Hello all,
In Caesar's on-line documentation for expansion joint's effective ID, it states that "If left blank, or zero, then no axial thrust force due to pressure will be calculated."
My question is when do we intentionally leave this data blank or zero? Which scenarios do we not calculate the pressure thrust for expansion joints?

Thanks.

I am not an expert but I suggest that you:

1. Continue reading the caesar II documentations regarding expansion joints and the answer is there.
2. Search this forum, a lot of experts have tackled expansion joints many times here. Some even gave very informative websites.

Goodluck.

For me if you model the expansion joint correctly you can always enter the effective ID. To leave it out can be dangerous. As a very rough figure from NB200 up use pipe OD plus 60 mm. Below maybe OD plus 40. Try to get the correct value always.

Where you do not need to enter?
- gimbal
- hinge
- tied compensator most of the time. To prevent problems I would enter always.

Then there is the issue of nozzle load. Full pressure thrust goes to the vessel anchor but only effective ID less pipe OD on the nozzle (if you do not have hinges or rods on other anchors).

Be carefull on this pressure thrust. It can be huge. Last item I designed had over 800 kN and the pressure is not even high. In this case there are 4 of them installed side by side. 4 x 800 kN. I would say you can move a building.

Jouko Sir, for 800 KN, the structure may be too high, why you didn't go for Inlined proessure balanced bellow in that case..

I saved the world ;-) I used word "had". Somebody made a small mistake. One weld missing and it was not axial pressure balanced. Four of them would have lifted the building with some equipment. Luckily I saw it. These things happen.

I know cases where the pressure thrust was not considered and lines went to service. Some lasted couple seconds some a bit longer. Even when the issue is considered there are failures. It doesn't take long to find under designed compensator on some company site.

oh ok...i didn't notice the word had sir.

I am using simple expansion bellow in hot water service. I switch off bellow effective Id & manually enter force at bellow pump nozzle interface ( force calculated based on bellow effective area –internal pipe area ) X pressure.
Whereas in other side of bellow total effective area x pressure force is considered. See attached sketch.
I notice that if I do not put any axial stop in the line I am getting very small axial load in both sustain & operating case.
Since I put both F1 & F2 force in a line so some balancing of force is there based on axial stiffness of bellow.
I doubt if my understanding with output result is correct or not.
This kind of thing do not appear when effective id of bellow in enter in Caesar, since in that case equal & opposite force is there in both side of bellow.
In my case I am considering unequal force & getting very small load on nozzle.

The pump nozzle load will (Face of flange) will be based on F1 alone the F2 force will push against the impeller or back wall of the pump case.

BUT without a tie rod set up the joint may become hyperextended due to pressure thrust!

I'm not a big fan of what you have here its a dicey bit of work!

i would suggest the bellows should be tied and on the equipment, not the pump. (better still -re-route)

by removing the ties you require a anchor within 4d of the bellows which will not help the equipment nozzle
By removing the ties you are pushing all of the pressure thrust into the pump (impeller).

Whilst this removes said forces off the pump nozzles, and helps you achieve pump code compliance, the pressure thrust is transfered to the baseplate.
you'd better make sure the pump can transfer that force and that the base plate is well built. grouting the pumps help here.

Not good.


Edit (JCL beat me to it- dang!)

Sorry, you cannot do what you are planning. You have a good change of popping that compensator (No guides) and you have nice 48 kNm bending moment on your nozzle. In addition you have lateral/bending movement on that compensator.

Cheap solution: Use 2 or 4 Viking Johnson flange adaptors with 2 chains per adaptor. Your forces are next to nothing.

Actually I spent a day at a power house looking at the same sort of ill advised design somebody else had cooked up. The huge pump ~ 15 feet from base to FOF discharge was wearing out seals like mad....

So I got to go out and look at somebody else's creative work. Anyhow when I pointed this out as a possible source of the problem all I heard was no thats not possible because the EJ takes all the load off the nozzle (as though it had magical properties). I had Dial indicators installed and zeroed with no line pressure in the line and then pressurized the line. The pump casing was moving over 1/8" in spots!!!!

So go ahead with your dicey design if it doesn't work out somebody else will fix it eventually.

Sorry Superpiper. My answer was to SHR post. Must learn to press correct button.

Yes, people do not realize the power of the pressure thrust. What about calling bellows as hydraulic cylinder? That should wake up some people.

Thanks everybody
For your suggestions.
I miss one point in previous sketch.
Sketch was for pump suction line & pump was sucking hot water from tank. The pressure in this line was only elevational head of the tank & that is 1.4 barg maximum as per process calculation.
Yes I agree we can not use a untied expansion bellow for pump discharge line due to heavy pressure thrust load.

Two years back I was working for a client who made typical plant & operate . Since they are expert on typical plant I just follow there guidelines.........
For cooling water discharge I was told to use Tied expansion bellow.
And for suction line ( pump was sucking water for tank with no much flexibility) use untied rubber expansion bellow.


Once I made something does not mean I should follow the same all the times.
But still I feel that I can think to use untied expansion bellow for Pump suction line , obviously not for discharge.
What is suggestion for our expert ?? please.

And If I can go for untied bellow for pump suction line my concept marked on previous sketch is valid ?
Because I am curious about sustain axial load 630 N on the pump nozzle when I do not put any axial stop on line. whereas F1 load as marked on isometric is 2930N

Thanks once again.

I disagree with all of this


Suction lines do see full pressure (walk around your plant and look for PI's on the pump suctions) don't ignore it its not SEP



No no no no no no no no no no ......................
Never EVER copy something without appreciating why it was done in the first place



If this was a concrete bunded CT then maybe, if the pumps were well fixed.
if it was to a steel tray CT then the vibrations and pressure thrust could be problematic, the allowables on steel tray CT's are Zero.


absolutly use an untied bellows, if its the right thing to do.
which if you = anchor the pipework, ensure the plinths are ok, ensure, the pump manufacturer is ok, ensure the bellows is appropriate then fine.



Not valid. see points above

Edit:
To anybody installing flexibles in a piping system.
There all different !!!!!!!!
You CANNOT put one in system A because BOB put one in system B!!!!
If somebody jumps up and down and shouts at you to put one in, you can't!!!
You have to sit down and think it through from start to finish.
In this situation, the use of bellows could be negated by good piping design which is surely better.

I HATE putting flexibles in pipework, simply because it takes time for me to reason it through and then have a second opinion check my reasoning.