NTIW TUBESHEET CALCULATION 1

[gengi]

Dear All,
I have to design NTIW (No Tube In Window) tubesheet of a vertical combined feed exchanger ( 1 pass tubeside with bellows at outlet ).
NTIW tubesheet is not in the scope of ASME VIII div. 1 UHX as well as TEMA standard since not perforated for all the overall circular area, but, with untubed parts at top and bottom (baffle cut is approx. 15%).

What is the normal design approach for NTIW tubesheet calculation?

My Client stated in the Spec's that, the exchanger, shall be in compliance with ASME and TEMA ( NO ASME U Stamp Required).

I calculated the tubesheets using either UHX Part or TEMA and I got bigger thickness using UHX respect to TEMA.
In addition the UHX Parts requires to increase the thickness of the barrel adjacent to the stationary top tubesheet.
In conclusion, with TEMA......COST SAVING.

I would proceed using the thickness found with TEMA with the support of FEA analysis ( as admitted by ASME CodeU-2(g))for the untubed parts.

Do you think this is a correct approach???

Anyway, please suggest me what is the normal approach used with your AI for NTIW of fixed tubesheet heat exchanger.

Awaiting for your opinion I thank everybody in advance.

GENGI

 

[SnTMan]

gengi, I believe this has been discussed earlier, but my experience has been that you just ignore that the NTIW is not within the scope, and do calculations per TEMA and/or Part UHX (or, these days, both). Never had anybody kick about it.

Since you are not U-stamping, your TEMA / FEA approach should be OK, but you could always run it by your AI and client first.

Regards,

Mike

 

[eliebl]

Is there going to be AI involvement? You stated above that it is not code stamped.

 

[gengi]

Dear eliebl,

I confirm U-Stamp is not required but there is an AI.

rgds

 

[eliebl]

I have seen designs like these built using TEMA, Part UHX, or FEA as the justification. It really depends upon what your AI and client are willing to accept. There was a link to the ASME whitepaper on the development of Part UHX posted previously. You may want to go through this document to understand what the assumptions of UHX are and how it could be applied to this case.

EJL