API 620, but I have also been using FEA as a gut-check on the design. As expected, the region around the top joint is in circumferential compression. I have explored using figure 5-6 detail 'g' and detail 'h-1' among others. The added material to satisfy API 620 cross-sectional area requirements resists the compression nicely. However, there is an area of concentrated high tensile stress directly on the inside of the roof-shell joint. The stress is well above yield. The cause is fairly obvious. It is a bending stress as the joint tries to "flatten" as the cone roof is pushed up and the shell is pulled in. Is this not a concern? API 620 does not appear to address it anywhere. It just assumes the joint is fine if the compression ring is adequate.

Please advise if I am over-thinking this or if this issue is addressed in another portion of 620. I am experienced with ASME and API 650, but not so mush with 620.

 

[JStephen]

One of the advantages of using a code-based design like that is that you have reasonable assurance that the design will work even if it is not all analyzed to the nth degree. So yes, the stress is really there, but no, it shouldn't be an issue and a tank roof meeting the code would be expected to perform satisfactorily.

Some thought problems on how all of this lines up. Basically, the compression ring area is trying to form itself into a knuckle, that's the effect you're seeing. Well, if you use an F&D head, how does that knuckle get in there? They stick a flat plate in a big machine and just bend the heck out of it well past the yield point, then build it into a tank and then it performs just fine. So you don't calculate those forming stresses in your FEA, but that doesn't mean they're not there, either.
Similarly, on most vessels, you take a plate, roll it to radius, which stresses it past the yield point, then weld it and use.
On typical flat-bottom tanks, the shell is restrained at the base, so the shell above the base deflects outward when loaded, and there is bending at the shell due to that restraint. But that bending is neither calculated nor limited by the codes. Similarly, where two shell courses differ in thickness, the radial deflections are different, but bending action keeps the plates together- but this bending is also not calculated or limited.

Somewhere in API-620, I think they have some commentary on issues such as this, perhaps in connection with a knuckle. It is actually permissible for some permanent shape change to occur in the tank, provided that removing the load/pressure does not cause yielding in the reverse direction. Basically, you're just letting a small amount of forming to take place due to the load, as opposed to forming the piece in some way, then building the tank from it.
On a small tank such as yours, I'm guessing this distortion won't ever be visible, though.

 

[HTURKAK]

Yes...It is a concern !!! ..

Let me express my opinion; when you say silo rated to 14.9 psi could be designed as per the rules of API 620 since the scope limit 15 lbf/in.2 you are right ... but i do not think it is reasonable if you use conical roof for that pressure and diameter..Hemispherical , elliptical , torispherical dished heads would be better options..

When conical head is used for high internal pressure, high unbalanced forces at the cone-to-cylinder junction will develop and shall be considered for the cone design junction and discontinuity analysis shall be done .. If you look ASME VIII-1 , the code limits the apex angle to a maximum of 30∘ because of these high forces...and for the apex angle, above 30∘,discontinuity analysis is a MUST.

I do not think that it is a good idea to have a cone roof with H=2D .

P.S. Sometimes you may circumvent the applicable standard but not the rules of physics..

I hope my respond answers to your thread..

 

[FoxRox]

I appreciate the replies. You guys have perfectly articulated both sides of the conflict in my mind.

 

[MJCronin]

With 14.9 psig design pressure and 12 foot diameter, what is happening at the flat bottom of the tank ?

... or did your MBA boss only ask you to look at the roof ?

Was another poor engineer assigned to evaluate the bottom anchorage and how is he doing ???

My opinion only ....

MJCronin
Sr. Process Engineer

 

[JStephen]

I don't see where they said the bottom was flat. With a cone, the question would be similar to the compression ring above. With a flat bottom, you'd anchor it with anchor bolts like other flat-bottom API-620 tanks. The uplift wouldn't be that huge on a 12' tank.

 

[MJCronin]

OK .... again, What is happening at the bottom of the 14.9 psig tank ?

Is it flat, coned ? ....or what ???

MJCronin
Sr. Process Engineer

 

[FoxRox]

MJCronin, thank you for your concern, but as I said in my OP, the rest of the silo is designed per ASME VIII-1. The bottom is a cone with a 30 degree apex angle. Until the cone roof came into the picture, the whole thing was designed to ASME, with the top head being torispherical. I only turned to API 620 as a basis for the cone roof.

 

[MJCronin]

So .... This is a silo that is being designed, constructed, welded, tested and certified to a vast variety of codes and standards ?

.... and your MBA boss has asked you to explore an API-620 style shallow cone roof evaluation based on his feelings that particular day ?

.... and for unknown reasons, a torispherical head (an ASME head type that is perfect for this diameter and pressure range) is unsuitable ?

.... and this whole design/analytical exercise is not a flight of fancy and is somehow related to an actual project for an actual client ?

Do I have that all right ?

??????



MJCronin
Sr. Process Engineer

 

[JStephen]

That is a pretty common situation for non-code vessels/ tanks/ silos. ASME doesn't adequately address flat bottoms with anchor bolts, API-620 doesn't address shop-fabricated tanks and doesn't adequately address F&D heads, and neither addresses granular products.