Heat exchanger boundary ASME

[Cluain]

Hello,
Say one designs a heat exchanger with the service side made of ASME code material, ending at the shell barrel flanges and where the process side is made of non code materials. The Non coded process side is mechanically held in place by tie rods extending from the code body flange to an end plate which is also made from ASME acceped material and calculated per ug 34/39 respectively.
Can one legitimately declare that the exchanger is code from endplate to endplate, even if the distance from shell flange to end plate is 12-16 " (ie the non code process side material seperates the two code declared items on either end).
If not, why not ?
Please base argument on Section V-III dv1
Thanks !!

 

[metengr]

The answer to your question is located in the Introduction of ASME Section VIII, Div 1. If you review U-1(e)1 (c) it states that the scope of this Division (Code section) applies up to and including the face of the first flange for bolted and flanged connections...

 

[Cluain]

Hello Metengr,
I considered that paragraph, but determined it as more of a general and not a limiting rule.
If one considers a series of bolted cylindrical columns all carrying the same fluids and no pressure differential, then one can, in my opinion Code this type of design from end to end. (I believe there is an interpretation which discusses the pressure testing of bolted shell sections.)If you look at the configurations in Fig UHX-3. Then you will agree that the paragraph you quote is more generic.
If you look at fig UHX-3(c) and imagine the channel and shell cover missing. Items 1,6 and 13,6 on the other end, are bolted with tie rods, code material. The intervening materials are non code, ie, stationary and floating tube sheet, but these may be anywhere from 3 to 16" thick. Most of the process side of the exchanger is constructed from non code material, except items 1,10 and 13.
Again, is it legitimate to stamp the exchanger as meeting code end to end, provided the U1 stipulated that the process material are none code except for bolted end flanges. In other words can one argue that the non code tube sheets act like a massive gasket between bolted end flanges and shell flanges.

 

[metengr]

No. The Code rules are designed for pressure containing components or vessels, period. If you have a sandwiched vessel design where some of the so-called non-Code pressure components are subjected to the same pressure conditions as the Code-approved material, I believe you have to include these items in your ASME pressure vessel design code space.

Look at the entire component (heat exchanger) from a pressure boundary standpoint. I would not use the argument that just because you have a series of bolted connections that are tubesheets sandwiched between code designed material that you can exclude these components because they behave like gaskets.

 

[Cluain]

Thanks,
From a mechanical point of view, why can this non code material which one uses for a tube sheet not act as a gasket. Lets say this tube sheet material is made from, shall we say, Teflon.
Note,I agree with you !! The sandwiched material(process side) should be encapsulated with code approved material and be capable of with-standing process pressure, as depicted in Fig UHX-3. However, if the tube sheet is made of such a material as mentioned above is it not a plausible scenario to say it doubles as a over-sized gasket. What limit am I not considering..

 

[GenB]

If the 2 vessels are tied together by bol;ts and are stamped
as one unit (each vessel with own stamp ) but one vessel with the master stamp it is possible one, two or three or more tied together all listed as a unit. (some inspectors will not allow the practice and will require separate stampings.
Now Put a barrier of non Code item/materials in between you have separate items, period.
Or even if you put a Code item in between with a different stamping: you are separating the items and it is not a unit anymore.- If the units were stamped as one, you loose the certification as one unit.
see the Code is not stupid and the AI
can not be duped as well.