Putting a partial penetration tap in a blind flange 1

[jgibbs22]

Hey Everyone:

Had a question come across my desk the other day I was hoping to get some input on! We are using a 16" bolted blind flange wiht a 1.5" hole drilled in it as the head of a pressure vessel. On the inside of the vessel, I need to attach a 150# 1.5" flange to the opening but my Engineer wants to drill out roughly 50" of the flange thickness and tap it to screw in some studs for the 1.5" flange. My question is does the code consider this reduction in the flange thickness or not because your fill it in with the stud? I recommended that he only drill and tap 25" for the studs and weld the studs into place, but I was hoping to get some more opinions on this. If you need other details let me know!

2) When you choose your allowable stress from Div II, part D for a flange head, can you use 1.5 x the given amount because your are really dealing with a bending stress and not a membrane stress when it is bolted on? Or is that logic faulted cause it is going against a value given by the Code?

 

[jte]

Jason-

You need to consider the reduction in thickness if you're drilling holes - removing thickness.

No, you canot use 1.5S for the equations even though you recognize that there is a bending stress involved:

Subject: Section VIII, Division 1 (2004 Edition, 2005 Addenda); UG-34, Unstayed Flat Heads and Covers

Question (1): Is it permitted to use 1.5S, as provided in UG-23(c), in lieu of S in the equations of UG-34 in order to determine the required thickness of flat heads and covers?

Reply (1): No. For those configurations for which it is appropriate to use an allowable stress equal to 1.5S, the 1.5S stress multiplier has been built into the C factor.

How 'bout a compromise solution: Weld some studs (check with matl's / welding engineer) to the blind flange rather than threading them in.

But... your approach of welding a standard flange on the inside is probably cheaper and easier and likely more reliable. That's the solution I'd choose.

jt

 

[jgibbs22]

JT, thank you very much for the response! My follow up question was going to be just that, could I simply weld the studs onto the flange instead of threading them. I asked our welder and he said he would like to drill out a 1/4 of an inch and set the stud in before welding so it is a little easier to weld. Would I still need to count this as a reduction in thickness or by setting the stud in the hole am "re-adding" the lost material?


Also, the gasket that is being called out is a spiral wound gasket with a y of 10,000 and m of 3. With this gasket I am not getting the blind flange to pass the UG-34 equation for thickness. Are there any other gaskets out there you could recommend that are made of 304 and can be used with liquid nitrogen temperatures?

 

[SnTMan]

jgibbs22, it is a common practice to blind-tap holes in a head, etc. UG-43(g) gives required tap depth and though not explicitly stated as such, I use UG-43(d) for guidance on thickness remaining below the hole.

Regards,

Mike

 

[jgibbs22]

Mike thanks for the reply!! One question though, for UG-43, I am slightly confused on the wording. It says

"Drilled holes to be tapped shall
not penetrate within one-fourth of the wall thickness from
the inside surface of the vessel after deducting corrosion
allowance, unless at least the minimum thickness required
as above is maintained by adding metal to the inside surface
of the vessel."

So does that mean you can only take away 1/4" of the material of you were starting from the outside surface of the shell? In my case I am actually taking away material from the internal surface so it seems like it slightly complicates this, unless I am over thinking it.

 

[SnTMan]

I would say, yes you are

 

[jgibbs22]

Why thank you for that lol!! Problem is removing the material from the inside increases the max stress the inside edge sees, thus increasing the stress throughout the thickness....

Does the Code mean you can tap up to 1/4" of the total thickness?

 

[doct9960]

jgibbs22,

I would not recommend welding of stud bolts because it may alter the mechanical properties of the bolt. Ask a material specialist before proceeding with this option.

Can you give more detail as to what component is attached to the inside of the blind flange? Why does it have to be bolted? Are there space limitations?

 

[jgibbs22]

attached is 1 1.5 inch 150# Raised Face Weld neck flange that attached to a pump spacer. This was a design issue that popped up last minute that just got tossed my way and I was looking at a couple possibilities.

 

[SnTMan]

jgibbs22, note the use of the words "minimun required" in UG-43(d). Also see UG-23(c) regarding localized stresses.

Personally, if I understand the construction and assuming you engineer wants to drill 50 percent of the thickness, I wouldn't have any problem with it.

Metallurgy aside you definitely don't want to weld the studs. What if you wipe one out?

Regards,

Mike

 

[jgibbs22]

Mike, what do you mean by "wipe one out"? Have not heard that before. What makes you say drilling out 50% of the thickness (so 1.44" gets reduced to .72") would be ok? my looming issue will still be telling the AI I have modified a 150# 16" blind flange so I can not just say it meets 16.5, and then a UG-34 calc shows it is not thick enough. Have not had to deal with an AI on a flange so pretty new to this world...

 

[TJOrlowski]

Mike thanks for the reply!! One question though, for UG-43, I am slightly confused on the wording. It says "Drilled holes to be tapped shallnot penetrate within one-fourth of the wall thickness fromthe inside surface of the vessel after deducting corrosionallowance, unless at least the minimum thickness requiredas above is maintained by adding metal to the inside surface

This paragraph assumes tapping a hole from the outside, but it means that if the minimum thickness of your plate is 1" + 1/16" of corrosion allowance, you must have 5/16" of thickness left after you drill your pilot hole. Also, check UG-43(g) for required minimum engagement of your 1/2" studs. The engagement needs to be at least 1/2", but no more than 3/4" (and these values don't account for the metal lost due to drilling a pilot hole).

-TJ Orlowski

 

[TJOrlowski]

my looming issue will still be telling the AI I have modified a 150# 16" blind flange so I can not just say it meets 16.5, and then a UG-34 calc shows it is not thick enough.

Once you've ironed out bolting your WNF to the inside of your head, this will, indeed, be your next wrinkle. If you haven't secured any material yet, you may consider using Appendix 2 flanges in lieu of 150# ANSI's - you won't be restricted by the max available cross sectional bolt area of an ANSI, which should make it easier to use a spiral wound gasket.

If that isn't an option, solve for m & y values using your ANSI flange, then consult your gasket supplier for a suitable gasket for your application with the m & y values you need.

-TJ Orlowski

 

[SnTMan]

jgibbs22, "wipe out" means it strips, breaks, corrodes or other bad things that happen to threaded fasteners.

On further reflection, you are no doubt right, you don't have a B16.5 part anymore, in which case you would have to prove it. So calculate it and make one, if necessary, can't be that big a deal.

You could either treat it as a loose ring per Appendix 2, or a flat head with opening per UG-34, either way I would disregard the stud holes in these calculations, but make them to comply with UG-43.

Regards,

Mike

 

[GenB]

have you consider a pad inside your blind so you can thread and make your bolted attachments?
but question: will that void the B16.5?

 

[jgibbs22]

Gen thanks for the response! PRoblem is I have already voided 16.5 when I tapped the 4 small NPT ports into the blind flange. Far as I can tell, 16.5 approves tapping a single hole in the center up to a certain diameter, but nothing else, which is why I am getting into the issues I am in now.

As I had worries about, and I believe someone in the string confirmed, you can not just take a 16,5 flange and modify it out of the scope of 16.5, and then still say I dont need to check the thickness because it is a 16.5 flange....

I could probably add pad, but at that point I am probably going to just make a new part, just really hoping I could work with what I have!

Thank you all for the great responses!!!

 

[chaulklate]

I could be wrong...but I am under the impression that we are not supposed to weld (not even back welding a tap) stud material to the pressure envelope as the stud material is not recognized as 'weldable quality' steel and has not been assigned a P-number. Stress values given in Table 3 are for bolting purposes only. If you are going the weld route, I suggest a poison pad first.

 

[GenB]

notice that B16 flanges can be over rated depending of the pressure and temp you are designing to.
if lower pressure and temp design, you may have extra T. on the flange.
you may now void B16 and calculate the flange as a flg made of
plate.'hope that will fix your problem. genblr