Flange leakage design continuation 1

[vesselguy]

Hi all,

The interesting post on flange load design got me thinking about the existing ASME/ANSI B16.4/B14.3 flanges and whether they would pass the rigidity factor test in Appendix 2. Has anyone done a project where they did a rigidity check of off-the-shelf ANSI flanges with piping loads, hydro, and bolting loads?

Also, I am writing up a standard right now and I wrote this requirement in for custom flange design.

"4.3.9 Custom designed flanges shall be checked for likely hood for in-service leakage by calculating the flange rigidity factor (J) per Appendix 2. Flange thickness shall be increased if J < 1.0. The applied moment (Mo) in the equation shall include the combination of all in-service loads including external piping loads."

What do some of you experts think about what I wrote? Critique is welcome to help me improve on this requirement.

 

[eliebl]

Prior to the introduction of the 3.5 design factor the requirement to check rigidity was optional. When the code case was written allowing a 3.5 dign factor one of the requirements was to consider flange rigidity. When this code case (I forget the number) was written into the code the check for flang rigidity was not mandatory until either the '05 or '06 addenda. The rigidity check is not mandatory.

I believe that your less than sign should be a greater than sign. If the calculated rigity exceeds 1 then the thickness must be increased. See 2-14(c)

EJL

 

[vesselguy]

eliebl,
Thanks for the sharp eye. You're right, it should be a > sign. I also typed the flange specs wrong, it should read ASME/ANSI B16.47/B16.5.

 

[SnTMan]

vesselguy, I may be on a little bit shaky ground here, but the applied moment you mention used in the rigidty calculation is the internal moment(s) calculated per Appendix 2 and is the result of only pressure and bolt loads applied to the flange. (I am assuming we are in Sec VIII, Div. 1 here.) External loadings will not be easily incorporated into this moment unless converted to equivalent pressures.

I don't know the purpose of the spec you are writing, but if I was a fabricator reading your para. I would say first "Its already mandatory for me to meet flange rigidty, why would anybody put that in a spec?", and second "I wonder if this customer has a favorite method to include the effect of external loads, or does he just want to second guess whatever method I choose?"

This is not intended as unfair criticism of what you are trying to accomplish, its just that my experience in the fabrication end of this business has exposed me to all kinds of unclear language in specs that cause no end of interpretation or misunderstanding. So, my point is, if you are going to put something like this in a spec, be as clear and unambiguous as possible, even to the extent of giving equations, etc.

Hope I haven't gone off an a tangent again...

Regards,

Mike

 

[vesselguy]

Hi SnTman,
No worries about any "unfair criticism" coming my way. I'm used to it; I get'em at home all the time. LOL.

Yes, my intention for the paragraph is to be clear to the fabricator that paragraph 2-14 in a Mandatory Appendix is to be applied. Reason I say something about this is on one project, a fabricator argued that paragraph 2-14 is "optional" and they want cost extra to increase the flange thickness on many large diameter custom design flanges on a Vac Tower. This fabricator raised a stink and gave our vessel engineer a hard time. After that, the standard included this pargaraph to make sure this kind of c*&% don't happen again. What you and I sometime think is common sense and deemed part of the code, may not be the same with another person. Kinda like shoving democratic ideas to countries that don't want it.

Yes, in another paragraph I have defined the method to convert the external loads to equivalent pressure end load H. Just wanna make sure the guy considered external loads into his calc.

 

[SnTMan]

vesselguy, understand about "common" sense. And if you need something that goes beyond the Code, such as including external loads in the rigidty calcs, much better to include the requirement in the spec. If you've got everything spelled out, unambiguously, as a fabricator I would have no problem.

Good luck with it.

Mike

 

[TGS4]

SnTMan i correct that the Mo moment in the rigidity calculations are the flange "internal" moment, and getting the external bending moment and/or external axial force included in the equation would be difficult.

As far as how I would go about including a calculation methodology, I've personally used EN-1591-1, but I would consider that to be overly onerous on a manufacturer to perform. There is no commercial software that does the calc, and it took me weeks of Mathcad work to implement it myself. I would not recommend it to anyone...

 

[SnTMan]

TGS4, I am not familiar with EN-1591-1, I have only used the methods mentioned in that other thread, similiar to what was called the "Kellogg" calculation, and as I said, I believe there is a similar reference in ASME Sec III, to convert to equivalent pressure.

Digging thru my folders full of paper I find "ASME Section III, 1977, NB-3647.1"

It's the old Peq = 16*M/Pi*G^3 + 4*F/Pi*G^2

Regards,

Mike

 

[TGS4]

Mike, are you going to the PVP conference? There will be some good papers on flanges. I downloaded the final program from the website -

http://www.asmeconferences.org/PVP07/

 

[SnTMan]

TGS4, no I never seem to plan on going to these things, I always feel it's probably over my head. Maybe I should try one sometime. It also seems there is too much work to do to be out if I don't have to.

I missed out on some software training this week because of a field deal that got postponed probably till NEXT week, when the conference is.

Ah, the life of an engineer

Bring us back some good stuff, TGS4. And enjoy San Antonio. Never been there.

Mike

 

[vesselguy]

TGS4,
I use the same method as SnTMan mentioned above. Never seen teh EN-1591 either. Anyway to get a free copy of it for light reading before bed time?
Yeah, I thought about asking the boss if I can go the the conference next week. But 4 days is costly and too much. Is there a way to get a copy of that flange paper for a look see?

 

[hartsa]

TGS4,
There are commercial softwares to performing calculations of flange connection according to EN 1591 (EN 13445-3 annex G). Here's links to few websites:

http://www.lv-soft.com/software/fachbereiche/apparate/index.htm

http://www.ohmtech.no/index.php/ohmtech/news/alternative_method_for_flange_design

We are using VVD-program (visual vessel desing by Ohmtech). The newest module includes flange calculations according to EN 1591.

Calculations according to EN 1591 are indeed very difficult to perform without any software because there are lots of iterations in this method. When I studied this method, I used MathCad. I got 40 pages of calculations for just one flange connection (I didn't know how to do loops in MathCad). Then I did Excel program with lots of VBA code. It worked quite nicely.

Nowadays we calculate our vessels with VVD. It uses formulas introduced in EN 13445-3 (Unfired pressure vessels). With VVD we can calculate almost every components in our vessels, eg. shell, ends, nozzles, flanges and so on.

Vesselguy,

Here's few links for more information about EN 1591:

http://www.sealeng.com/ase/index.html

http://www.unm.fr/en/general/en13445/

(look for Background to the rules in Part 3 Design. Click the acrobat icon to open the pdf-file. Then look for annex G.)

 

[innovation2]

Isn't there a code case that says that flange rigidity does not need to be required. I am pretty sure that we will see this so called requirement removed from the code books in coming years. Personally I think it is a bit of overkill.

 

[eliebl]

The code case is intended for manufacturers who have flanges in stock which were calculated to previous editions of the code to be able to be used until they re-calculate and meet the rigidity requirements.

That code case is not acceptable in one jurisdiction as the rigidity calculation was mandatory for flanges 40" diameter and above. It was part of the 3.5 design factor code case but for some reason was not rolled into the code with the design factor change.

EJL

 

[gr2vessels]

For flanged connections subject to external loads, Dennis Moss provides a formula in Pressure Vessel Design Manual, Appendix G, formula 6 (same as provided by SnTMan above). Also, many of the Compress like sofware do calculate and include the equivalent pressure in the design of the flaged connections.
TGS4, any news from the conference?
Cheers,
gr2vessels

 

[TGS4]

Yes, I have news from the conference. The paper that I was anticipating stated that for Class 150 and Class 300 flanges, using the equivalent pressure method yielded reasonable results using a limit of TWICE the ASME B16.5 rating pressure PROVIDED that a sufficiently high assembly bolt stress/load was employed. (Unfortunately, the papers are not yet for sale at the ASME store...). This was applicable to flanges using spiral wound gaskets only. I appears that each combination of gasket and assembly bolt stress might yield a different conclusion. The results for the higher classes showed that the limit was somewhat less than 2X rating pressure.

That said, I would be very comfortable (for initial design purposes) using a value of 1.5X rating pressure when using the equivalent pressure method.

Since the failure mechanism of flanges is leakage, I would certainly apply the rigidity check (or at least impose a limit of 0.3 degrees of flange rotation at the actual install bolt loads) to all flanges. It has been demonstrated numerous times that excessive rotation is a sure-fire way to ensure leakage. It is also very difficult to improve rigidity after the fact.

My 2 cents.

 

[gr2vessels]

I could not afford the cost of ASME VIII re-write just as yet, but does anyone know if this "equivalent pressure" issue has been addressed in the new edition?
On a different note, while browsing the ASME PVP 2007 website, I noticed that the chairman of the executive commitee of publications is employed by the Dept. of Mechanical & Astronautical Engineering at the Naval Postgraduate School in California. Do they have a High Speed Railway Dept.?
Cheers,
gr2vessels

 

[TGS4]

Nope - not even addressed.

The flange design rules are due for a serious overhaul, but the Code Committee responsible for flange design (SWG-BFJ) did not get Div. 2 rewrite marching orders until too late. There was one presentation at PVP07 about this, but the development of rules are extremely complicated, so it may be a while...

Not a specific high-speed rail dept at NPS, but then again, when has the navy been interested in railways?

 

[gr2vessels]

Since the creation of Aeronautical department, I imagine...
Cheers,
gr2vessels

 

[TGS4]

Naval aviation is pretty big business (something about force projection, etc etc) not to mention the aeronautical issues with sub-borne tactical nukes. That this is pretty off-off-topic...