Metal Expansion Joint

[marchieV]

Hello.

It is my first time to design an expansion joint and I need some advice or tips regarding the design of expansion joint.

Actually, I need the parameters for Bellow Stiffness (for CAESAR II Input) and according to my research the vendor will provide those stiffness.
I was wondering on how the vendor will come out on such values for stiffness if the system in the CAESAR II did not yet run?
Do they have a pre-qualified selection/table? (The same like in the case of spring selection that the vendor will usually provide the table for selecting the spring based on the generated CAESAR II load of the pipe.)

In short, what data does the vendor needs to know upon requesting for the design of expansion joint?

Please help me to understand the process of designing an expansion for I have no experience on it.

Thank you and hope for your support.

Sincerely,

marchie

 

[JG MechEng]

I believe the vendor develops the stiffness with a combination of FEA and or EJMA calculations. The vendor will need to know the allowable expansion joint travel in each direction. This is some times identified in client specifications and if not, you will need to work with the vendor for a reasonable number. they can tell you what is commonly specified depending on the type of joint being specified. You will also need to know if you have to worry about product condensation corrosion in your bellows so they may need to be insulated to keep temperatures above dew point of the process fluid. There is also a variance in the number of cycles is reported/calculated between EJMA and ASME. This is also an area I would clear up with the client and vendor. If you have a double bellows system, how will you monitor the internal pressure...again that has been an area that was always a discussion? I would start with EJMA, and then get an expansion joint vendor on the phone. I have rarely not been able to call a vendor and ask for help. you really wont be doing much "design". you should put a place holder in your model with the overall size, weight, movements, and spring constants / stops in the directions you require. the vendor will be able to give you starting numbers and when you get his design nailed down, you plug in and make adjustments.

Its been a few years since I have specified an expansion joint, but if you want to talk more in depth, PM me.

 

[marchieV]

jguil,

Thank you for your knowledge.

Actually, I am doing an investigation regarding the failure of an expansion bellow.

The situation is like this...

The existing pipe (built: year 1981) has an existing expansion bellow but in several years these bellow was broken. They decided to replace it with new bellow (using the data from old requisition/specs.) but with different vendor. But upon installation and testing as the maintenance reported, the new bellow only last for 3 hours and there is already a leak in the weld joint. They decided to temporarily install a pipe spool in replaced with the broken bellow and they wonder about the effect of the spool (although the leakage was solved after installing the spool). They were not sure about the future effect of it in the pipe.

Upon investigation, I decided to ask for their oldest bellow (the one that last for several years) and I saw a flat bar (not tie rods) welded in both flanges of the oldest bellow. Now my question is, Is this right or acceptable according to your experience? I thought tie rod is the only way to restrict the axial movement/thust. The new bellow that last for 3 hours only used tie rod. And I was suspecting maybe the welded flat bar (OLDEST BELLOW) vs. the tie rod (NEW BELLOW) is the main difference of their design. Am I right? Correct me if I am wrong.

Please help me clarify the issue and hope for your opinion.

[See attachment for the photo. I deleted the word "NOT"]

Sincerely,

marchie

 

[DSB123]

The flat bars are effectively constraining the axial growth/contraction of the bellows unit. This could be as per design such that the expansion of the length of pipe is taken by the bellows convolutions and the pressure thrust is accommodated by the flat bars and not transferred to the adjacent pipe.

 

[marchieV]

DSB123,

I agree with you. Does the effect of these flat bars will be the same with tie rods?
The client actually wondering why the new bellow (w/ tie rods) only last for 3 hrs while the old bellow (w/ flat bars) last for several years? Both old and new bellow have the same design specification but different vendor. What's wrong with the new bellow?

 

[LittleInch]

Can you get more details / photos of those flat bars because its difficult to see how they are attached or not at each end.

Are you sure these bars were meant to be permanent? They look to me like they were there to protect the bellows during shipping??

If they are meant to be permanent and assuming one end is free, they are limiting the bellows in x and y planes, but not Z. If both ends are fixed then it is not clear what the bellows are doing.

Tie rods only restrain movement if it goes beyond a certain axial( Z) component and do very little in X and Y, though this does depend on the design and fixing of the tie rod ends - do you have a photo or drawing of the replacement bellows?

however I can't see any serious welding on either end of those flat bars so it's a bit of a puzzle at the moment.

I can only guess at the moment that the tie-rods are not limiting some sort of movement which is outside the bellows movement limits which the flat bars prevented. However we don't have the data yet.



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

 

[pennpiper]

marchieV
Question: The picture of the old Expansion Joint appears to have the two flat bars welded at 120 degrees apart. Is that correct? Is there a third bar on the far side where we do not see it in the picture?

If there are three bars this would result in a very rigid element and the bellows would be nonfunctional. Why did the original (1981) installation require/have an expansion joint?

You did not provide any information on the commodity. What is it? What is the operating Temperature? What is the Operation Pressure? The Piping appears to be Class 150, is that correct?



Sometimes its possible to do all the right things and still get bad results

 

[DSB123]

pennpiper,
"..the bellows would be non-functional". This is an incorrect statement. Yes one end of the bellows cannot move relative to the other du to the tie bars but the bellows would accommodate the expension between the two ends, similar to an axial tied bellows. So unless you know the exact reason for the bellows installation you cannot say "the bellows would be non-functional.

 

[pennpiper]

DSB123
If those bars are welded to both EJ flanges, there will be no movement between the two ends. Therefore the joint is worthless.

Another question: Why is there no EJ on the Suction side of the Pump?

Sometimes its possible to do all the right things and still get bad results

 

[Ehzin]

Agree with the two of you on this one for separate comments. I'd imagine the bars lock the face to face of the expansion joint making it a fixed length component regardless of the what the bellows wants to do. It will at least remove some axial thermal growth if that's all it's meant to address. And agree with DSB on that fact that a lot more information is needed to say anything definitively.

Temperature and a rough idea of pipe layout would likely give quite a bit of information about this situation. It may be that the original design info on the datasheet was inadequate but you happened to end up with a more robust expansion joint than necessary. And now the replacement was designed to just meet the design conditions and it doesn't meet what it's actually experiencing in the field. A bit of speculation there.

Thanks,
Ehzin

 

[marchieV]

Dear All,

Thank you for all your comment.

Actually, I did not take those pictures but I guess it might be 3 flat bars welded at 120deg and the flanges used are slip-on & 300#. Also they just installed those spools without consulting first that is why all the information that was given by the client is all history. The datasheet they provided are very old and incomplete. But based on my initial analysis (using the spool) the pump nozzle is exceeding the allowable (API-610) at Moment that is why the old design has a bellow. I am currently asking the vendor regarding the design specification of the new bellow (that last only 3 hrs), maybe there is some discrepancy with the old and new bellow's design specification.

It is really hard to gather all the information specially if the plant was built 3 decades ago and I did not saw the actual situation.

Sincerely,

marchie

 

[pennpiper]

Someone out there Please! tell me (us) how that original Expansion Joint with the Three Bars welded to both the inlet (lower) flange and outlet (upper) flange is going to work? To my way of thinking it was a "rigid" non flexible element.

Sometimes its possible to do all the right things and still get bad results

 

[Ehzin]

Pennpiper, I don't think you'll find a logical answer to that question since I don't believe there is one. If it only had bars at 180° it could allow for minor lateral movement between the two flanges but 3 bars at 120° would restrict this. The easiest visualization of this would be a long horizontal exchanger with minor temperature increases connected to large bore piping that runs straight underground over a short length. The nozzle location will lengthen out and the short large bore piping won't be able to provide necessary flexibility to limit stresses. The expansion joint would largely be there to accommodate this axial movement so locking it in at a fixed length may not be that problematic.

A possible scenario could be that the original expansion joint design was always inadequate but the bars kept displacement/distortion from damaging the expansion joint. It just happened to lead to excess pump flange loadings, although within reason since you haven't noted any major pump issues. When you installed it the "right" way the joint failed because the design was wrong from the beginning. Once again, this is speculation.

If you've any kind of modeling performed for the pipe routing with the expansion joint look and see what the axial/lateral/angular displacements are on the joint and compare to the data sheet. I'm not entirely positive what expansion joint data looks like in Caesar II reports but I imagine this information could be gleaned from it.

Thanks,
Ehzin

 

[Gator]

I agree with pennpiper - that EJ doesn't do anything. Maybe it "seemed like a good idea at the time" and then they realized oops and then welded the bars on to effectively make it a rigid piece.

How much linear expansion is there going to be in a vertical pump discharge line that's maybe 20 feet long?

Then again, LI said: "They look to me like they were there to protect the bellows during shipping?"

 

[europipe]

marchie, have You seen the stresscalcs?

 

[XL83NL]

Ive bought a metallic expansion joint in the past with such bars on them, if I recall also 3 at 120 degrees seperated. The bars were used for transportation, and to protect the convolutions during handling, before the joint would actually be serviced for use. The manual mentioned this, and included a paragraph on removal of the bars before taking the joint into service.

 

[LittleInch]

This though requires the pipe fitters to read the manual......

After a short period of time maybe everyone thought they were part of the design!

that's why nowadays things like this are painted yellow with a remove before use painted on it.

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

 

[XL83NL]

Ah, I forgot; the bars on the joint I mentioned were indeed paitned yellow/black to indicate removal before usage.

 

[LittleInch]

Maybe they didn't bother - it's still my vote at the moment because I couldn't see any significant welding on those bars - they almost look stuck on, added to the fact that they seem to invalidate the use of the joint...

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

 

[marchieV]

Dear All,

​

Our client decided to design new expansion joint but still I need your help because its my first time to design an expansion joint.

I just want to ask on how to design new expansion joint. The following questions are:
1.) What are the information that I need to submit to the vendor of expansion joint?
2.) How to determine the movements (Axial, Lateral, Angular)? Is that through CAESAR II displacement report from pump expansion (because the line is connected after the flange of pump discharge) OR from the location of the expansion joint?
3.) How to determine the thickness of the bellow?
4.) How to determine the no. of ply if it is connected to pump?

I will also read the EJMA but still I need your expertise to guide me through.

Thank you and hope for your support.

Sincerely,

marchie