Preload on B31.1 Code flanged joint

[jives]

I have a question on bolt pre-load for an ASME B31.1 Code pipe flange joint.

Should the total load on a bolt in the joint, including applied load from pipe internal pressure and pre-load, NOT exeed the allowable stress of the bolt material at temperature ? Or can total bolt load be safely brought up to 50% of Yield strength.

Also, what is the recommended joint safety factor (ie: ratio of pre-load to applied load), to make for a leak-free joint ?

Anybody out there have experience with this ?

Jonathan Ives
Project Engineer
Transalta, Alberta Thermal Plants
Duffield, Alberta

 

[TGS4]

Jonathan - I would highly recommend that you obtain and read ASME PCC-1. It will provide some guidance in flange bolt-up. That said, there have been many papers written that advocate maximizing the bolt pre-load.

I've done a lot of work on this, and I'm in AB.

 

[MJCronin]

jives...

On all of the ASME B31.1 and B31.3 work that I have ever done, (BTW, AARP sent me an invite ten years ago) every AE has developed a "field guide to bolted joints"

For piping systems these torque values were the same as those developed for Flexatallic gaskets many, many years ago...

See this link.....

http://www.flexitallic.com/pdf/designCriteria.pdf

-MJC

 

[tothepoint]

ASME PCC-1 details flange assembly procedure and, as guidance, suggests bolt preloads equal to 50% yield of a B7 fastener. Not real helpful.

Assembly preload is optimized by knowing the stress limits of the three primary components of the joint: the flange, the gasket and the bolt. Bolt torque tables do not take all of these factors into consideration, so be careful using them.

 

[MJCronin]

tothepoint,

You are correct,.... however, for the hundreds of flanged joints in piping systems of large complex projects, there must be a simple method.

Again, for piping and pressure vessels, the only practical method to get a large job completed and to ensure reasonable protection against leakage, is to perform the correct tightening sequence and to use some type of bolt torque tables.

Bechtel, Fluor, Duke et. al. all use a variations on this method. A "custom preload" for each piping joint does not seem practical to me...

-My opinion only

-MJC

 

[TGS4]

tothepoint - just to play devil's advocate for a moment, please explain to me why you would need to consider the stress in the flange during operation? Is there a stress-related failure mechanism that goes on in the flange? Who cares if there is a local region of plasticity in the flange? How detrimental is permanent deformation (and can you quantify what level of permanent deformation is unacceptable)? I'm asking these questions both from a strength-of-materials perspective as well as a leakage perspective.

(Many of the answers to these questions are in this paper

http://store.asme.org/product.asp?c...-ICPVT11TC-3&product_id=PVP2006-ICPVT11-93075

but I want to hear some other opinions).

 

[jives]

I'm not worried about flange stress. The problem here is that the bolt material on the existing flange is unknown. PMI testing indicated that the closest match was A193 B16. So I assumed 85,000 yield strength for the bolts to be on the safe side. Then I calculated the load on the bolts from the flange internal pressure of 3550 psig. Then I considered the minimum crush required on the gasket, which necessitated minimum 30,000 psi bolt stress. We torqued the bolts to gain about 38,000 psi bolt stress. This gave us a joint safety factor of just over 3 (ie: force on bolts from torquing exceeds force on bolts from the internal pressure by a factor of about 3).

Jonathan Ives
Project Engineer
Transalta, Alberta Thermal Plants
Duffield, Alberta

 

[tothepoint]

TGS4 and Jives:
You're "not worried about flange stress", and "how detrimental is permanant deformation"? You're kidding me right? If you avoid gross plastic deformmation and excessive flange "bending", your flanges remain flat and you have a much easier time sealing these joints; time after time after time. Excessive bending of the flanges during assembly reduces the gasket contact area and creates high stress areas towards the OD of the gasket. Higher gasket stress is great- until you exceed the maximum limit of the gasket (which by the way is often lower than you think, and for some materials reduces with temperature) and the gasket fractures. Flange bending is even more of a challenge with spiral wound gaskets. In this case the flange bends and lifts right off of the spiral winding.

Avoid plastic deformation of the flanges and your gaskets work more effectively, more consistently and you're not replacing flanges during shutdowns.

 

[TGS4]

tothepoint - the paper that I provided a link to does a pretty good job of refuting a number of your comments. I suggest that you buy the paper, read it, and then respond.

I appreciate what you are saying about the flanges rotating and the contact stress in the gasket increasing. However, most spiral wound gaskets have centering rings which make contact with the sealing surface, thereby moving the rotational center toward the bolts, and limiting additional rotation. Besides, I would be surprised if many flanges out there weren't sealing solely on the centering ring.

As far as gross plastic deformation is concerned, I might agree with you, but what constitutes "gross". How much of the flange needs to be plastic before the deformation can be deemed "gross"? Would you recommend that ALL plasticity be avoided - I don't think so...

 

[tothepoint]

TGS4:
I read the paper months ago. Goes to show that you shouldn't blindly believe everything you read. I'm providing input from field experience not theory. You're right about the spiral wound gaskets- rotate the flange and you end up sealing (?) on the guide ring (not adviseable- that's why they're made with a sealing element). While the metallic outer guide ring in a spiral wound will "take" this high stress at the OD of the raised face, many sheet type gaskets cannot. They shear at this interface and leak. An avoidable leak had gross plastic deformation/excessive rotation been avoided during assembly.
Again, I agree with your comment on GPD. There's no "standard" for this. Bottom line- if you're visibly bending flanges during assembly, and they're staying bent- you're applying too high of a preload.