Allowable Stress for Pipeline Hydrotest Fabricated Blind 2

[McDermott1711]

Hi everybody,
I am going to fabricate a hydrotest blind from CS plate and stiffening structural steel and analyze it for maximum Von Mises stress by Ansys software. Would you please tell me that this stress shall be compared by which stress based on the ASME B31.8?
FYI: Location class is 3 and therefore test will be done by 1.5xOP.

 

[LittleInch]

Why? You can get perfectly good blinds and blind flanges to 16.5.

Every time I've seen a manufactured test blind it has been a disaster. I urge you to think again. In any point 31.8 is not your correct code IMHO, but 16.5

For info what size and pressure and what steel and thickness were you thinking of? Any drawings or sketches?

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[McDermott1711]

Thank you LittleInch for your reply:
This natural gas pipeline is 42", and line class is 900#, material of pipe is API 5L, Gr. X65.
The reasons are:
1- B16.5 supports up to 24".
2- Blind or cap is long-lead one and we past the deadline.
3- It is a common practice (Link) to fabricate a blind using thinner plate, stiffened from the outside.

 

[BigInch]

BPV VIII

you must get smarter than the software you're using.

 

[MJCronin]

Little inch, as always, gives good advice....

But are you going to design and fabricate a massive piece of steel, and only use it a couple of times ?

Unless you are going to use this component repeatedly......Why not rent ?

http://www.usaindustries.com/paddle_blinds.php

MJCronin
Sr. Process Engineer

 

[McDermott1711]

Thank you MJCronin for your reply.
1- This blind will be used more than once.
2- Please get back to my question, which was allowable Von-Mises stress for a given material (CS) under hydrotest situation.

 

[LittleInch]

Let's see if I've got this right.

Someone (maybe you) has forgotten to order a 42" #900 blind (not sure if this is a blind flange or a spade type). These are available to ASME B 16.48 and / or MSS-SP-44 ( above 24" they are the same thing). At 231mm thick [sic] it is a bit of a monster.

So now you want to fabricate something not as thick but add reinforcement instead (makes me think this is a blind flange), somehow weld this steel to the flange to stiffen it thereby introducing lots of forces and stresses, analyse this with a finite element program and then presumably present it to someone for their review and approval? The excuse that "We didn't have time / schedule to buy the proper equipment" hasn't much of a chance in court....

People may have been doing this and getting away with it for years, but it doesn't make it right and for this size and pressure it will still be a big thing. Line blinds are less of an issue than blind flanges, but still need to be designed properly and only when any failure is not a destructive one.

You're looking in the wrong place for any attempt to prove this thing (a sketch would be great to see). You need to use ASME VIII or 31.3 or one of the flange codes, not a gas pipeline code - Annex A of MSS SP 44 gives some guidance on how they did it, referring to ASME VIII and providing some stress values they used for many stresses, not just von mises.

I would be spending your time trying to find one of these blinds and get it to your location as fast as you can.

The real problem is that you'll do this, use it more than once and then think it's OK, right up to the point where it fails.

which country / region is this being executed?

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Gator]

Wait - 231mm thick? Even if there's a missing decimal point there, something seems wrong with this number.

Piping Design Central

 

[LittleInch]

Actually 231.8 mm

Remember this is a class 900 42" flange. At max op pressure you've got 3 MILLION lbs force on 24 bolts.

Even the weldneck flange is 206.4 mm thick, bolts are 3.5" diam. This is one big mother of a flange....



My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Gator]

Wow, that's a big flange.

Piping Design Central

 

[McDermott1711]

Please note that there isn't any flange in the line. This is an U/G pipeline carrying natural gas and all joints (including valve ones)are welded.
It seems if someone wand to make a test spool with stiffened (instead of an elliptical or dish-head cap) at the end, there isn't such a criteria for the allowable stress. Am i right?
What about ASTM A530 "Hydrostatic Test Requirements" which says "... Except as provided in 21.2 and 21.3, each length of
pipe shall be tested by the manufacturer to a hydrostatic pressure which will produce in the pipe wall a stress not less
that 60 % of the minimum specified yield strength for carbon and ferritic alloy steel pipe, or 50 % of the specified minimum
yield strength for austenitic alloy steel pipe."? Is it in accordance of the governing ASME B31.8?

 

[LittleInch]

A test spool with a stiffened end is commonly referred to as a flange and blind flange.....

I now think you're in even more dangerous territory if what someone (you?) is trying to do is weld on a flat plate at the end of the pipe and "stiffen" it with what are presumably some big bits of steel beam?

Have you ever wondered why high pressure pressure vessels all have dished ends?? What you're trying to do is design a pressure vessel so you need to use pressure vessel design codes, not pipeline design codes which are meant for design of pipe, so no effective stress applies in the same way.

Not sure what your point is about ASTM A530. Nearly all decent pipe will be tested in the mill to some fraction of SMYS for about 5 to 10 seconds. This mill test is simply there to find gross defects in the pipe manufacture and in terms of a hydrotest to B31.8 is irrelevant and worthless.

A sketch of this end plate would be really quite good to see....

Better still either buy a ready made end cap or a flange and blind to MSS-SP 44 / ASME B 16.48. You are messing with some massive forces here and it is goes even a little bit wrong could have huge consequences.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[McDermott1711]

Yes, LittleInch, you are right. This will be dangerous if I fabricate and test it without pre-calculation. But what I'm gonna do, is exactly, that calculation. FYI, a sketch of the idea is attached.
All I want to know is what is the allowable stress for the presumed materials (say A516 Gr. 70 or API 5L Gr. 70) which are used for fabrication of this blind.
Certainly, I'm not going to build it, if results of FE software show that maximum Von-Mises Stress is more than that allowable stress or I shall redesign it in such a way to bring stresses below the allowable stress.

 

[LittleInch]

WOW

That really is something else. I struggle to see how building that monster of a thing - it must weight several tonnes - is any quicker and easier than getting a 42" flange and blind, or a cap end but that's your call.

You are definitely into pressure vessel design here, not simple von mises stress and with all the welding going on you probably need to heat treat it as well.

Do the analysis and let us know how it goes - it would have been a bit easier to comment having this up front, but hope you got something out of the discussion...

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Duwe6]

McDermott, think about inserting this cap about 1 to 2-ft into the pipe. Old boilermaker's trick for hydroing an open header. The portion of pipe that extends outboard of your cap will be nozzle reinforcement. Further, the fillet weld reinforcing cap on the insert weld puts that part of the weld in compression, thus normally 'fatal' defects will be forced shut.

 

[TGS4]

It's taking a big leap going from an allowable stress basis and nominal stress levels in a hydrotest (your ASTM spec, for example), to an FEA. There is an entire section of the Boiler and Pressure Vessel Code (ASME Section VIII, Division 2, Part 5). It's not just a matter of keeping all stresses below a certain stress level (in an elastic stress analysis), because not all stresses can cause certain failure modes from acting.

So, to your original question: there is no easy answer. You would be best to consult an expert in applying FEA to pressure equipment.