Hello Folks,

I am working on a problem and there are 2 different sizes of spring hangers located at the same location. This is a vertical run of pipe supported by spring cans through trunnion. My question is why there are 2 different size spring cans selected for the same location. I could not find the previously done stress analysis on this piping, but even if I let the CAESAR-II to reselect the springs, it is doing the same, i.e selecting 2 different sizes of the springs.
Could the experienced forum members assist me regarding this issue, If it is okay to have 2 different sizes spring cans at the same location. If not how can I force CAESAR-II to select the same size spring cans.

I hope I am making sense.

Regards

I'm going to assume you have a vertical run of pipe with two trunnions, separated by 180 degrees, each being supported with a spring.

If the line itself has any amount of rotation, the two springs are going to have a different load through the various design cases. CAESAR doesn't really understand that these are two springs at virtually the same location.

There's more than one way to approach this. My preferred method is as follows:

1. Place a single spring in the center of the vertical pipe at the trunnions, and tell CAESAR quantity 2. Save this file as "spring size." Run and read the cold load and spring rate the output.

2. Input these values as user-sized springs on the trunnions. Run normally, but also make sure you look at the load range of the springs and ensure they both meet the manufacturer's recommended load range. If not, you're going to need to manually size and input springs until they do pass.

Thank you Michael for taking the time to reply and help me clear my doubts. I have followed a little different approach to tackle this. I have put a +Y support at the node and read the sustained load at the point and dividing that load in 2 and putting those values as operating load in the Hanger Control window.

This job is having 18 springs and I have found if I manually change the cold load values to greater values on some of those spring cans than before, I get failure at some reducers in Sustained case. Any insight on that scenario?

I thought your checking on the working plant, isn't it?

If the installed spring can has different sizes, than the spring rate or the set load could be differ. Check out the spring vendor catalogue or the detail drawing if you can.

Before any advice your second question,you must check why and how you change the spring load. The spring hanger can be variable type or constant type. Depend on the type,different approch shall be taken. Also, you need to check the restraint summary thoroughly. Check the displacement at the sustained case(too much lift-up or sagging), or the changing operating load caused lines to bend or reaction force from other support.

Spring hanger's operating load can be adjusted for a small amount of load, but it has its limitation(See MSS SP-58). So, your first approach may not realistic.

Right - changing the cold load means literally adding cold spring to your system, which is not permissible for getting pipe to pass, and is partially permissible to get equipment to pass (assuming B31 code).

Well, I have checked all the detail drawings and datasheet for all the spring cans installed in the plant. After the field check, I found all of them are out of setting i.e. none match to the Datasheets' C.L and H.L.

I have redesigned some of the spring hangers but will be replacing different size spring cans at the same location.

As I am unable to find the previously done stress analysis on this system, I believe the reason for having different size spring cans at the same location is because CAESAR-II gives 2 different sizes.

Michael, Cold load setting on variable spring hangers always make the piping out of balance in sustained case, as we approach to balance the piping load in operating condition. So isn't that also means we are adding cold spring to the system?

Michael, by B31.3 319.2.4 Cold spring is the intentional deformation of piping during assembly to produce a desired initial displacement and reaction.

Cold Load of a spring support is a force applied to piping system in "cold" status.

It is true that literally "Cold Load of a spring support" can be shorten to "cold spring", however proceeding in this way we can speculate everything we want. Anyway, for me it's the same connection as to cold war and cold box...

cold load setting will make the piping system out of balance in operating condition right?

Jay, it is hard to answer because it depends what you mean by "out of balance". You may consider "cold load" and "hot load" as 2 sets of forces applied where the spring supports are placed. The forces are active forces, rather than "reaction forces" as for regular supports/ restrains.
On each status- "cold", "hot" and any intermediate one, the piping system is in equilibrium, if proper designed.

But the spring forces are not arbitrary, their selection is a result of an algorithm with a target that you may call "balance status". In this case I would answer "yes" to your question.

SandeepTiwana,

Based on what you say, the previous analyis or site work(installing) has not been properly done, or piping has been damaged due to primary or secondary stresses.

Owner or operator may have "hanger record sheet"(or whatever it called), and it's good to track how spring supports changed.

My opinion was if you were not have authority to change pre-installed spring, you should not fool with pre-defined spring load. In your case you should do on your own analysis.

For what you say 'CAESAR-II gives 2 different sizes.' if the line expand like a bow or the support point has rotation, that could happen. Place guide or stopper near there and for the spring please follow what Micheal suggested.(whatch for the spring travel range and variability though)

If the cold load setting of the spring is 2000 units and the cold force from the piping onto the spring is 1000 units, we can rest assured that the spring, if released, will deform the pipe to the tune of 1000 units, potentially leading to failures.

It sounds like this has already happened based on Sandeep's query.

With regards to my statement, the code states:



...which is word salad for if your system behaves cyclically, it doesn't count, and if it doesn't behave cyclically, it counts. Can you guarantee the operation of such a system? The only lines you can guarantee are this way are the lifeline of the facility, and I'd do my best to make them make it work without cold spring.

The code also states:


... which is actually different from what I remember, so it's not actually "partial credit," but "design 2 additional models."

On the same logic, any cold load of spring support that is not able to give mathematically exact 0.0000 inch as deflection in that point of piping system would be classified as cold spring.

I wouldn't consider mathematical, manufacturer imprecision, or minor adjustments as "intentional" cold spring.

I should not have stated that is flat out impermissible. The code then goes on to provide examples where you might use cold spring:



My impression of the "word salad" in the previous post is you can use cold spring in the following cases:
1) Pipe failure in a sustained case, but no pipe failure in an expansion or occasional case, and a spring that's not cold-sprung doesn't cause it to pass, but a cold-sprung spring does.
2) No pipe failure, but nozzle/support failure with non-cold-sprung springs, where cold-sprung springs work.

In both cases, you must run the analysis in triplicate. At 1x specified cold load, 2/3, and 1.25x. All must pass.

So from the above discussion, should I say it is not a good engineering practice to manually tune up the spring can to hold more weight in the Operating Case?

Also it wont be good as a responsible engineer if I say CAESAR-II told me to select those spring hangers with the program delivered Cold Load and Hot Load.

Michael, no doubt, by definition, cold spring is an "intentional deformation" but in back of the definition was the fact the cold spring was supposed to be made by "cutting short" or "cutting long". In this case, the intentional deformation is rather an intentional strain.

In an old Intergraph/Coade training document there is a nice definition:
"Cold Load (CL) is the intentionally incorrect load at which the spring is pre-set, in order to get to the Hot Load after moving."

What you essentially said is that that applying a force (load) in a piping system results in deformation of system and must obey to rules of "cold spring".
It is your interpretation and maybe is in harmony with the definition, but I do not second your opinion. On the same logic, applying forces on a beam simply sustained will get "deformation" of the beam, however nobody defines this fact as "cold spring".

You may have in practice spring supports with SUS displacement as a quarter of inch. One contractor may consider 0.25" is within construction tolerances or as you said "manufacturer imprecision".

In the end, nothing wrong to consider such deflection under CL as "cold spring"- just curious you tell Caesar there is a "cold spring"?

I remain on my opinion: a spring support is a force in system.

Hi all,

I subscribe entirely to "Mariog" thoughts synthesized above.

Really, it's strange and beyond sound engineering judgement how people may assimilate spring supports as a cold-spring spring provision.

Here is a quote from an older Ceasar II Technical Reference Manual (...those old and good times when engineers used to be real engineers, preoccupied of sound judgement and fundamentals understanding, and less of graphical interfaces and "magical power" of dialog windows' buttons...):

<<
Cold spring is the process of offsetting (or pre-loading) the piping system with displacement loads (usually accomplished by cutting short or long the pipe runs between two anchors) for the purpose of reducing the
absolute expansion load on the system. Cold spring is used to do the following:
- hasten the thermal shakedown of the system in fewer operating cycles
- reduce the magnitude of loads on equipment and restraints, since often, only a single application of a large load is sufficient to damage these elements.
>>

So, there is no logical association or connection between cold-spring and spring support provision.

As "Mariog" pointed out above, cold spring/support is just an aplication of a force - in fact it's an ELASTIC Restraint, which provides a reaction force following pipe deflection.
Sound engineering practice recommends NOT TO DEVELOP initial deflections and/or stress states at installation by the spring supports. Such approach is a "cheating" practice and might yield to unpleasant incidents during installation operations. In other words, it is a wrong and unprofessional approach to set the installation/cold load value a high values to induce initial upward movement and/or initial restraint's loads opposite to those develop under operating/"hot" conditions.

On the contrary, Cold-Spring is the development (at installation) of an initial stress & internal forces' state that is opposite to that developed under operating/"hot" conditions.
It should be noted that cold-spring does not reduce the displacement stress range, it just reduces the restraints' reaction loads level.

I suggest to "beginners" or less experienced "engineers" (!), to read again and carefully Caesar II manuals, some reference specialty books (still recommended on Coade website) and to search and read older posts on this subject, which may be found on this Forum. Fortunately, high level professionals (such as Mr. John Luf or Mr. John Breen) left available a lot of interesting topics on this matter and many others...

Regards,

I do not know if your this comment "I suggest to "beginners" or less experienced "engineers" (!), to read again and carefully Caesar II manuals," was directed at me, but I beg to differ on your suggestion that cold load values should not be tweaked. Even CAESAR-II has a very good tool "Operating Load" in hanger design window to fine-tune springs. Once analysis is done, if we want to tweak a spring, increase a load or decrease a load, we could use a number here to override the calculated dead weight at that point. This is a way to do some more precise work around rotating equipment.
I would also like if Dave or Richard can also give their viewpoint on this topic.

Dear Sandeep

Lets say you are using a special type of constant supports rather than spring supports. Long time ago- but I've seen such supports in an old refinery in 90s- a constant effort support was a pulley carrying the pipe on one side of a wire rope and on the other a mass having weight as desired. The tension in both sides of the rope is equal such that if the pipe moves downwards the weight moves upwards with zero effort or change in rope tension. With that special construction, it was quite evident that we talk about a force rather than a support.

With such construction, your system would be one having restrains -otherwise the system is rather a mechanism- and active forces applied in points where you placed such "supports".

Now you may imagine your system as having rest supports instead forces applied. With Caesar you are able to calculate the rest supports loads. Having the numbers you may choose exactly the weights of the "special supports" I've mentioned above. I would say that you have balanced the system in cold condition and the shape will be exactly as the system is on rests.

But you can choose the weights based on other criteria. The system will have the same restrains and other set of active forces= weights and will have other equilibrium shape. Some engineers would say you are wrong because you placed "intentional deformation" by no transparent criteria, the contractor may be terrified because the deflections that they can see, etc. However it may happen that you solve your target (acceptable stress, desired forces/ moments in restrains), even the result is "ugly" for the majority (and some will claim you made a mess of your system).
But a good question is what is really wrong there? In my understanding nothing wrong looking only to "cold status". Can be wrong when the system is "hot" and there are not only the "forces" but also the temperature as load. This is the "hot case". In case the analysis of "hot case" gives acceptable results vs criteria that you (and the Code) define, what is really wrong?

Now, you have the same situation when choose the cold load of constant supports. With spring supports it's different because the forces depend on the local displacement. But, in my opinion, what I've written above still hold.

As Dorin pointed-out, sound engineering practice recommends NOT TO DEVELOP initial deflections and/or stress states at installation by the spring supports. Just keep in mind that, in this case, Sound engineering is a result of opinion of majority. I rather belong to a minority for this case and it is my opinion that you do not need to worry about the cold condition displacement. You only need to specify the Cold Load and the Hot Load and the travel between them, a proper load variation, a proper length/ height geometrical definition of spring supports and goes away with a proper analyze of the piping system. keep in mind that Caesar has a good and performant algorithm to choose spring supports but you have the freedom to input User data and analyses what happens with your system. And, as I said, I do not consider the problem of "cold spring" as argument of "avoiding deflections" in cold status (with the notable exception when Michael will be my Client!).

Thank you Mariog for the response. I fully agree with you. I did tweak the cold load a little but not as much that my system will have high deflection in cold position. The maximum deflection in the system (spring hanger location) after I tweaked the cold load is around 1.2mm which is negligible.

Also as per L.C Peng, Industry standards [5] limit the load variation to no more than 25%. This means that, regardless of the pipe movement, the spring will carry no more than 125% and no less than 75% of the properly balanced portion of load. The unbalanced load will be shifted to the neighboring supports or equipment.
In actual applications, a smaller design load variation may be specified for critical piping systems,such as the ones operating at creep range or connected to rotating equipment.

Mario,

In response to your question:


I would only model in cold spring in order to emulate an existing phenomenon in order to attain the current status of something requiring investigation.

But if I were to include cold spring, I would indeed replicate it within CAESAR, despite it being a time-intensive task.

In this day and age of GUIs and window button "magic powers," we're expected more than ever to provide concrete numbers to our answers in a timely and cost-effective fashion.

Just a few points...
If the spring itself is treated as a piping component - one with very low (axial) stiffness - the preload on that spring can be thought of as a cold pull (cut short). That's what you are doing when you adjust the spring to carry a pre-defined load. This "pull" is much better defined than cutting out a length of pipe since the manufacturer determines the true performance of the spring and the sets the provided preload at the shop. Conceptually, you are "pulling" the other end of the spring so that a predetermined load (along with the requisite stiffness) is applied to the system.
What complicates things is we seek to set a load for the operating position while we check our spring settings in its installed position. Since the spring moves from installed to operation, the installed load is out of balance (either too high or too low). (A low load variation will limit this out of balance condition.) Matters are further complicated if the system load in the installed position (empty, e.g., WNC) does not match the system load in the operating position.
Any additional spring adjustment in an attempt to transfer load away from sensitive equipment may bring additional imbalance to the installed state.
Do what you have to do to get the designed operating loads correct and check your work. Then make sure these springs show the proper operating load in the field.

Ok, not sure I should open a new thread (because my post is rather focused on the cold spring) but I try to continue here.

I would like to propose you an exercise.

Let's say a contractor erected a piping system 4"-Std, A53B, CA=0(!) with just 2 anchors, 2 vertical legs (3000mm, 2000 mm) and 2 horizontal legs (5000mm and 3500mm) described as bellow.
100 (anchor)- 200 (bend), 3000 mm up
200- 300 by x=4500m, 300-400 (bend) by x=500 mm
400-500 (bend), 2000 mm up
500-600 (anchor) by x=3500 mm
The above description is the weightless theoretical shape or "as design" shape. The Contractor build it up in shop without any geometrical error. Of course when they erected the system, its shape became "as build shape". The system is deformed under the gravity load.
In node 300, the deflection under gravity is theoretically 10.5 mm and the value is sensitive to the actual capacity of anchors to restrain the rotations. The piping system would be considered as poor sustained vs. common practice, however cannot be classified as out of Code in Sustain case, unless other troubles are foreseen. But this is a "no flow" line, inside located and stayed in place for 6 months with "exactly" 10.5 mm deflection in node 300.


1. A very late inspection revealed the situation and the decision was to place a rest in node 300. A structural beam was available around, it is enough space to use a hydraulic jack, to lift the pipe up in node 300, install the rest and carefully release the jack. The piping rest now on a support as it was better to be from the beginning i.e on the "as designed" geometry. They modified the "as build" shape and now, in node 300, the deflection vs "as designed" shape is zero. But fact is that deformed the old "as-build" system with 10.5 mm in node 300.

2. The theoretical analysis shows that rest load in 300 is a vertical force of 1005 N for Sus case. That means that applying a force of 1005 N reduces to zero the deflection of "as built" shape in node 300. In fact they already applied that force because the hydraulic jack provides a force; the deflection of 10.5 mm is the result of the force applied to allow support installation.

3. The line is used once per year for a special procedure handling hot gas, low pressure. Instead to put a rest in 300 I would prefer to place there a spring support. It may be one with cold load of exactly 1005 N and everybody will start saying that it is not a good procedure. I shall use "the good procedure" that indicates to put there a spring support of 960 N as cold load. When install it, this result in a down deflection of 0.5 mm vs "as designed", but to install it I have to jack up the system with about 10 mm in node 300. And, I repeat, do not be focused on what will happen in hot case, it was analyzed and everything is OK.

4. The Contractor was stupid or clever (the matter is still under debate) and decided to choose a spring support with cold load of 900 N, with different size vs point 3. They can do some structural changes there and they planned to jack the system up to place it. But they purchased one of 1100 N cold load of different size and jack it up there. Let say in both cases the hot analyze still looks well- do not be focused on the details.


5. Let say they purchased the support of case 3 but they succeeded to adjust it (non-recommended by Manufacturer) to 1050 N. They use the same procedure to install it, i.e. jack the pipe up by about 12 mm.

I would like to know in which case of this stupid exercise I introduced the cold spring and how to count it.

The point of this example is that the magnitude of deflection under gravity in absence of the support is greater than the "adjustments" operated by changing "cold load" and to install the support I have to reverse the gravity deflection by applying forces in the elastic system.

Thanks.