Fatigue Assessment 2

[DSB123]

Hi Fellow engineers,
I would like your opinion on the subject of Fatigue assessment of Piping. For example if we were to undertake a Caesar Pipe Stress analysis of a system and then used the results from that analysis to undertake a fatigue assessment in accordance with say ASME VIII Div 2 (since we have cyclic loads other than just thermal) what is the level of stress we should use for the assessment. I ask since we know that the peak stresses in the piping system are twice the values reported in the caesar analysis so is it correct to use the reported values from Caesar (or any other pipe stress analysis program for that matter) to perform the assessment or should we increase the values by a factor of 2?

Regards

DSB123

 

[DSB123]

No opinions out there??

 

[BigInch]

Not yet.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[davefitz]

How do yu know the actual peak values are twice the Ceasar values, and how do you know the Ceasar program is correctly modeled?

Sounds like a pipe support or bumper is not correctly modeled.

One similar case ( recently published) could not understand why the Ceasar program did not predict the experienced pipe ratcheting ( TVA Bull Run). Then I advised the author that I had heard that unit had once restarted following a major turbine outage without removing the temp pins from the support hangars, whihc reuslted ina 12" creep growth of the steam lne and routine contact with fixed bumpers. Suddenly things began to make sense.

 

[DSB123]

BigInch,
Smart Answer..... But not very helpful!!

Davefitz,
I was talking generally, not specifically about a particular situation. I was trying to enlist people's opinion/open a dabate as to the correct approach. We know that the SIF's for tee's, bends etc for a B31.3/B31.1 pipe stress analysis are developed based on a butt weld in a straight pipe (which is nominally set to 1.0 - whereas the actual value is nearer 2). My question is whether it is correct to extract the stress results from a pipe stress analysis performed to B31.3/B31.1 and use directly in a fatigue assessment or not?? Should the stress values be doubled for a fatigue assessment?

Any constructive responses welcome.

 

[JohnBreen]

DSB123

I have been holding back because I am out of pocket at the moment and I do not have access to the resources I would want to reference for a comprehensive answer. However, I will offer a "general" comment or two or........

Well, of course the B31 Codes are in fact "fatigue based Codes" since the 1955 edition with the introduction of stress intensification factors. The B31 analysis using software that is "built around" the B31 Codes does not calculate true elastic stresses for the reasons that you cite. You cannot depart from the B31 allowable stresses as the Standard of comparison for the "equivalent stresses" that you calculate (with the use of appropriate SIF's). Now we say in the B31 Codes that if you have better information regarding SIF's than that in Appendices "D", you should use it (e.g., more appropriate SIF's). Those SIF's can be calculated using, e.g., FE/Pipe - but they are still SIF's and B31 rules apply. So you are using beam theory and B31 Code rules and comparing to the B31 allowable stresses - that is all appropriate. When you use the B31 rules you must apply ALL the B31 rules. If you are using B31 SIF's you are bound to all the rules of B31. You DO NOT increase the calculated "equivalent stresses" by any factor.

In B31.3 we say fatigue (due to internal pressure variations and due to thermal fluctuations "shall be considered". Of course the "pat answer" is to design the system with countermeasures to avoid cyclic loadings but we all know that is not always likely to be an option. Generally, the rules of ASME "Standards and Guides for Operation and Maintenance of Nuclear Power Plants", OM-S/G, Part 3, can be used for guidance. The ASME OM document provides a methodology to evaluate vibration using high cycle analysis and the methodology can be applied to other cyclic loadings and to any piping system.

When you opt to use other ("more rigorous") methodologies (the B31 Code tells you that it is permitted to apply "more rigorous methodologies") you must use all the rules of that methodology (to include the allowable stresses). The recent rewrite of ASME B&PV Code, Section VIII, Division 2 introduces the "structural stress" method (developed at Stanford) for evaluating fatigue in weldments associated with pressure components (as an alternative methodology) and I think the ASME B31 Mechanical Design Committee will be having a close look at that (don't hold your breath). This method is also described in API 579-1/ASME FFS-1. The method requires a finite element analysis of the component of interest. Stresses are linearized over the cross section and substituted into a fatigue master curve for welded joints. This method is far less mesh sensitive than other finite element-based fatigue analyses, and it has been validated with numerous fatigue tests of welded joints.

Of course, once you get into Section VIII, Division 2 analyses, you are no longer calculating B31 "equivalent stresses" and since you are now calculating "true elastic stresses" (yes, almost twice the B31 calculated "equivalent stresses") you are bound to the allowable stresses of Division 2.

I hope I have touched upon some of what you have asked.

Regards, John.

 

[BigInch]

DSB123, See. Now I'll bet your glad that I waited for John to check in.

John, perfect answer as usual.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain

 

[DSB123]

BigInch,
My point is why post when you did not contribute!!!Seems like a waste of time to me!!!

JohnBreen,
Many thanks for your constructive response. You basically concurred with my thoughts. It's always good to have a second opinion. The case in question is where a Contractor has used the results from a B31.3 analysis and then performed a fatigue assessment in accordance with BS5500. I was concerned about the stress values being used in the assessment.

It's always nice to read your responses to technical queries.

Regards

 

[BigInch]

Sorry. It was quitting time.

http://virtualpipeline.spaces.msn.com

"What gets us into trouble is not what we don't know, its what we know for sure" - Mark Twain