Piping Stress Analysis Questions 2

[Ccorvera]

Hello all,

I am an early career engineer who has recently been doing a lot of work with pipe stress analysis using CAESAR II software. I have a few questions I was hoping someone could help me with:

1. In the CAESAR load combinations setup window, a coworker advised me that the analysis should use Ec (Piping Elastic Modulus at ambient temperature)for all load cases, including those that include thermal effects. Results of interest for us are generally pipe stress and support reactions, and his reasoning was that this approach would yield conservative results for these values. This reasoning seems to make sense to me, but I wanted to confirm with other knowledgeable engineers. Also, obviously if deflection was a result of interest, using Ec would be non-conservative, correct?

2. What is the difference in classification of load combinations as sustained (SUS) and operational (OPE)? These sound like identical classifications to me (any “operational” load seems to me like it would also be “sustained”). Any place in the code I can look at where these terms are defined?

3. I also sometimes work with pressure vessel stress analysis, and I am pretty confused by the stress classifications (membrane, bending, “linearly varying”, secondary, etc), as well as the process of stress linearization. Does anyone know of a good resource to teach myself these concepts?

Any help on any of these would be greatly appreciated.

 

[Snickster]

It has been a while since I did stress analysis using Caesar but here is my ideas

1. Not sure but I believe using cold modulus is correct. I don't recall ever having to calculate a hot modulus by interpolating tables of modulus versus temperature values.

2. For B 31.3, the Code cases to be analyzed are SUS, EXP and OCC. OPE is not a code case but are cases that can be used to build code cases or used to analyze cases to get worse case loads on pipe supports, equipment nozzles, etc., where the stress in the pipe may meet the Code case but in actual operation you want to know the maximum loads on supports and equipment that you have to really design for. I also made OPE cases out of any combination of loads I thought would occur in actual operation and used as design loads on supports and equipment, or determining maximum displacements, etc. Attached is a list of load cases to be analyzed. Note how the EXP case is built by subtracting the SUS case from the OPE case.

3. A good description of stress categories can be found in Pressure Vessel Design Handbook by Bednar Chapter 2 - here is link

https://www.academia.edu/44346522/Pressure_vessel_design_handbook_2nd_Edition_Henry_Bednar

 

[RVAmeche]

I don't have a comment on the EC approach - by default Caesar uses EC as you can see in the load case editor. I've taken the 3-day Caesar training and done other online training and never seen someone modify that field in the load case editor. Edit - A quick google found this in a Caesar webinar Q&A:

Sustained is, by definition, the as-installed condition with just ambient temperature and pressure. Generally Caesar assumes this to be 70F but if you know for sure the system will be installed in the dead of winter at 30F you should adjust it for that model. Otherwise, at least in my work, we don't really know when the piping will be installed so we stick with 70F. So in the Sustained case you may have all your supports in contact, but during Operating cases of say 200F you may see lift off on some supports, which will transfer weight/forces differently. This is why Caesar now includes the "alternative sustained" case for each operating case. ASME B31.3 Section 319 talks about displacement stress ranges between extreme displacement conditions (operating temps) and as-installed conditions (sustained case). I don't think the terms are explicitly defined in basic terms but I could be wrong. Similar to Snick's post see attached sample of load cases that also has a tiny bit of discussion on them.

The basic allowable stress equation in Section 302 has SL, the stresses due to sustained loads. Previously the "conservative" thing to do was assume SL = SH (basic allowable stress value at hot temperature) because it was too intensive to calculate the actual sustained stress throughout the system. Caesar by default now uses what they call "the liberal allowable" aka calculating SL for the system and using that in the equations so you have more allowable leftover for loads on the system. It's "less conservative" but it's more accurate than assuming the worst case value every time.

I can't comment on pressure vessel analysis be, with respect to Caesar, remember it is pipe stress analysis and not FEA. Everything in Caesar is a stick with a stiffness assigned to it and it does gigantic matrix math to solve spring equations. There are no meshes or surfaces in Caesar.

 

[TGS4]

CCorvera - I see that you’re in Washington State. You should be attending the ASME PVP Conference this week. There was a tutorial by Dr. Charles Becht on B31.3. Lots of opportunities for training, too.

 

[NovaStark]

1. In the CAESAR load combinations setup window, a coworker advised me that the analysis should use Ec (Piping Elastic Modulus at ambient temperature)for all load cases, including those that include thermal effects. Results of interest for us are generally pipe stress and support reactions, and his reasoning was that this approach would yield conservative results for these values. This reasoning seems to make sense to me, but I wanted to confirm with other knowledgeable engineers. Also, obviously if deflection was a result of interest, using Ec would be non-conservative, correct?

There are some codes that specify that the hot modulus E[sub]h[/sub] can be used for equipment nozzle loads given that Ec may be too conservative. So it depends on what you are doing and how sensitive your results at a particular location.

3. I also sometimes work with pressure vessel stress analysis, and I am pretty confused by the stress classifications (membrane, bending, “linearly varying”, secondary, etc), as well as the process of stress linearization. Does anyone know of a good resource to teach myself these concepts?

This might be also helpful (from a Becht blog). I can't recall if this is specifically in ASME VIII Division 2 where these would be more prevalent.

 

[Ccorvera]

Not sure how to reply correctly on this forum but thank you everyone for your answers and provided resources, this was very helpful. Everyone seems in agreement that Ec is the conservative value to use for pipe stress and reaction forces. However would you also all agree that this value would be non-conservative if the value of interest is pipe deflection? And if not, why?

Also @TGS4: I was in attendance at the PVP conference and was able to sit in on one of the B31.3 tutorials by Dr. Becht. I was only able to attend two days however.

 

[StressGuy]

Ec isn't the "conservative" value for calculating pipe stresses per the piping codes, it's the requirement. For reference, B31.3 2022 section 319.4.4 calls this explicitly for flexibility stresses (EXP in Caesar).

For loads/reactions, you can use Eh at the operating temperature, though I don't typically take this approach unless the temperatures get to be above 800°F or so. That's where Eh drops enough compared to Ec to start making a meaningful difference. Caesar defaults to using Ec. You can choose which E value you want in the load case editor now. You could have a W+T1+P load case run with Ec for the code stress calculation and another case run with EH1 corresponding to T1 to give you restraint loads calculated based on the lower stiffness that EH1 should generate.


The fundamental differences between SUS and EXP loads/stresses are:
SUS - the stresses don't change due to system response. Weight keeps pulling down due to gravity regardless of how the system deflects. Pressure keeps pushing outward on the pipe wall regardless of deflection of the pipe shell. Exceeding the yield stress under a sustained load will lead to continued yielding until failure. Stresses are limited to Sh, which is the basic allowable stress at the temperature of interest. Sh is generally going to be the yield stress with a safety factor reduction that will vary a bit from code to code.

While we like to think of the sustained stress as due to the pipe in the cold/ambient condition, that's not really how it works. It is due to the sustained loads like pressure and weight, but must be considered with the supports as they are in the operating state. This means that if your piping system heats up and lifts off of a support such that it is no longer in contact with the pipe, it doesn't count as a support for the sustained stress calculation. This was a topic of dispute for many years, but the code committee finally clarified this and has it addressed in one of the included example problems. That's what the "alternate sustained" cases in Caesar are setup to address.

EXP - The fundamental feature of "expansion" stresses is that they are self limiting. When a piece of pipe heats up, it will expand. That expansion deflects the piping system and creates a stress state. However, once the system reaches temperature and finishes expanding, it will not continue to deflect the piping system further. It's very difficult to fail a piping system under one a thermal cycle. Thus the allowable stress value for the EXP stress check is 1.25Sc + 0.25Sh and has a basic expectation of being good for 7000 full temperature cycles. There's some more nuance, but that's the intro version.

As noted above, at least for the B31.3 and B31.1 codes, there isn't an "Operating" stress case to check.


Edward L. Klein
Pipe Stress Engineer
Houston, Texas

"All the world is a Spring"

All opinions expressed here are my own and not my company's.