B31.1 sustained stress in Autopipe. 3

[papiesz]

Dear all

Anyone knows why Autopipe when calculating sustained stress according to B31.1 uses corroded wall thickness by default. This is of course more conservative but ASME B31.1 para 104.8.1 states that nominal thickness should be taken into calculation. Also Caesar II takes nominal thickness

This is my first job in power piping and I haven't time to look deeper into the code so I'm little confused. Could someone tell me what is the common practice in power industry or what piping (maybe live steam) require more conservative approach.

Regards

Pawel

 

[mizzoueng]

B31.1 para 102.4.1 states that you should include the corrosion into the analysis but it is up to the designer to decide if the wall thickness is to be increased to account for it.

THere is no other mention in Part 1 about including corrosion, at least not that I have read yet. I would assume that Bentley used the minimum wall thickness instead of the nominal thickness as a "safety factor" so as they are not liable for failures when you say there is corrosion.

You could post this question to Bentley on their forums, they usually get back to you within a few days.

 

[richay]

But B31.1 doesn't say to compute flexibility stresses in the corroded condtion - as B31.3 does in 302.3.5. This is why CAESAR II doesn't (by default) corrode a B31.1 job.

If you want to consider corrosion in CAESAR II in a B31.1 job, go to the "SIF & Stress" section of the Configuration, and turn on the option for ALL_STRESS_CASES_CORRODED.

Richard Ay
COADE, Inc.

 

[KevinNZ]

I agree with mizzoueng that b31.1 does required you to consider corrosion. You can not ignor it.

When you thick about it how can you say a pipe will as strough when corroded as it was when new.

Kevin

 

[UtilityLouie]

I've worked with Caesar II, Triflex and AutoPIPE.

One of the things that I was mentored on from the start was to always consider the corroded thickness. So depending on what program I've used I make sure that all analysis is on the corroded thickness. This is the most conservative path and usually doesn't require too many changes.

 

[ColonelSanders83]

Whenever I run a code stress analysis in AutoPIPE I look at both the corroded and un-corroded states for the follow two cases.

1) Corroded: The corroded case will give you the worst case stresses in the piping within your system for the given corrosion allowance thus ensuring adequacy of the plant for the long term.

2) Un-corroded: If there is any equipment in the system with Nozzle load limits (and there always is) I will run the analysis in the un-corroded state, this will ensure the Nozzles can handle the greater loads imparted by the now less flexible pipe. Remember Pipe stiffness is related to wall thickness and if your corrosion allowance is a 1/16" on schedule standard pipe you may find that a pipe which was marginal before is now failing the allowable.

It generally only takes a few minutes to do and generally gives me good piece of mind, and remember the code is merely documented minimums, one must still use sound engineering judgment in the design.

 

[SOROURA]

Sometimes if you do not understand the Code, you might be mislead or interpret the Code incorrectly.

The B31.1 Code allows you to be conservative, both with respect to the pressure stresses caused by internal/external pressure, as well as for the sustained and thermal stresses. It's only how the Code does it is where the mystery lies.

The Code's conservatism for pressure stresses is covered by equation 3/3A. When determining the minimum wall thickness, the Code directs you to take allowances (of which corrosion is only one contributor) into consideration, because you have to account for it throughout the lifetime of the Plant, meaning that near the end of the plant's life span, the pipe wall thickness (corroded) should be enough to ensure that pressure stresses do not exceed the allowable stresses (SE or SF).

The Code's conservatism for sustained and displacement stresses is covered by para. 102.3.2. The Code does not explicitly state what thickness to use (even though it has "tn" in the equation for para. 102.3.2 (A.3)), but using the nominal thickness is actually more conservative when determining the stresses due to Dead Weight and also -as ColonelSanders83 stated- for thermal displacement stresses and loads on equipment. It's also more conservative for occasional loads, because more thickness implies more inertia, so this is also to be considered in the design of pipe supports.

I hope I clarified this point. Please let me know your opinions; and if I am mistaken, please correct me.

Thanks

 

[JohnBreen]

Hello,

The ASME B31 Codes for Pressure Piping are the MINIMUM requirements for piping design. These Codes allow you to do more rigorous analyses if you can and on the other hand they allow additional layers of conservatism. The piping engineer should THINK and decide WHEN the conservatism of the Codes is sufficient.

The ASME B31 Codes are a "simplified approach" and they include a level of conservatism that includes the consideration of the limitations of that approach. IF you opt to do a more rigorous evaluation, it may be appropriate to relax the degree of conservatism to some degree. It would always be prudent to document the approach, assumptions and any other considerations in the design brief. "Factors of Safety" are "indices of ignorance" - the more you know about all the design variable and the more rigorous your methodology, the less design margin is needed. When there are uncertainties more design margin is needed - "When in doubt, build it stout". If you can reduce the "doubt" you can build it "smart". Engineering is all about reducing the unknown to a level that is consistent with due diligence. I have heard it said that an engineer is somebody who can design/build something right with one dollar that any D***ed fool can design/build with two dollars.

Wall thickness calculations must include consideration of all the possible intended and unintended issues that may reduce the wall of the pipe under consideration (and long seams if appropriate). Yes, corrosion to be sure (but how many of us actually obtain accurate corrosion data for the combination of environmental conditions, the product to be carried by the pipe (at temperature), and other associated factor (CUI comes to mind) and apply that data). There is also erosion, abrasion, grooving (machining at Grayloc type fittings), threading, mill tolerances (although here we may be too conservative at the moment) etc. Then of course, after we calculate the MINIMUM required wall thickness with consideration for all these things, we round that number UP to the next commercial schedule (more conservatism). Reduction in wall thickness has a much greater effect on the safety margin involved in pressure calculations (circumferential ("hoop") stress) than it does on the margin of safety involved in the calculation of stress(beam bending stress (S = M / Z).

What is my point? There is a lot to think about and that is what engineering is.

Regards, John.

 

[DSB123]

John,
Very good points highlighted in your post. There are many people who beleive that the ASME Codes are "Design Guides" and not just the minimum requirements to mee tthe Code. As you say it is up to the Engineer/Designer to ensure the piping system is fit for purpose and additional considerations must be given where appropriate.
As you are connected with the Code committee (I beleive) on a slightly different subject - appologies to the original poster - have the Code considered if/when the Code SIF's may be changed/updated?

Regards

DSB123

 

[JohnBreen]

DSB123,

We are working on it. ASME has several active research projects funded and under way and Paulin Research Group is working on several aspects of this project. PRG has a testing machine that is patterned after the original Tube Turns machine used by Markl, Rodabaugh et. al. (only better) so we can build upon the original fatigue data base. Computational methodologies have also developed exponentially so 3D FEA models can be compared to testing results to help bring both technologies to common ground.

The research on which the original Code SIF's is 60 years old (wow) and we know a lot more now. The research will find a very orderly path into the ASME B31 Codes for Pressure Piping but it will not be in the next year or two. Maybe in my lifetime.

Regards,John

 

[DSB123]

John,
Thanks for the update. I have a license for the PRG FESIF package and comparing the results with the present Code values is fascinating. As you say maybe these approaches may get into the Code before I retire or expire hopefully the former. What would be the Code Commitee's view for Engineers to adopt the FESIF values in lieu of the Code values?

Regards

DSB123

 

[JohnBreen]

DSB123,

The Codes tell you that the SIF's included in Appendices "D" ARE TO BE USED IN THE ABSENCE OF BETTER DATA. If you have batter data you should use it but make sure you DOCUMENT the use of that data and provide a reference so that the client's verifier can check the validity of the data.

Now I am really feeling bad about high-jacking this thread - my bad.

Regards, John

 

[papiesz]

Hello again

I see that "my thread" got second life. I'm glad especially seeing John and other forum giants comments. It is very valuable for me.
I have almost 4 years experience in B31.3 calcs in Autopipe. Now I turned to B31.1 still autopipe but first stage of project was done by another company in Caesar. Some calcs which have to be repeated had minimum changes in routing but much greater in results. Comparing calcs in these two programs and their "default" code interpretations is very interesting but time consuming. I have some new questions but I think that would be better to start new threads

Regards

papiesz