Flange Load - ANSI Check?

[pepeuy]

Hi community,

I've been reading about flange load calculations all over the site (ex:

http://www.eng-tips.com/viewthread.cfm?qid=191806),

but I have a doubt about the equivalent pressure method used in Autopipe and stated on the Kellogg "Design of piping systems".

The Autopipe software clarifies (haven't seen it on the Kellogg) that the force used in the equivalent pressure calculation is only for tension components, it is reasonable since it will tend to open the joint and produce leakage. My problem and doubt is related to that equation.

On my system the external force is of compression, and on Autopipe the ANSI check syas that the flange is not acceptable since the equivalent pressure is greater than the flange rating; the compression force isn't used at all as stated in the help documentation and results.

Doesn't the external compression force aid in closing the joint and thus reducing the possibility of leakage? Is this a correct reasoning?

Best,
José

In Vino Veritas

 

[C2it]

The Kellogg method is a very crude, very simple check and it is abused more often than not by Autopipe and others. Basically if your flange passes uder that check, jolly good. If it fails that way, think again, don't change anything.

It is silly to expect a simple one-liner equation to tell you anything about a body under such complex loading and with complex geometry as is a flange.

If you want a half-way decent answer, use ASME VIII Div 2 or preferably EN 1591 to do your flange analysis.

 

[DSB123]

C2it has given the correct response again. Caesar II program uses either the Equivalent Pressure approach or the ASME leakage check but both are simple assessments. As C2it says the better assessment would be to use ASME VIII Div 2 or you could opt for a FEA assessment if you want to spend some money!!!

 

[XL83NL]

Here's what I would do. Go to the AutoPIPE pipe stress analysis forum, run by Bentley. They may provide you a better in depth answer as to how AutoPIPE deals on this subject specifically.
Furthermore, AFAIK the EN 1591-1, albeit very complex (and still containing a s***load on errors and typo's), is the current best method for flange analysis.

Unfortunately, Im not that in depth educated on flange leakage methods, However I recently picked up this thread on the linked ASME-PVP Pressure Vessels and Piping Division group.
Few guys who have considerable background on BFJ analysis (incl. Warren Brown) stated that both the VIII-1 and VIII-2 methods have nothing to do with leakage. Not saying that that was (in)directly said by others above, but one may draw that conclusion.

 

[pepeuy]

Thanks for the answers.

I have been reading the EN 1591 and as I reasoned, it takes into account if the external axial force is of compression or tension.

I still can't figure out why autopipe only takes a value for tension. I will have to do some research on the bentley forums.

On my design using autopipe, I have a 10% larger equivalent pressure than the flange rating. There are 3 more steam generators on site, one of them has a pipe design (same materials, diameter and schedule, operating pressure and temp) and dimensions similar to the one I'm dealing with. I've analyzed that pipe, and to my surprise the equivalent pressure is twice the flange rating. This pipe and steam generator had being working for over 7 years without any problems.

So I will take the new design as satisfactory without further calculations.

Anyway I will study flanges in greater detail for a future design.

Best,
José

In Vino Veritas

 

[TGS4]

I showed in a paper a couple of years ago that 1 x rating pressure as a comparison to internal pressure + equivalent pressure was excessively conservative. In fact, I showed that, provided the initial bolt load is appropriate, you could easily go to 2 x rating pressure.

The key to leakage is not a design check, or any design number, but the ASSEMBLY!

 

[racookpe1978]

True. But that (2x limit) does assume that the flange in question is properly supported and that the "closing force" across the flange is not used to lever pipes around during assembly and to correct pipe fitting errors and bad welds.

 

[TGS4]

Racookepe1978 - completely agree. I would add those loads in calculating the equivalent pressure (assuming, of course, that you could calculate that a priori).