What is the determining factor on whether or not an expansion joint is needed 4

[BronYrAur]

I have a long length of chilled water pipe. 2-1/2" copper pipe drops down through a floor and runs in a tunnel for 300' with no elbows. Then the pipe turns 90 degrees and runs an additional 20' The pipe then turns back up through the floor. Nothing is rigidly locked above the floor on either end, so there is "wiggle room"

It is chilled water pipe that will be operating at 40 degrees F. Absolute worse case ambient during installation will be 100 deg F. I can calculate the the expansion easy enough; it is approximately 2" along the 300' run. My question is, how do I know when expansion compensation is necessary? How do I justify when to install expansion joints ---- and when not to. I need to be able to make the case either way.

 

[pennpiper]

Always consider Expansion Joints as the last (and the most risky) choice. They are more prone to failure than pipe and fittings.
- First try routing (with anchors and guides) for flexibility and clearance for expansion movement.
- Second try routing with Loops, anchors and guides to control expansion movement.
- Last choice accept the need for Expansion Joints.

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

 

[BronYrAur]

Thank you for the comment. Can you offer any recommendations about my specific example? Do you think any expansion compensation is needed? And if so what would be the justification one way or the other?

 

[BigInch]

If the thermal expansion over-stresses the pipe and you cannot add any further flexibility to the pipe configuration to reduce stresses, consider an expansion joint.

Your pipe is small diameter. The run is long, but temperatures are not excessive and you have 20' on the end in which to do some flexibility bending. The moment developed there should not be all that great. I doubt if you need an expansion joint. You could try using an anchor to force growth to the opposite direction, away from your floor penetration. Maybe try two anchors, one at each end of the 300' run and S-snake the pipe a little in the middle. Expn joints are generally reserved for placement nearby sensitive equipment where piping is extremely tight, with no room for any expansion loops or doglegs and expansion directional control by use of anchors was found to be ineffective.

 

[LittleInch]

for more complex systems you would do a stress analysis to see if the stresses in the pipe from pressure, bending and torsion exceed the allowable pressure.

Without that it is difficult to exactly determine if you need an expansion joint or not, but from your description ( an isometric sketch would be useful) it looks like you have enough flexibility. The key is working out where it is in fact finally fixed ( it must be fixed to something somewhere) and what the likely forces are at that fixed point.

As this pipe will be in tension you don't really have too many issue with the pipe buckling or falling off simple supports, but you will need guides somewhere to limit movment.

I echo all my fellow posters - use expansion joint s only if nothing else works, e.g. a straight line between two fixed points.

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

 

[Lore80]

Hi!

in my opinion the use of expansion joint can be useful when:

- you have high temperature and a piping routing with limited space..so you cannot introduce flexibility by working on the piping routing
- you have very low allowables at connection to be respected

But the starting point is that first of all I Always try to solve the problems by working on piping routing and supports....the last choice is the use of expansion joint and hangers

Lorenzo

 

[JohnBreen]

As you describe it, your system will be installed at “ambient temperature” and the rise in stress will be caused by the reduction in temperature as the pipe material CONTRACTS when the system is placed into service. In concurrence with most of what has been said above (thanks pennpiper) the first step is to determine if you will have enough “flexibility offset” in the system as you describe it (as it is designed). The approximations described in so many books: (Grinnell Piping Design and Engineering, Pipe Stress Engineering (Peng and Peng), Process Piping (Dr. Charles Becht) , et al provide a sufficient approximate flexibility (stress) analysis. Of course, today this would be a trivial exercise using available analysis software. If this “approximate” evaluation approach indicates that additional compensation for the in-service contraction of the piping is needed to assure appropriate levels of stress, you would want to provide a “contraction loop” (with required anchors and guides) in an appropriate location in the system and indicate on the drawings the locations for the required “anchors” and “guides” needed to assure appropriate levels of operating stresses. Infrequently, it may be necessary to include expansion joints (e.g. there is not enough room to install an “expansion / contraction” compensating loop and associated anchors and guides) and in these cases it would be prudent to consider ALL the options that these components provide. There are devices that provide axial compensation using bellows or “sliding” options and there are devices that can provide “offset” compensation using “ball-joint” or similar options. This of course would be a “last resort” as all these devices have a finite operating life expectancy and they require periodic inspection to assure that they are still serviceable.

 

[Zurck]

Expansion is not your problem in your system. Contraction is....try to provide a little flexibility on the route of your copper line by adding few bends and turn to absorb the movement in the line.

 

[BronYrAur]

Thanks all for your replies but I'm really wondering if the system needs compensation as I have described it. In other words no offsets along the long 300' run to absorb the contraction. How do I know if the system I have laid out in words above will need something. What's the threshold that determines whether or not compensation is needed. I can calculate the amount of contraction and the angle of deflection it will put on my short 20 foot long pipe assuming the vertical pipes through the floor are rigidly locked - which they aren't. That would be a worst-case, but I just don't have a good feel for what amount of deflection would be a cause for concern

 

[pennpiper]

Why don't you know the amount of contraction in Copper pipe for your maximum low temperature?

Per your Post:
Installation Temp. = 100 degrees
Max Low Operating Temp.= 40 degrees
Delta = 60 degrees (but temperature wise in the opposite direction)

So, use the coefficient of expansion for Copper and calculate it as a negative (Shrinkage)and that should be close enough.

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

 

[SandCounter]

You need to translate your thermal contraction into tensile force. What force would be required to stretch 300' of pipe by 2"? That force you would then apply to the 20' run as if it is a 20' cantilevered beam fixed at the floor penetration. If the deflection of the 20' pipe is within allowable safety for yield, the 20' run is ok. If you do not get the full 2" of deflection from the 20' section, back calculate the force from the difference on the 300' run and check tension in the 300' run.

My concern would be at the elbow, if threaded, is not designed to take much transverse torsion.

I used to count sand. Now I don't count at all.