Leg materials and MDMT temperatures

[michigander]

SA-36 is now becoming increasingly unavailable as the new stronger materials become popular. A-992, for example is a structural material that is recognized as a code material, but the recognition does not include Sect 2 Part D properties. Would you just always use a code recognized material for a poison pad for attachment? Would you use the generally accepted properties available on the web for design?

I've looked back over the discussion on MDMT for saddles and attachments. Are folks doing an MDMT evaluation of structural attachments used as legs? If you weld an SA-36 beam leg onto a tank with the the toes of both flanges in contact with the tank, are you actually using the beam flange thickness as a governing mdmt thickness? For example a 3/8" thick 516-70 tank with a W10 x 39# leg has an 1/2" flange on it. When attached to the shell, the beam flange thickness is the governing thickness and at 1/2" I cannot get -20 out of 1/2" thick. What to do?

I am trying to get a handle on what is reasonable and rational and what is just "code chasing".

Your insights would be appreciated.

Michigander.

 

[jamesl]

For poison pad, most shop uses the same material as used on the vessel.

To use A992 as legs, you don't need to use Section II allowable. Legs are usually designed per structural steel code, such as AISC.

When you weld 1/2" thick attachment to 3/8" thick vessel, the governing thickness is the thinner part, which is 3/8". See UCS-66.

Don't forget there is a UG-20(f) exemption for carbon steels. For UCS-66 Curve A material (structural shapes), it is 1/2" for MDMT not colder than -20F.

 

[jte]

Presuming that your code of choice is VIII Div. 1...

"Poison pads" can be used, but if the head/shell are carbon steel, I wouldn't bother with them unless you need them as reinforcing pads due to stress issues. I'm not familiar with A992, but presume its a weldable carbon steel material. In that case, it probably complies with UG-4(b) and can be shown to comply with UW-5(b).

Material for nonpressure parts, such as skirts, supports, baffles, lugs, clips, and extended heat transfer surfaces, need not conform to the specifications for the material to which they are attached or to a material specification permitted in this Division, but if attached to the vessel by welding shall be of weldable quality...

I've not seen much concern for MDMT of supports - though if they were substantially in tension I'd consider it. Generally it's hard to sell the argument that a skirt in compression is susceptible to cracking. Yes, some parts of saddles and offset legs may be in tension; apply good engineering judgment. But Div. 1 does not (yet) directly address the design of supports.

jt

 

[jamesl]

Although Div.1 does not have design procedure for skirt and other vessel supports, it does require impact test evaluation of vessel attachments that are essential to structural integrity of vessel. In a Div 2 code interpretation, it was stated that skirt and saddle fall under this scope.

Skirt are not always under commpression. For a tall column under wind, the skirt side facing heavy wind is usually under tension.

 

[jte]

jamesl-

Good point about the tensile side of the skirt. However, I've never seen the tensile side of a skirt govern the thickness - its almost guaranteed that the compressive side will govern the thickness. Thus, I'd wager that in most cases the tensile stress developed will be less than half the tensile allowable stress.

Worth checking, but I don't think tensile stresses in the skirt will be likely to govern the MDMT.

jt