ASCE Buried Pipeline - Combining Stresses ? 2

[AlbertaMecchie]

Hi,

I am using ASCE buried pipeline guidelines to evaluate stresses in a buried slurry pipeline.

I have determined the through wall bending stress from earth and live loads, hoop stress from internal pressure, and then compressive thermal stress due to thermal expansion and hoop stress. Wall crush check and ring buckling check is OK.

ASCE gives criteria the following criteria under Appendix A:

- Hoop stress to be within allowable by code - My code is B31.4 and stress is within allowable
- Through wall bending stress < ).5* Yield stress - Stress passes
- Thermal expansion for secondary loading - Within B31.4 code allowable

My question is, do these stresses somehow have to be combined to determine some kind of equivalent stress, or is it sufficient if they are within allowable independently ? The ASCE guideline doesn't seem too clear on this.

 

[LittleInch]

Why not use one design code like B31.4 which has an equivalent stress check?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

They must each be checked individually and also checked for combined stress when using the ASME codes. ASME B31.4 is apparently your specified design code. Use the provisions found within and in its referenced publications to define all minimum criteria of your design.

You must not combine codes. Use only your specified design code to define the minimum design criteria and requirements. You are free to use any method or quideline, company spec, etc.to confirm design adequacy, or to provide additional design assurance, but whatever method you use in addition to your specified design code must always result in a design that at least meets your specified design code's minimum requirements.



Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[AlbertaMecchie]

Thanks for the replies !

The pipeline I am dealing with is a restrained slurry pipeline, as it will be buried.

The issue is that the ASCE Guidelines for Buried steel pipe (americanlifelinesalliance.com/pdf/Update061305.pdf) have no method described for combining the stresses into a combined or equivalent stress. It mentions "The designer should appropriately combine the effects of concurrent loads when evaluating the adeqacy of the buried pipe". The Guidelines are just that - guidelines and not a code. Under the Appendix A, it says "The criteria for a buried pipeline should be established on a case by case basis".

It then indicates in Appendix A Suggested Acceptance Criteria for loading condition for hoop stress, allowable stress should be within "code allowable" for internal pressure - in this case, I was using B31.4 for the code check on that, because the ASCE guideline is not a code. It then describes in the Appendix allowable stress limit of 0.5xYield stress for the through wall bending stress, which is an equation given in the ASCE guideline.

The ASCE guideline gives an equation for maximum compressive thermal stress, which is compressive longitudinal stress due to temperature differential and is: Young's modulus x coefficient of thermal expansion x (operating temperature - installation temperature) - Poisson's ratio x hoop stress. The ASCE guideline Appendix A indicates thermal expansion stress should be withing "code allowable" for secondary loading.

The ASCE through wall bending stress due to soil cover over the pipe and live load (in my case ground pressure from a Cat D8) is not the same as beam bending stress (M/Z), which is based on the pipe section modulus and unsupported span. Referring back to ASME B31.4, the equation for equivalent stress for a restrained pipeline is based on Sh (hoop stress) and Sl (longitudinal stress = thermal expansion stress + Poisson ratio*hoop stress + M/z) and Ss (torsional stress). There is no tortional stress in this case.

I'm pretty sure I can't use the B31.4 equivalent stress calculation for a restrained pipeline in this case, because through wall bending stress in the pipe is not "M/z" or beam bending stress from an unsupported pipe span. Through wall bending stress acts perpendicular to, and not along, the pipe longitudinal axis.

So I guess my question is how do I combine the hoop stress, longitudinal thermal stress, and through wall bending stress ? Or do I need to combine the stresses at all and can I more simply use the ASCE Guideline Appendix A criteria which gives acceptance criteria for the individual hoop stress, through wall bending stress, and thermal expansion stress ? To me, combining these 3 particular stresses seems a little tricky, because the through wall bending stress has both a tensile and compressive element, and how do you combine that with hoop stress acting, more or less, in the same direction ?

 

[1503-44]

The same hoop bending stresses would occur in the API 1102 method, although they are not considered to be directly added algebracially to hoop stress in 1102, you could add them to the hoop stress if you feel that such a treatment is necessary. The hoop bending would increase hoop tension stress at the OD and decrease it by the same amount at the ID. You could then check the total hoop stress the OD against max allowable hoop stress, noting that the hoop tension stress would also be slightly reduced from the effect of addeding the compressive load in the wall from the D8 pressure. Considering the average stress in the wall, with bending stress cancelling out at the NA and the additional compressive load from D8 pressure reduces the net hoop tension anyway, which is probably why it is ignored by 1102.

If you don't want to ignore through wall bending stress, just add it to the hoop pressure stress at the OD and check that total against the allowable hoop stress.

Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Alex Matveeff]

EN 13941 and ISO 14692 codes has equivalent stress that include ring bending stress from the soil. But ASME B31.4 and B31.8 not.
Also you write "I am dealing with is a restrained slurry pipeline, as it will be buried". If the pipeline is buried, it doesn't mean that it is restrained.
It could be buried, but unrestrained or semi-restrained! Need to analyze each pipe segment behavior.

I'm the PASS/START-PROF Pipe Stress Analysis Software Developer

 

[AlbertaMecchie]

Thanks again for the assistance.

Given that the pipeline is not likely fully restrained, what I have done for now is to estimate an X axis stress from compressive thermal stress and a Y axis stress from hoop stress/thru wall bending, and then calculate estimated principal stresses and Von Mises stress. It seems to me a reasonable approach, given that neither the ASCE buried pipeline guidelines nor B31.4 code give any guidance or equations for determining a combined, or equivalent, stress accounting for thru-wall bending stress from earth / live loads.

I suppose I could take a look at API 1102 (provided my company will buy a copy), although I'm not sure how relevant this would be. The live load is not a "highway" wheel load, it is from a tracked vehicle (Cat D8).

 

[1503-44]

Use Tresca stress if B31.4 or 8 will define minimum code wall thickness. If VM results in a thicker wall, you are free to use it. If less, then no, under 4 or 8 you cannot reduce wall thickness to VM thickness.

Generally the longitudinal restraint condition of buried pipelines are not entirely well known and after time many initially unrestrained points become fully restrained, thus both conditions should be addressed, selecting the thicker wall for use wherever positive means are not installed that will guarantee flexibility underground will be permanent.

Using 1102, you would have to make a reasonable conversion form tracked load to wheel point load. That can be easily done, as the track should be roughly perpendicular to the pipeline. There's no way I would design a crossing point only for tracked vehicles, so if you do not use a typical truck load, don't forget the load limit warning - tracked vehicles only. Generally that is never done and a 3ft clear burial is considered adequate for the odd typical farm tractor, a car or light truck crossing above with no additional special protective measures. Tracked loads are less.


Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[AlbertaMecchie]

I notice in B31.4 under 402.7 Combining of Stresses it says you may combine stresses using maximum distortion energy theory. Which I believe to be Von Mises approach. So I used hoop + thru wall bending as stress in X direction, and thermal compressive stress as stress in Y direction, calculated principal stresses, and then Von Mises stress from principal stresses. Seems reasonable to me.

My crossing point is in an oil sands tailings facility, it will be crossed by D8s and maybe the odd 3/4 ton pickup truck, nothing more than that, so I think this will be good.

 

[1503-44]

There is an ASME clarification dated 1992 prohibiting use of Von Mises Distortion energy (the basis of my previous comments) here. I do not have a copy of the latest edition of B31.4, so it would be helpful if you could post a copy of paragraph 402.7 I and maybe others would like to see it, as it may be a significant change to previous editions, which also may affect their previously issued clarification that apparently has not been withdrawn. Maybe we should ask for a clarification to the clarification.

The clarification to which I refer as well as other historic clarifications published by the ASME code committees can be found here,

https://cstools.asme.org/Interpretation/SearchInterpretation.cfm

Von Mises Distortion Energy.

https://cstools.asme.org/Interpretation/InterpretationDetail.cfm?TrackingNumber=13339

Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[LittleInch]

From 2019 version which is the latest current version.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Thanks LI.

This is interesting too.

https://www.eng-tips.com/viewthread.cfm?qid=275926

So, "alternatively" you can use either answer.
I feel a clarification coming.

Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[AlbertaMecchie]

I ran some calculations per API 1102 using the Cat D8 LGP track ground pressure and for what appears to be the highest stress scenario of maximum operating temp, lowest installation temp (it gets cold up here !) and maximum internal pressure, API yields effective stress somewhat lower than the B31.4 equivalent stress calculated. I had added hoop stress to thru-wall bending stress in my B31.4 calculation so as you mentioned perhaps 1102 treats circumferential stress differently.

 

[LittleInch]

You haven't actually said what size, wt and pressure your pipe is at and I can understand why you want to prove that it will be not stressed, but I think you could lay this pipe virtually on the surface and so long as the tracks didn't chew it up it would be fine.

Out in the desert with surface laid pipes they just drive over them....

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.