Reverse flange weld size - Compress / ASME VIII-1 (2019)

[marty007]

It's been a long while since I designed equipment utilizing a reverse flange. We use Compress for our design calculations, and I was surprised by the specified weld size they came up with. I brought this up with Compress and they referenced a code paragraph that I'm not sure applies. I'd love to hear your perspectives.

We are using a loose ring type reverse flange per Fig 2-13.2 - with an attached shell thickness of 3mm.

Per Appendix 2 paragraph 2-13(a)(2):

The shell-to-flange attachment of loose ring type reverse flanges may be attached as shown in Figure 2-4, sketches (3a), (4a), (8), (9), (10), and (11) as well as Figure UW-13.2, sketches (c) and (d). When Figure UW-13.2, sketches (c) and (d) are used, the maximum wall thickness of the shell shall not exceed 3/8in. (10mm), and the maximum design metal temperature shall not exceed 650°F (340°C).​

Per Figure UW-13.2 sketch (d), the required weld sizing is: a+b shall not be less than 2ts. In addition 'a' shall not be less than ts, and there is a limit on the setback from the the outside face of the reverse flange. If the reverse flange diameter is equal to the shell OD, then a=ts, so the requirement becomes b shall not be less than ts. In our case with a shell thickness of 3mm, this would mean 'b' shall not be less than 3mm.

Compress however is resulting in a required weld size 'b' of 6mm. After contacting their customer support, they referenced paragraph UW-13(e)(2) which states:

(2)For flange rings of bolted flanged connections, as shown in Figure UW-13.2, sketches (m) and , the sum of a and b shall be not less than three times the nominal wall thickness of the abutting pressure part.​

So here's my question... paragraph UW-13(e)(2) specifically mentions Figure UW-13.2, sketches (m) and , does it also apply to Figure UW-13.2, sketch (d), when it used for the attachment of a reverse flange?

(I know the sketches in the code are never to scale in any which-way, but as a point of interest I measured the weld size in Figure 2-13.2 relative to the drawn shell thickness. That figure would suggest that weld size 'b' from Fig 13.2, sketch (d), would equal ts...).

Thank you!
Marty

 

[SnTMan]

marty007, I have to admit I am not very sophisticated when reading the Code. I also have to admit I don't always implicitly trust the software.

So first, not being very sophisticated I tend to interpret the Code as it reads. In your case I'd interpret that sketches (m) and do not apply, also taking into consideration that the language in 2-13(a)(2) restricts use of details (c) and (d). To me, it is similar to rules in UG-36(c)(3) exempting a larger opening in a thinner shell than in a thicker one.

Second, the designer has responsibility for the design, not the software. Now if I were writing the software I might chose to just apply the blanket rule in UW-13(e)(2) rather than apply the tests in 2-13(a)(2). Let the user overrule it if he wants

Regards,

Mike

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[Geoff13]

At the moment I can only get my hands on my 2017 Code, and paragraph UW-13(e)(2) does not include the words "sketches (m) and ". Any possibility Compress just hasn't caught up with the latest wording?

Sketches (m) and look like integral flange weld details, whereas (c) and (d) look like loose type welding details. Limiting the UW-13(e)(2) rules to (m) and seems to make sense.

As SnTMan notes, you have the final responsibility, but you can point at the change in the Code requirements as your justification.

Geoff

 

[SnTMan]

Near as I can tell, the language of UW-13-2(e) is the same back thru the 2004 Ed. That'd be kind of a long time to inadvertently overlook a change in the Code. I wonder if other software treats this differently?

The problem with sloppy work is that the supply FAR EXCEEDS the demand

 

[david339933]

2019 has "m" and "n"....Compress has put out a few updates for the 2019 ed. Hard to believe they would miss it.

 

[marty007]

Thank you all for your responses.

Geoff13, I think you've accidently stumbled across a key change in the code wording! In 2017 paragraph UW-13(e)(2) was:

(2) For flange rings of bolted flanged connections, the sum of a and b shall be not less than three times the nominal wall thickness of the abutting pressure part.​

In the 2019 edition, specific reference to certain Fig 13.2 sketches was added:

(2) For flange rings of bolted flanged connections, as shown in Figure UW-13.2, sketches (m) and , the sum of a and b shall be not less than three times the nominal wall thickness of the abutting pressure part.​

Final responsibility of course falls on me, I've accepted my fate in this regard... I just want to make sure I'm following the rules of the code without being forced to over-weld the flanges due to potential software company misinterpretation. If I had just blindly followed the software, we wouldn't be having this discussion!