Single shell course?

[KLee777]

I'm reviewing a calculation of an existing vessel that was done in Compress. The existing vessel has four shell courses. The designer input the vessel as a single shell course over the OAL of the vessel.

Which is the proper method? I suspect entering each shell course separately, but I keep talking myself out of it.

Any advice?

Thanks.

 

[TomBarsh]

I don't think there's really a "proper" way. Either method can be used. Certainly it's easier to enter only a single cylinder versus worrying about where to located girth seams and multiple sections. There are pro's and con's to each approach. In the end, however, it's probably best to try to spot girth seams at appropriate locations.

ahhh, I just re-read your message. This is an existing vessel so the locations of the seams are known. This changes things a bit.

Remember that in some cases the radiography performed on the circumferential seams may affect the joint efficiency permitted by ASME Code for the longitudinal seam (or seamless vessel section). See rules of UW-12(a) for full RT (and UW-12(d) for seamless) and rule about UW-11(a)(5)(b).

Thus by ignoring the existence of the girth seams an incorrect value for joint efficiency might be applied when determining required thickness based on circumferential stress. Inspection of the data report for actual radiography performed can give some guidance here.


Tom Barsh
Codeware Technical Support

 

[KLee777]

Hi Tom, thanks for your response.

I have performed my own Compress calculation, and applied the joint efficiency as noted on the U-1 report. In this case, that happens to be spot radiography for the shell, and 100% radiography for the heads.

Since the locations of the weld seams are known, I have built the model using the four shell courses. I believe this is more accurate since the circumferential joint efficiency can be applied properly.

I have found other errors in their report, so I will make recommendations accordingly.

Thanks again for your help.

Kelly