Expansion joint 6

[Dj364]

Hello Every one,
I am running a 70” straight line (46 meters). I am designing a Flexible joint to absorb thermal expansion as well as thrust force.
Technically, the thermal growth is absorbed by the expansion joint and thrust force by tie rods,
I put expansion joint without tie rods, the flexibility is Ok
I put tie rods (1”1/2 diameter, spring rate is around 805333 lb/in), with gap forward to allow compression on bellows, consequently large forces caused by thermal growth.
Could anyone give me some guidelines to resolve the problem? Dealing with pressure thrust and flexibility.
Thanks in advance
Dj.

 

[C2it]

If you must use an axial bellows to absorb thermal expansion whilst avoiding end thrust forces, it really has to be a pressure compensating device. Any combination of tie-rods and pre-load is a numerical fix that cannot be relied upon to work in practice.

 

[GregLamberson]

Dj364

Assume you've done a flexibility analysis? You can find information on tie-rods for bellows in Appendix X of ASME B31.3. You should look at cycle life (fatigue analysis).


Greg Lamberson, BS, MBA
Consultant - Upstream Energy
Website:

http://www.oil-gas-consulting.com

 

[Dj364]

Thanks for your reply,
the bellow is acting in tension due to thrust force, and in compression due to thermal growth.
what kind of bellow is suitable? inline pressure balanced or single bellow with tie rods. I used them , it didn't solve the problem...
I 'll appreciate your recommendations
Dj.

 

[C2it]

In a straight line presumably between two fixed points with a bellows somewhere central, I can't see the relevance of a flexibility analysis, except with regard to the actual bellows convolutions. Nothing else is flexible as such, and axial forces are only a hand calc.

If you are attempting to absorb thermal expansion and are not concerned about pressure end thrust, use an axial unit with suitable pipe guides. If pressure thrust is significant, there is no otion but pressure balanced bellows in this locked in geometry. Tie rods will help at all in reality.

 

[jistre]

I agree completely with C2it. In a straight run of piping, a single bellows expansion joint with tie rods cannot solve the combined problems of loads on equipment caused by the pressure thrust of the joint and loads caused by the thermal expansion of the system.

Imagine an expansion joint that doesn't have tie rods. You pressurize the system, and find that the pressure thrust generated by the joint is too great and you're imposing unacceptable loads on your equipment. (Ignore squirm and other real world pipe control stuff for now) So, you turn to tie rods to carry this generated pressure thrust and your equipment is now unloaded.

Now, imagine that you increasing the temperature in this pressurized system. The pipe begins to try to expand and compress the joint. To do this, the expansion joint faces must move off of the tie rod faces. At the instant that occurs, the tie rods are no longer supporting the pressure load. That load goes into the piping system, and ultimately to the equipment or anchor/support. So, the pressure loads that earlier couldn't be tolerated reappear. Not only that, you also are going to be getting the loads generated as the joint begins to compress and you see the effects of the spring rate of the joint resisting compression. As C2it mentioned, carefully setting the preloads on the system to try to eliminate these effects is foolhardy at best (and lethal at worst) because even if the installation crew does everything perfectly, it's only a matter of time before the universe notices how you tried to used a finely tuned, perfectly balanced setup in an industrial application and punishes you accordingly.

The only way I can think of to alleviate both of these effects in a straight run of pipe between two anchors or pieces of equipment is the pressure balanced expansion joint. Alternatively, if you have turns in the piping system, it MAY be possible to use expansion joints with tie rods to do what you wish. Instead of putting the joint in the section of the line that is causing the thermal problems, you put the joint on the other side of an elbow in a section of the line that is at 90 degrees to the long portion causing problems. Then, the tie rods will take up the pressure thrust created by the joint, and the expansion of the problem portion of the line will be acting to displace the joint laterally instead of trying to compress it against the pressure thrust. Granted, this may not work and it may not be desirable depending on what your restrictions are, but it's an option.

Good luck!

 

[0707]

Go to for you expantion joint calculation

http://images.google.co.jp/imgres?i...n+bellows&ndsp=20&svnum=10&um=1&hl=pt-PT&sa=N

cheers
Luismarques

 

[JohnBreen]

Hi jistre,

Good posting. I would only add that if, in the configuration that you suggest, the lateral offset at the joint(s) is significant it might be wise to consider a "tied-universal" type joint (two bellows elements with a staight piece of pipe between them with the tie rods spanning both bellows) so that each bellow element will have to accommodate only one offset angle.

Regards, John.

 

[StressGuy]

the bellow is acting in tension due to thrust force, and in compression due to thermal growth.

No, no it isn't. A bellows can only act in compression or extension. You are on your way to a disaster with your line of thinking.

You have two anchor points of some sort with a straight run of pipe between them and you need to compensate for the desire of the pipe to expand some length based on the distance between the anchor points and the temperature of the system.

A simple bellows in this system will have no capability to restrain pressure thrust. You can put tie rods on the joint. However, as soon as the pipe heats up and the joint goes into compression, the tie rods will unload and your two anchor points will be taking the pressure thrust.

What you need to do is first determine what happens to your end points if the pressure thrust load is applied to them. If they can handle the load, then you just need to make sure the pipe is adequately guided near the joint to prevent squirm.

If the anchors can't take the pressure thrust, you either need to reroute the pipe to a "Z" pattern and take the movement across a tied universal. Or, if the pipe routing can't change, you need to look at an in-line pressure balanced bellows design.

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.

 

[StressGuy]

Strange, when I pulled up this thread, Jistre's, 0707's and JohnBreen's posts did not show. Now that I've written a response, they do show and I look like I'm just piling on.

That was not my intent. As is blatantly obvious now, I also agree with jistre's comments.

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.

 

[JohnBreen]

Hi Ed,

Please feel free to pile on any time you wish

Actually, there is sooooo much more to say about this issue that several of us coming from various directions compliment each other.

I should also have mentioned that in the "Z" configuration suggested by jistre, it would be prudent to install gapped guides near the universal joints so that thermal cycling would not cause the long straight run of piping to continuously "walk" toward one of the joints (in one of the perpendicular legs) thereby causing a joint to exceed its maximum angle of deflection over time.

Regards, John.

 

[Dj364]

Thanks to every one, I appreciate it "The Light" very well.
as from your recommendations, and my understanding in line pressure expansion joint will placed approximatly in the middle of the straight pipe run (46 meters, inluding a valve) to absorb axial movement, in such a way the two anchors( static equipement) will be relieved of pressure loading.
however, could you tell me what s the best way to model the in-line pressure balanced bellows?
Dj.

 

[jistre]

If you use AutoPIPE, there is a step-by-step example in the help files.

Essentially, what you do is model the whole joint as what it is in reality: three separate expansion joints sized correctly and connected with tie rods in a special way. The idea behind these things is that the middle section of the joint generates twice the pressure thrust as each of the two outer sections. The tie rods are connected so that as the middle joint tries to expand, it tries to compress each of the outer two joints. Hence, the pressure loads of the outer two joints counteract that of the inner joint, and the only load seen by the equipment is what it takes to compress the joint against its spring rate.

The simplest model of this should be to model three expansion joints in series. Begin at one end and number the nodes across these joints 1, 2, 3, and, 4. Use a rigid member to represent the tie rods and connect node 1 to 3 and node 2 to 4. Put the proper properties in for each of the three component joints and there you have it.

However, I prefer to do the more in-depth model that AutoPIPE outlines. It is fairly complicated with 24 structural members to model the tie rod assemblies, so I modeled it once all by itself in a new model, then saved it. When I need it later, I can copy and paste it into whichever system needs it, change the material and pipe properties, and I'm off and running.

 

[jistre]

I forgot to note that as in all things with modeling software, the simpler you choose to model the system being studied, the greater the chance you're going to get inaccurate results and the greater the chance the universe notices you taking shortcuts and punishes you accordingly.

You have to decide whether or not your situation warrants the simple dirty model I described or the more in-depth model I prefer along with its increased complexity and computing time.

I wanted to make sure I didn't imply that one would be equivalent to the other.

 

[Dj364]

Thanks Jistre and all of you guys for the explanation,
yes I am using autopipe.
How about the tie rods, should I model them with gap forward? ( in gap free to move both ends of the in line pressure expansion joint to allow compression/extention the two balanced bellows).
I ve done other analysis ,66" inlet/exhaust blower but I am not sure about the result, I ve put pressure balanced bellows in order to get lower loads but stiil over NEMA requirements.I would appreciate if you have some guidelines on such analysis.
Thanks in advance.
Dj.