Stress analysis of patchwork quilt pipe, strength/flexibility governs?

[e123344]

A client wants us to stress check a system comprising new pipe designed by my co. and tie-ing in to exisitng eqpt and pipe system already installed and operating.
Alas the clients docs are ambiguous in the definition of the pipe in-situ, varying from HFS32 to ASME 106 grade A or B. The operating temp of the system is 60 deg C, design is 280 deg C. Op pressure is circa 47 Barg.

The client is suggesting that we run the stress check using the lowest yield of the possbile materials, but i am unconvinced this is good or even safe practice.

The piping in question is inaccesible, and so PMI is out of the question until the unit comes off-line during the shut-down at whcih point the designed new piping will be installed.

To cover ourselves i believe we should consider the most onerous permutations, but as to what those are i am unsure as i do not know which will govern, i.e. is it always the case that one or other of material strength (or more appropriately weakness) or the system stiffness is likely to lead to a failure of the stress check. Essentially i'd like some guidance as to which combinations of highest / lowest stress / yield / UTS / wall thickness properties of the three materials will give the worst case scenario. It is also important to note that there is a seismic condition to be satisifed - assume it is SSE.

Furthermore, can anyone advise if there would be any stress reduction factors applicable for the use of mis-matched CS materials in this fashion ? i.e. would the sigma for the material with the lowest allowable actually be the lowest
allowable in the design or would it be reduced further due to factors already described; welding of dis-similar materials with different allowables, taper boring that will be required.

Many thanks.

MDW

 

[JohnBreen]

If you want to do a legitimate piping stress analysis you first have to determine what Code or Standard is to be used (this will be the Code/Standards that the local jurisdiction mandates). Then you will look at the mandated Code/Standard to answer most of your questions. Or, alternately, identify the Code/Standard and ask a series of focused questions in this forum.

Regards, John.

 

[e123344]

John Breen,

thanks for response.

I made some asusmptions re. my post and the recognition of my engineering background (dubious as it evidently is). I am reasonably well versed in codes and standards, but without knowing the minutae of 31.3 for example.

The design code will be BS, and whilst the code will specify how to do the analysis if the entire system properties are known, i am fairly sure it won't tell me how to determine the most onerous combination of material strengths v other mechanical props. I am not seeking the load case combinations to consider, i am looking for the combination of mech

I take your point re. my posting being perhaps too complex, so breaking it down :

question 1) are there any rules of thumb for determining whether the lowest allowable stress or the system flexibility will govern in a failure case.

Q2) Should i only be considering worst combinations of allowable stress and wall thickness in a system of unknown pipe properties (but known to be one of three specs and grades) or should i also consider UTS ?

Obviously we could plug all the possible combinations into the analysis software and see what turns out, but that takes time and essentially money, which if you'd seen the manhour estimates my engineering dept produce is not something to take lightly !!

 

[JohnBreen]

Answering from the B31 point of view:

Q1) are there any rules of thumb for determining whether the lowest allowable stress or the system flexibility will govern in a failure case.

There are NO "RULES OF THUMB", the rules are all explicitly spelled out in the Codes. Flexibility will affect the calculated stresses.

The ASME B31 Codes address PRIMARY (non-self limiting) longitudinal stresses due to sustained weight (dead weight and live weight) and pressure (SL) differently than SECONDARY stresses (self limiting) due to thermal expansion/contraction (and other cyclic loadings such as wave motion - SE).

The equations for calculating the stresses due to sustained weight and pressure (SL) differs from one B31 Code to another B31 Code (also see ASME B31.3 Code Case 178). The B31 Codes set the maximum ALLOWABLE primary longitudinal stress (Sh) due to weight (dead weight and live weight) and pressure by providing allowable stress tables in Appendices "A" of that Code. The allowable stress is the listed allowable stress "at operating temperature".

The B31 Codes provide equations for calculating "displacement stress ranges" (these are also know as the "flexibility stresses"). The B31 Codes provide an equation for calculating the maximum ALLOWABLE "displacement" stress RANGE for comparison to the calculated "displacement" stress RANGE. In the case of "displacement stress range" you would calculate the "full stress range" from the lowest temperature to the highest temperature (if it is a system that ordinarily operates at temperatures above the installed ambient temperature the temperature range would be from the coldest winter night when the system is shut down to the warmest temperature caused by operation or ambient air conditions). You would add the absolute value of the displacement stresses due to the temperature excursion from the installed temperature to the coldest temperature to the absolute value of the displacement stresses due to the temperature excursion from the installed temperature to the hottest temperature to get the calculated stress range.

The allowable stress range is calculated as:

Sa = f*(1.25 Sc + 0.25 Sh) or as

Sa = f*[1.25*(Sc + Sh)-SL]

where
f = a stress range reduction factor which addresses cycling.

Sc = the allowable stress from Appendix "A" at ambient temperature

Sh = the allowable stress from Appendix "A" at the highest operating metal temperature

and SL is the longitudinal stress due to sustained weight and pressure.

For occasional loadings that are in excess of the normal operating loadings (and limited as to the number and extent of the loadings) the B31 Codes allow greater allowable stresses. You will have to read the Code regarding occasional loadings and combinations involving occasional loadings.

So, the calculated "sustained" stress must be less than the Code "allowable sustained stress" AND the calculated displacement stress RANGE ("flexibility stress") must be less than the calculated allowable displacement stress RANGE. Note that the allowable stresses in the B31 Codes are a fraction of the yield strength of the material or a fraction of the ultimate tensile stress of the material and these allowable stresses are provides (at temperature) in appendix "A". In some cases the creep properties of the materials affect these allowable stresses and this is also considered in the allowable stresses provided in Appendix "A".

Q2) Should I only be considering worst combinations of allowable stress and wall thickness in a system of unknown pipe properties (but known to be one of three specs and grades) or should I also consider UTS?

You should be addressing calculated stresses (calculated stress ranges) and comparing them to the Code allowable stresses (stress ranges). The Code allowable stresses do consider the ultimate tensile strength (and other properties) of the materials. Follow the Code and the Code's allowable stresses (and stress ranges).

In the case where you know all of the materials used in the system but do not know where in the system these materials are used it would certainly be conservative to use the Code ALLOWABLE stresses for the material known to have the least allowable stress (or calculated stress ranges calculated using the Appendix A data). The better approach would be to do material identification testing. If you do not know the material having the least allowable stress you cannot do a legitimate structural (stress) analysis.

You must use the proper section properties of all the pipe (the wall thickness affects these) to perform a legitimate structural (stress) analysis.

Regarding the loading cases that are to be analyzed, look to the mandated Code to provide these combinations. These will be credible combinations - i.e., loadings that actually can happen. You will note that some Codes will not require some "occasional" loadings to be combined e.g., seismic loadings to be combined with excessive wind conditions.

You MUST read the Code and understand the Code.

Regards, John

 

[e123344]

John,

thanks for the exceptionally detailed response. I have 12 years experience in oil and gas industry so adherring to codes is very much in the blood. I fully understand that codes must be read and understood, and they dictate the allowable stresses to be used and the load case combinations. My problem as you picked up in the bottom is that we don't know the exact material installed, but we know it's one of three. Identifying excatly what we have and where is out of the question, we are customer driven, so we need to make the engineering judgement as to how to provide a design that is safe, and inherently in such situations that means onerous and/or conservative.

I am not a pipe stress specialist, but i have a reasonable grasp of the basics of mechanics and code requirements. My concern is if we assume that the whole system is comprised of the material with the lowest stress, will the additional flexibility that inherently infers give a potentially incorrect result, as in piping systems i am aware that rigidity of a pipe can be as fatal a flaw as it's lack of material strength. I don't have the experience to make the call whether the allowable stress of the material, or the increased rigidity of the system is more likely to produce a failure.

I totally agree with you re. using the correct configuration of the piping system, I am checking that my engineering team and my client know what they are doing, for as i infer if not state, in my eyes, just assuming the lowest allowable for the system is to put it politely "not best practice", possiby even negligent.

thanks again.

 

[BigInch]

If they are doing it, they probably don't know what they are doing.

While it is conservative to assume the lowest allowable stresses of any material used in the system as the pass fail criteria, that measure alone will not guarantee you obtain the correct stresses from the analysis.

You must know the expansion properties of the materials, their Young's modulii and their geometric properties of the components in their correct topology within the system to obtain correct stresses and SIFs. To analyze a piping system with any degree of confidence and where so many permutations are possible would be a nightmare. I wouldn't think it could be done, except perhaps by employing some kind of statistical method, which are not addressed by any code I now about. As any analysis using unknown properties would essentially produce a statistical result, code compliance would be impossible. As such, to imply to a client that you could bring the plant piping into code compliance, or even give some measure of reassurance that the piping would be operating below allowable stresses at maximum design temperature, given such unknowns, would be highly questionable.

You don't mention why there is apparently such a wide divergence in operating and design temperatures. Has this piping been derated, or is it not operating at higher temperatures now, because of its unknown components?

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[BigInch]

What I would consider doing is a study of the system solely intended to identify potential problems in relation to the minimum yield stress of all known materials, BUT in no way imply and clearly state that you will not be using any particular code as a basis and that any results contained therein could not be applied to any code compliance issues, past, present, or future.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[e123344]

Big Inch,

thanks for the response. You have confirmed my fears, re any assumptions being unsafe.

the design temps are so varied because of an upset condition, i.e. potential CV failure leading to hot gas going down the wrong pipe, or safety venting.

p.s. please don't respond to the air v pressure testing post again. my email is full of notifications already.

 

[BigInch]

Sorry. OK. I certainly don't plan to.
You can turn off e-mail notification. Bottom left.

So that temperature load would be very rapid as well. Not the best way to warm up the system.


**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[cloa]

Couldn't you assume each pipe has minimum strength and maximum rigidity- artificial as that is, at least it would give your worst case scenario.

 

[BigInch]

That could define the range of possible reactions, but as far as the system internals are concerned, I think that would be equivalent to saying the system response could range somewhere from a truss to a beam. If a truss, no bending resistance would max the axial forces; if a beam, maximum bending and no axial forces. Then in the real system, you could never be sure where a part of the system acts more as a truss, or if it acts more as a beam, which effectively shifts whatever forces it can't take to other parts of the system. You would never have any real idea of what is occuring at any localized component, which is afterall what you really need to know. Even if you did get some recognizable result, what would you do if it was unreasonably high or low? What part of the system would you change and what justification could you have for changing it?

I think it might be a better bet to start with the average strength and regidity, see if problems show up and move towards component attributes that might push those to a worsened condition, however in the end even that method sounds like tons of work, with intermediate results hard to interprete, difficult to manage and not very indicative of true stresses at any given component. But a run using average properties just might indicate where the system is especially susceptible to hi stress and point out areas that need to be investigated in a detailed manner at the next shutdown. It might let you concentrate available resources to those areas and identify which segments of the system could be treated in a lighter sense.... maybe not.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/