Expansion joint application in piping 1

[farhadsh]

Hi,
I have a question about EJ used on piping and appreciate any feedback on it.
My piping is connected to lube oil filter from one end and to heat exchanger from other side. It is elbow and loops on it to add flexibility. It is water cooled heat exchanger and plate type. for simplicity assume a very simple system. Allowable nozzle loads on H.E is very low. Installation temperature and max oil temperature are 20 and 60C respectively. I have constraints and can not add any pipe support to piping between two equipment. I can not modify pipe model and add extra flexibility or loop if needed.

While modeling piping in Caesar it meets the stress code but nozzle loads on H.E goes beyond allowable loads. when load case is sustained load plus thermal expansion load the H.E loads will become 2x or 3x of allowable. So I decided to use EJ with tie-rod.

I assumed piping first see pressure then thermal expansion. when piping is under pressure EJ will try to expand due to thrust while tie-rod prevent from expansion. so in this case there is no extra force on H.E nozzle.

when piping temperature increases, piping thermal displacement pushes the EJ to compress but since in my case thrust is greater than thermal force it won't let the EJ to compress and so thermal displacement will add extra force to H.E nozzle. the extra force is equal to thermal displacement force.

It appears using EJ is not going to help piping system or to lower the nozzle loads on equipment

Am I right about this?
Is this valid for EJ when piping sees thermal expansion?
Is the solution to put EJ with tie-rod perpendicular to line?
Thank you,

 

[LittleInch]

This needs a diagram / iso or sketch to understand what's going on here.

EJs need to be guided so that they can operate correctly.

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

 

[farhadsh]

Hi LittleInch
Let me see if I can add sketch.
going back to the concept, do you agree that EJ won't compress unless pipe force from thermal expansion exceeds the thrust loads developed inside EJ?

 

[LittleInch]

I recommend you look up this lot and get a copy of their standard.

https://www.ejma.org/

The trick is to avoid the pressure thrust....

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

 

[Snickster]

Looking at Axial forces at HX only:

For the case without the expansion joint the pressure thrust on the first downstream elbow is balanced by tension in the pipe back to the heat exchanger nozzle. This assumes that there is a straight piece of pipe connected to the HX nozzle and downstream is at least one elbow and offset pipe leg (or more) which provides flexible piping between the HX and filter (like an L-shaped pipe arrangement for example). This is because the axial stiffness of the straight pipe to the HX is far greater than the stiffness of the flexible piping downstream of the first elbow so pressure thust would necessarily be taken up by tension in straight pipe back to HX nozzle. In this case all of the thermal expansion of the straignt run of pipe to the HX will cause deflection of the downstream flexible pipe downstream of the elbow and produce a thermal expansion force on the HX nozzle equal to change in length of straight run times the stiffness of the downstream piping K(dX thermal). The net force on the nozzle would therefore be sum of the Pressure Thrust directed away from nozzle and thermal defection force directed into the nozzle.

With expansion joint the maximum force on the HX nozzle will always be the net compression of the expansion joint times the axial stiffness K factor of the expansion joint. The pressure thust will be acting on the downstream elbow as before so before heat up to temperature the pressure thust will again be taken by tension in the straight pipe and transfered through the tie-rods back to the HX nozzle. As the pipe heats up and expands it will want to compress the expansion joint but if it does then the only thing resisting the pressure thrust now is the downstream flexible piping. If the downstream piping is very stiff then it will be able to resist the pressure thrust not allowing the pressure thrust to move the downstream piping much so all or most of the thermal expansion is pushed into the expansion joint causing compression. In this case the pressure thrust is taken by the downstream piping causing some deflection of the downstream piping so that the net deflection of the expansion joint is equal the the overall thermal expansion minus the deflection of the downstream piping due to the pressure thurst.

If the piping downstream is very flexible then the deflection of the downstream piping due to the pressure thrust could equal to or greater than the thermal expansion of the pipe. In this case the gaps on the tie-rods will close and make contact with stops. In this case the pressure thust will be shared by the flexural resistance force of the downstream piping and the contact force of the tie-rods nuts on the expansion joint stops. Any pressure thust not taken resisted by the downstream pipe will then be transfered to the HX nozzle but no axial thermal forces would exist.

So as can be seen there is a differnce between having or not having an expansion joint.

Likewise with expansion joint any lateral thermal expansion of the downstream piping would be taken by the lateral deflection of the expansion joint. The lateral deflection force would be based on the lateral stiffness of the joint and the effects of the tie-rods in increasing the lateral stiffness of the joint.

 

[FacEngrPE]

Caeser is capable of modeling several kinds of Bellows Expansion Joints.
Examples for these
[ul]
[li]Universal Expansion Joints - Simple Model[/li]
[li]Tied Bellows Expansion Joint - Simple Model[/li]
[li]Universal Joint with Lateral Controls Stops - Comprehensive Tie Rod Model[/li]
[li]Hinged Joint[/li]
[li]Slotted Hinge Joint - Simple Model[/li]
[li]Slotted Hinge Joint - Comprehensive Model[/li]
[li]Tied Bellows - Simple vs. Complex Model[/li]
[li]Slip Joint[/li]
[li]Gimbal Joints[/li]
[li]Tied Bellows Expansion Joint - Complex Model[/li]
[li]Dual Gimbal[/li]
[li]Simple Bellows with Pressure Thrust[/li]
[li]Pressure-Balanced Tees and Elbows[/li]
[li]Universal Joint - Comprehensive Tie Rod[/li]
[/ul]

There is also an example for accounting for the equipment & flange flexibility.

Either your joint must be pressure balanced or restrained, or the pressure force will be transmitted to the HX flanges, and available anchors.

 

[StressGuy]

You are correct in your thinking that, assuming your joint is inline between your exchanger and filter, then in order to compress to relieve the thermal load, the pressure thrust load will be removed from the tie-rods and shifted to the pieces of equipment as they become the main anchors.

For a joint to be effective, you really need to put it in bending. That's going to require at least one offset leg.

The other option is to use an in-line pressure balanced joint. They are pricey, but if you're stuck with a routing that you can't change, that is a way to get it to work.

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.