Has there been a change in the way Caesar handles the lifted up supports in determination of the sustained stresses?

If yes, when this was done?

Thanks.

No, why do you ask?

CAESAR II does not automatically "handle" support lift-off the user is expected to provide supervision to the software solution...

Hello all!!

I have been reading the Technical Reference Manual 6.19 about suStained stresses and it is really interesting. It surprises to me that removing supports THAT lift-off during operation (as I usually do) when calculating the sustained stresses is not neither clear nor evident and some people think is necessary and others not (as CAESAR responsibles).
CAESAR technical reference 6.20 says that "if the pipe became overstressed, it would yield and sag back to the support". Do you think it would be so in the case of lifting, i.e, 140 mm??
I have read somewhere that a lift of up to 1/4" may be considered as negligible but when it is bigger I am not sure if it could suppose a problem...
I would like to know the opinion of other people, about negligible lift-off, if they remove the supports with any lift or only if they exceed certain limit,etc..

By the way, why B31.1 doesn't state it clearly and there would not be any doubt for anybody??!!

Best regards

Removing the support is "wrong". Here is why:

Example 1: Consider a header with three identical branches to pumps. Two branches are operational, one is standby. In an operating branch, a +Y lifts off. Removing this restraint for the analysis is incorrect for the standby line.

Example 2: Consider a system with more than one operating temperature, where some supports lift off in some, but not all, of the operating load cases. Removing these restraints is incorrect for those cases where the pipe would sit on the restraint.

OK, but to make the problem easier let's imagine a system with one operation temperature and the whole system at the same temperature. If one +Y support lifts 100 mm during operation, is it "correct" the sustained stess calculated with CAESAR? (let's forget the influence of that support for expansion range for example). Is it reasonable to consider that "if the pipe became overstressed, it would yield and sag back to the support"?. I see that the yield and sag back is clear for small lifts but in case of doubt what I usually do is to make a copy of my model and run a cold sustained case with the supports that lift in operation including a Y gap of the same length than the lift during operation in order to check the sustained stresses.
Which would be the stress in the pipe to sag back 100 mm?

Regards and thank you for any reply

Hola...

See http://www.coade.com/newsletters/jan01.pdf

and then post any further questions you may have!

"Wrong" is such a severe word. I tend to think of it as "conservative". Seems to me there are two schools of thought on this subject: is lift-off a primary stress issue (SUS) or is it secondary (EXP)?

For the pump example, stresses typically aren't the driving force in the design. Allowable pump nozzle loads generally dictate the use of spring supports or some other means to take the load off the nozzle. The multiple temperature example is different.

My approach: look at the SUS load that is being eliminated by lift off. If the "dead weight" load is small (compared to adjacent supports) I feel comfortable removing this from the analysis since the re-distributed load will be small. If the system is not overstressed without the support, this is conservative. However, if it fails, I place a spring support at this location, or find some means to re-distribute the loads so the system will not be overstressed.

A number of years ago, a client had an audit done on some compressor piping. The auditor used a load case combination that seemed very unconventional. The majority of the time I use the load case combination

1 W+P+T (OPE)
2 W+P (SUS)
3 L1-L2 (EXP)

or variations thereof. However, I believe the auditor used

1 W+P+T (OPE)
2 T (EXP)
3 L1-L2 (SUS)

or something like it.

Basically it boiled down to the issue at hand, dealing with lift-off. He was of the opinion that his load cases correctly dealt with lift-off. We agreed to disagree. Regardless, his comments were relatively minor and not related to overstressed points due to lift-off. We incorporated his comments to the satisfaction of the client.

Mr Ay,

Respectfully, I too disagree with COADE's position on +Y lift off.

The examples that you give above are very real situations that occur, but that's not a justification to treat the effects of +Y lift offs as secondary loads and secondary stresses. For me, thermal loads due to pipe expansion or contraction are secondary and dead weight loads are primary. When a pipe lifts off a support, the dead weight loads in the system are redistributed to other supports. The stresses resulting from the redistributed dead weight load are primary.

For me, the answer to the examples you cite is not to allow +Y lift off to be treated as a thermal condition, but rather to create additional files to define the separate restraint conditions for each situation as required. It's not the easy way but I believe it's the right way.

Best regards,

Kevin,

You said that the redistrubted dead weight loads are primary. I disagree. Any difference between the two load cases is secondary. This fact is stated in the French piping code, CODETI.

If you disagree, read the section on "hot sustained stresses" in Chapter 6 of the Technical Reference Manual. This sets up a SUS load case the way Richard (Havard) lays out above.

In actual fact, as is pointed out in the newsletter article referenced above, you may need to evaluate the SUS stresses in more than one (boundary) condition of the piping system. Simply removing the +Y support may address (one) hot sustained condition, but the resulting expansion stresses are incorrect. This is why I used the word "wrong" above. You can't just remove the +Y from the model and call it quits.

And another thing....

Check for Sag.... if the line is very hot do your sag check with active supports and a hot mod E....

Richard (Ay),

After reading the well-written article referenced above, and realizing you were one of John's editors, I was a bit surprised by the apparent difference of opinion on this matter. However, after consulting with John, I realized the "correct" analysis is to leave the support in to capture thermal stresses AND create a seperate file without the support to check the sustained stress. If my understanding is correct, maybe the chasm isn't so wide or deep.

My next question is, if the support is removed for the thermal analysis, would this yield overly conservative results? Or would they be severly underestimated?

I don't think you need the second file (model). You can do all of this in one run. Refer to the "hot sustained" discussion in the Technical Reference Manual.

Neither/Both - I think it depends on the specific model being analyzed.

Thank goodness I was afraid there was going to be a necktie party down in Houston!....

Rich and I are in agreement....
The most important things are...
1)Review the model results carefully.
2)Capture and evaluate the supports that lift-off in some manner as opposed to doing nothing!

And finally if it gets too messy and its possible to do so eliminate the lift-off supports by....
a)deletion
b)strapping the line down with a full y uplft restraint or
c)use a can

BUT DON'T SIMPLY IGNORE THE SITUATION!

Nobody is wearing a necktie in Houston today - it's Friday ...

After you have carried out the stress analysis, one thing to be cautious of if you have a support lift off (esp larger lift offs) in your piping is the desire on site of someone to wedge something into the bottom gap thinking there is a problem with the pipe support.......

Hey Richard (Ay ... can I call you Ray?),

We don't wear neckties....ever!

Miss Itchy, good point. However, the ones that would put a shim in to reduce the bottom gap, are probably the same ones that forget to pull the travel stops on springs. And I've seen too many springs with travel stops not removed to worry about shims.

Hi Richard H,

We deal with pretty hot piping, so when we see lift offs on existing piping on site they can be pretty large and it is more an issue of people sticking 2" x 4" bits of wood in the support gaps rather than shims.......Mind you, if the lift-off is supposed to be there or not is another issue.

After reading carefully John C. Luf article about sustianed stresses in Coade Jan. 01 I have a question. It is said (and I agree) that a seperate new file without the support to check the sustained stress must be run. I suppose that if the lift off during operation is 0.1 inch, 0.5 inch or 1 inch mustn't be the same but in that new file, removing the support completely, there is no difference between them . Moreover, what would happen if the support is lifting in operation i.e 0.2 inch and the sag in the new model is bigger than that in that point? I think that the calculated sustained stresses wouldn't be correct. Instead of completely removing the lifting supports I think they should be considered in the new model but with a Y gap of the same length than the lift-off during operation. What do you think?

Regards

Carletes,

If the sag was greater than the liftoff then the support would not have lifted off in the OPE case.

Ineveitably the few times when I have done this hot sustained model without the support in place I have found that sag is not equal or greater than the lift-off. Also I have seen only a small percentage of times where the calcuated Sl with the inactive support removed created an overstressed condition....

I remember several years ago, while describing the nonlinear capabilities in CAESAR II, an engineer said to me that lifting off a support is poor design. To him it was a yes/no, good/bad or black/white issue. I am sure that this is (was?)the opinion of those earlier pipe stress engineers who could handle a key punch machine. They just didn't allow it to happen.

So now we have software that can handle the structural response of piping systems with nonlinear boundary conditions. And yet we do not have clear guidance from the piping codes. The latest B31.3 interpretation of which I am aware states that sustained stresses should be examined in ALL support configurations.

Now I've been working with analysis for many years and, with only an exception or two, I am quite confortable with how CAESAR II does it. It is my opinion that "the Code" calls it sustained stress because it is always there and can lead to failure by collapse. Also, all changes in stress are secondary and these lead to fatigue failure.

I think this subject has a lively set of comments because, since the piping codes have not yet "told us what to do", we are all doing what we believe should be done. But remember, the piping code shouldn't be seen as a guide to good design.

What we should do is all get together at some bar after work and throw out a few conversation starters like lift off, sag, shakedown, hot tap, rachet, hot modulus, or friction. Until then, I think this thread would be of greater benefit if, instead of saying CAESAR II is right or wrong, we offered and supported our own opinions on any of these subjects.

Here is one opinion - I have not found a situation where (Operating)-(Thermal), as offered by Richard Havard's auditor, did not produce the same sustained stresses as the linear "hot support configuration" (with "unused" restraints removed).

Here's another - any yielding that occurs while the system moves to its hot position is shakedown.

And here's my question to the field - if a 1mm liftoff is OK but a 50mm liftoff is not good, what measure of "goodness" are you using? Is it Itchy's "make sure no one sticks something in there" or "make sure no one removes the (maintenance) support" or "boy this stress is getting up there" or what?

I hope I'm not hijacking this thread...

Dave

Its even murkier than that....

If a support bears down on a +y lets say 650n with the Sustained case of W+P but during one of its OPE cases it bears 640n what happen to the delta?

You have to ask the question what will the long term effects be of the load re-distribution....

Unfortunately if you read the original query of this thread.... "Has there been a change in the way Caesar handles the lifted up supports in determination of the sustained stresses?"

It seems that a simple yes/no automated approach to this difficult topic is desired.

If it is feasible to do so eliminating lift-off / load redistribution is the most straight forward way of taking care of this complex issue.

Oh and by the by I'm an ex-punch card dynaflex/mare island guy!

Life is complex at times.... thinking is always a good thing!

Mr. Luf's statement :

"Unfortunately if you read the original query of this thread.... "Has there been a change in the way Caesar handles the lifted up supports in determination of the sustained stresses?"

It seems that a simple yes/no automated approach to this difficult topic is desired."


It was clarified by Richard Ay. on the first reply That Coade's position on this issue has not changed.

Wow... we are sensitive are we not! Your original question led me to my own conclusion as to intent.

As to what your intent was or was not only you can say. But if any person thinks this is a simple pass fail situation hopefully they will come away better informed.

As to what your intent was well if I have mis-interpreted your intent I am sorry but I think whats more important is the threads contents as a whole.

My humble opinion (on pipe lift off)

Aside from the technical issues,
The issue of pipe supports lifting off in the field is mostly a visual issue and usually results in some very important but technically incompetant client ranting and raving in your bosses office obout pipes left dangling off supports.
I find that in the majority of cases, the success or failure of the stress analysis input on the piping design is measured on a non technical and visual basis by 'tea break'engineers.
So, i generally try to make sure that the pipe sits on all supports, all the time.
Personally, i agree with the old engineering maxim of not letting support lift off. it just look bad.

Regards.

Dave, I like your idea about a pipe stress round table discussion at a local watering hole after work. Maybe we can maximize participation by scheduling these when one or more of the out-of-town folks are in Houston. Keep us advised.

While I disagree with John's and COADE's position on lift-off, I 100% agree with John's comments above posted May 06, 2005 12:26 PM. No matter what your positon, this advice will lead to a successful analysis.

Best Reagrds,

Hmm... Mr. Monroe you have voted "negative" without comment which is a no no under the bylaws of the ASME code committee.

I respectfully request you expound on your point(s) of disagreement so that the discussion may continue.... The committee is a consensus based organization and as such encourages open discussion. I feel this is a good thing and would like that to occur here as well...

Unfortunately I'm not down in Houston but if you guys do meet I'll buy the first round in absentia!

Sorry John, I thought I stated my position earlier in this thread.

One of the qualities of secondary stresses is that a small amount of localized yielding will reduce the stress. When a pipe lifts up off of a support, the dead weight load that was on this support is re-distributed elsewhere in the system. Unlike thermal forces that reduce with some yeilding, forces due to the weight of the piping system in the hot position will remain constant after any yielding so these stresses, which in my opinion are primary stresses, will remain constant as well.

This is why I feel that removing the +Y support or clamping it down is the more correct way to evaluate sustained stresses. Still, I know this is not perfect. It doesn't account for those situations where a support reaction is 3 kips in the sustained case but only 100 lbs in the operating case. Though the pipe still touches the support, for all practical purposes this support is gone and the dead weight load has been re-distributed elsewhere. For these reasons, I not only evaluate the +Y supports that lift off but other supports that have significant differences betweeen the sustained and operating loads. A complicated system may require more than one CAESAR run with different support definitions for a satisfactory evaluation.

I'm not completely convinced that the load cases advised by Richard Havard's auditor correctly evaluates this situation either. The "T" alone is not an accurate behavior. This load case set-up may be OK if all of the Y supports were coded as double acting and only the ones that want to lift are coded as +Y.

About SUPERPIPER's comments, I don't know how you make all of your supports sit all the time, even generally. If you have hot piping with vertical offsets, you're going to get lift-off quite often. Many times that support is not required, so there is no reason to add flexibilty just so it doesn't lift off of the steel.

Also, your choice of the words 'dangling off supports' lets me think that your experience with lift-off has resulted in unstable piping installations. For flexible systems, I like to add a guide at the supports that lifts off for stability.

That's my 2¢,

Let me offer some additional remarks to my previous post.

I wanted to clarify that although I keep talking about the +Y support that lifts off, this support is not the primary focus of my concerns except for over-span situations.

My main concern is to evaluate the system for sustained stresses as a result of the redistributed loads. The weight reaction that was supported on the support that lifted up is now reacting on other parts of the piping system. I remove the +Y that lifts to evaluate the piping system with the increased weight reactions on other parts of the system as a sustained case.

Mr. Monroe,

Thanks for your expounded remarks.....

I tend to want to lock down or provide vertical restraint wherever possible/ practicable so as to remove this issue from the plate of design issues.

In the old days (Punch Cards) systems were quite commonly "floated" based on T only displacements and W only loads... many a spring can was used some perhaps unnecessarily to the detriment of the installed systems cost and ongoing maintenance.

I have used the load case setup on and off for curiosities sake that Rich has laid out and it has agreed with the inactive support removed analysis. I suspect that it should be fine... however eliminating the lift-off eliminates the need for this unique load case.... I have always preferred simpler whenever possible/practical.

I do not agree with the philosophy that only "good" supports don't lift off. Often times refineries will have turn-around supports for the cold posistion that assist in maintenance. These act as a firm boundary in the cold posistion and cannot be ignored.

In the end this condition creates a bi-polar stress maybe its a collapsing load and maybe its a self-limiting thermal displacement... our job is to figure out what is most likely... and not to ignore the problem as many do.

In code parlance your remarks might be construed then as either a negative with comment or maybe an approved with comment. Many times I have had a more difficult time reconciling ballots that are approved with comment!

Keep up the good thinking....

I tend to agree with Kevin. If the operating case shows that there is lift off, I'll run a case with the support removed to check sustained stresses. For myself, as a compromise, if the load decreases, but doesn't go to zero, I take the sustained stresses with the support in place. I expect that, if the support has not actually been lifted off, then there will be some relaxation due to the secondary nature to bring load back on.

It is, perhaps, an arbitrary point, but it's the one I'm comfortable with. I think there is justification to say that there is some secondary nature involved, which is why I allow the support to "count" if there is still some load on in operation.

I've heard a suggestion for the cases where the load at a point goes from say 1000lbs in the SUS case to 100lbs in the OPE case. That is to make a model with a constant 100lbs force to replace the support point and check the sustained stresses that way. For me, that seems too extreme.

I have never been comfortable with the treatment of lifted off supports described by Richard. Admittedly, I haven't got both feet in the stirrups on this one or whatever you say in Tx so I am not offering a definitive statement.

The Caesar method handles only the warm up case and refers to the pipe sagging back onto the support. IMO stress range is a 2 way thing. In stress calcs we consider the cold to hot load case. More specifically in calcs we consider the range from the installed position to hot position at the end of the first heat up.

In reality we should really be calculating stress range based on actual cold to actual hot positions and the reverse.
It is easy to show that the installed to first hot position and actual cold - hot - cold positions are not always the same. (Not just for lifted off supports). If the lifted off support has sagged back onto the support when hot, my gut feel is that the worst case stress range would be from the actual sagged hot to actual cold positions. It's been on my good intentions list to try to calculate this but I have never got around to it. The horse won't stand still for long enough.

Being a simple sole, I assume that if a support can be shown to be not required in the hot case, then there is not much reason to have it in the cold case. I do the calcs with the lifted supports removed from the calc. I'd always check to make sure it didn't have a negative affect if it was physically left in.

If the support is requred then I'd look at a reroute or a spring. I seem to use fewer springs that some folk I know, so I don't think the simple approach is having much affect on pipe system cost.

(By "actual" I mean the position adopted after the pipe has settled down after several cycles).

Andrew

Yes, this is very possible. For a discussion of this, read the July 2003 M.E.N. article on expansion stress evaluation, begining on page 11.

Dear Mr.Richard Ay,

For a Hot-sustained Case,

As per Caesar-II recommendations the following are the load cases,

L1 = W+T+P - Operating
L2 = W+P - Sustained
L3 = L1-L2 - Expansion
L4 = T - Expansion
L5 = L1-L4 - Hot Sustained

My query - Is it necessary to check for the expansion stress (i.e.) L4=T also during the hot sustained or only the hot sustained stress is sufficient (i.e.) L5 = L1-L4?

Please clarify.....

That case (case 4) is not a "range" as the Codes require for the Expansion check. It (case 4) also doesn't account for any possible effects of non-linear boundary conditions. So my answer to your question is "no, case 4 is a construction case only".

Just a curiosity,

How many linear systems do a stress engineer encounter in a refinery?
In our modern time I think that we have no more.

We try to make them non linear. can be a poll open for this issue?

Best regards,

Dear Mr.Richard AY ,

While I proceed Hot sustained stress check as defined in CII technical paper , I found there is something problem as below,

L1 = W+T+P - Operating
L2 = W+P - Sustained
L3 = L1-L2 - Expansion
L4 = T - Expansion
L5 = L1-L4 - Hot Sustained

When I see the "restraints summary extend" report ;however,the liftoff support node is shown FY=0 and Dy=-0.005 in. in case L5 (hot sustained). How to explain why the displacements is still "negative" without any vertical supporting force ? As we know (OP temp > ambient temp) , if Fy=o is shown at any support node , it means the support node displacement should be larger than zero.I don't know how to treat the new "phenomeon" , please give your advice.

Best regards,

L5 is built to calculate stress. Displacements and restraint loads for these subtracted loads (L1-L4 or even L1-L2) have no structural meaning.

Dear Dave,

While we see the "restraints summary extended" report in case L3(L1-L2),
The "sign" of vertical force vector is always same as the vertical displacement vector.In case L3 (EXP), my understanding is that there will be some liftup (dy=+value) resulted from thermal force is positive and it is indeed.
So I think the restraints summary extened report in case L3 is still helpful for me to see how the piping system deforms and get an appropriate
explanation.
However , in "Hot sustained" case , I can't understand what's the meaning.
If we ignore the restraints report and just check stress in "Hot sustained" case , maybe it should be reported individually by CII.

Best regards,

The problem is that you are trying to apply the "hot sustained" method over a non-linear system. In this case your results may be not the one that everybody expect. For example in the attached file you may note that even if you actually do not have lift-off supports you have different results for "sustained" and "hot sustained". I expect that it shouldn't be any difference but it is.

Therefore in case of non-linear systems, use other method.

Best regards,

Operating minus thermal can provide a useful way to collect "hot sustained" stress in nonlinear support systems. Restraint loads from this case do not have similar value. If the W+P (SUS) gives the restraint loads when the system is cold and W+T+P+D gives the restraint loads when the system is hot, I do not see a reason to consider other restraint reports.

Dave, you are right. The only thing that is not clear for me is why the EXP load case (OPE-SUS) is not equal in stress result with T load case.The only explanation that I have is that T alone make no sense. The pipe will try to expand where it can, even up. Therefore we will have vertical displacements even where they do not exist. If these displacements do not reflect reality, the stress will not reflect reality because is directly related to displacements.

Maybe I can not make myself clear, sorry for my english.

Best regards,

Using your terms (OPE-SUS) will equal T alone in linear systems. Nonlinear supports (typically vertical) complicate things.
Here on earth, gravity has the pipe resting on the support when it is cold. As you add thermal strain, the pipe most first overcome this deadweight load before any movement occurs. And the final position of pipe will incorporate that deadweight sag along with the thermal strain.
The piping codes are quite clear in their fatigue evaluation - it's a range of displacement. That's why we use (OPE-SUS)

Think of your T alone as an analysis in outer space - no gravity.

I agree. This is why I think that "hot sustained" is not valid for non-linear systems (i.e. on earth)

I see the things this way:

- there is a concept "hot sustained"
- this concept need to be validated
- validation can be done against something reliable
- need to take a system with no lift supports and to apply this "hot sustained" concept (please see the file)
- results need to be the same

But the results are not the same.

Which one is not right? "hot sustained" or classic way?

I simply can not believe in an answer type "both are right"

Best regards,

The longitudinal stress due to sustained loads is not constant - we often call it sustained stress but the Code does not. Let me call it SL here. Up until this point, we have been discussing SL based on support configuration; you get one number without liftoff and another number with liftoff. But SL doesn't "jump" from one value to another. It is a gradual transition based on the amount of thermal strain added to the system. If you could monitor the operating load on a +Y (vertical) restraint, you will see the load drop with no movement until the thermal strain up equals the deadweight down. At that point, the support then disengages. So what happens to the force-based (i.e. deadweight) load distribution as the system heats up? It changes. So SL changes.
It's not so black & white; SL changes with thermal strain. But looking at black & white will set the limits to SL and that's all we need.

My problem is that I get one number without liftoff and another number without liftoff. (please see the file).

But I agree that after all we have only aproximations of what is real and that Caesar is only a tool (even if is the best on market).

Personally I will continue to use the method with another file with lift off support removed.

Best regards,

Please see attached a slightly modified file.

How can be explained this?

LOAD CASE DEFINITION KEY

CASE 2 (SUS) W+P1
CASE 5 (SUS) L5=L1-L3


Piping Code: B31.3 = B31.3 -2006, May 31, 2007


CODE STRESS CHECK PASSED : LOADCASE 2 (SUS) W+P1

Highest Stresses: (lb./sq.in.) LOADCASE 2 (SUS) W+P1
CodeStress Ratio (%): 48.8 @Node 120
Code Stress: 5272.7 Allowable: 10800.0
Axial Stress: 933.0 @Node 128
Bending Stress: 4339.7 @Node 120
Torsion Stress: 40.7 @Node 109
Hoop Stress: 1920.0 @Node 20
3D Max Intensity: 5272.7 @Node 120

CODE STRESS CHECK FAILED : LOADCASE 5 (SUS) L5=L1-L3

Highest Stresses: (lb./sq.in.) LOADCASE 5 (SUS) L5=L1-L3
CodeStress Ratio (%): 109.7 @Node 69
Code Stress: 11846.7 Allowable: 10800.0
Axial Stress: 924.8 @Node 98
Bending Stress: 11006.0 @Node 69
Torsion Stress: 2030.5 @Node 80
Hoop Stress: 1920.0 @Node 20
3D Max Intensity: 12131.3 @Node 69


LISTING OF STATIC LOAD CASES FOR THIS ANALYSIS

1 (OPE) W+T1+P1
2 (SUS) W+P1
3 (EXP) T1
4 (EXP) L4=L1-L2
5 (SUS) L5=L1-L3


3D Max Intensity: 12131.3 @Node 69


Best regards,

agreed with auditor,

it gives good result,

as some times in perticular support sustain load showa 1000 N force while at Operating case it shows 500 N then where the remainning 500 N goes, ecxactly it will distribute to other supports.

by using this method we can find out Hot sustain stress too.

How about this file. It is a simple flat turn. It shouldn't fail, yet......

Best regards,

Read all of this thread.

http://65.57.255.42/ubbthreads/ubbthreads.php?ubb=showflat&Main=2046&Number=8306#Post8306