Heat Exchanger body flange

[vesselguy]

Hi,

I will be starting my research on this topic but thought I'd get some ideas from some of the more experienced people out there.

I'm told we got a shell & tube exchanger that has its body flanges drilled with 2 centering pin holes (180° apart) between bolt holes. Nothing unusual about that. Problem is the centering pin holes are so large that, when looking from the face of flange view, the remaining metal between a flange bolt hole and the centering pin hole is only 1.5mm (typical both sides of the centering pin hole). I hope describing this in a manner people can understand. I'm trying to reason it out whether or not this is a problem or not that needs rectification.


Honestly, its been a long time since I did a flange calc. But, I recall the cals don't take into account of how much metal is there transverse of the cross sectional area. A flange is primary designed for bending and a pressure through its cross section at the bolt hole.

Anyway, like I said, I'd be doing some calculations/research into this today, but I would like to hear some input/experience from other experienced designers on this.

Thanks.

 

[Christine74]

I don't think there's any standard flange calculation method that you could use to evaluate the stresses in this particular flange, so all you're left with is FEA. ASME VIII, Div. 1 Appendix 2 doesn't even consider the effect of the material lost in drilling the bolt holes, so it wouldn't be of any help to you.

Appendix 2 is actually quite conservative, in my opinion. Many exchangers that were designed with a maximum allowable stud stress of 20-25 ksi have been torqued to a much higher stud stresses, but without any ill effects.

-Christine

 

[waskillywabbit]

In my opinion, sounds like bad design work at best, recipe for a failure at worst...restating the obvious. Almost sounds like an "afterthought" that was added in after the design work was done, obviously for centering things.

With so little metal on one side of the bolt hole I would be worried about crushing it out when torquing things down and getting improper seating on the flange/gasket, leading to leak(s) and/or a failure.

I'm not sure what "calculation(s)" you can do to help your current flange(s), but it sounds like some redesign of the body flanges might be in order to eliminate the sloppy centering pins and incorporating some that are properly placed. Perhaps even some filler weld metal and incorporating properly sized centering pins that don't weaken the bolt holes.

I've only had about 2 sips of coffee this morning so far, thus I am not held responsible for my ramblings.

Brian

 

[vesselguy]

Christine,

You are right about Appendix2 in that it does not take into account of the metal adjacent to the bolt hole. Same as other flange calcs. It is times like this that makes one pay attention to the basis of the trusted cook book formulas that we all have been using all these years without thinking.

Mr. Wabbit,
Elmer Fudd says it was not a last minute design thingie. Centering pin is a pretty standard design in body flanges on S&T exchangers, but this time the damn pin hole diameter was too freaking big. Eventhough the EPC specified it, the Italian Fabricator should've threw up a red flag to us to give us a heads up before drilling the damn thing. Ahhh..

 

[SnTMan]

vesselguy, your pin holes are not thru holes are they?

Mike

 

[vesselguy]

SnTMan,

Yes, they are thru holes.

 

[SnTMan]

Well, that's unusual.

 

[CodeRef]

I recall a similar issue, but with smaller holes. The issue was resolved (as an afterthought) by including the design as part of a UG-101 proof test. Initially, the proof test was required solely for the flat head/body flange openings, that were spaced closer than allowed by Section VIII Div 1.

When we realized the flange guide pins/thru holes were not considered by formulae in Appendix 2 of VIII Div 1, we made reference to them in the proof test package and satisfied the Code requirements in that fashion.

Code does allow proof test without destuction of part and reuse of the part provided yield is not exceeded and permanent deformation is not exhibited. Gaskets failed long before that anyway.

 

[alamin]

It is obvious that the locations for centering pins (dowel) were poorly chosen during design phase. However, all is not lost. You have options on your plate; such as, weld fill the centering pin holes, drill new hole for centering pin by staggering (upward) between the two adjacent flange bolt holes and carry out req'd NDT as req'd by the code, or you can just do away with the centering pin. Centering pins only make it easy to locate during body flange bolt-up but it's not a must have. This was a recent feedback from the work crew in the plant I work where there are heat exchangers with 2m dia body flange.

Cheers,
Amin