Rubber Bellows vs Metal Bellows Expansion Joints 4

[simon17]

Good morning.

I haven't had much luck recently in finding information on when to use rubber expansion joints and when to use metal. My expansion joint experience is limited. If either can be used, what information is critical to choose which to use? How do they differ?

Thanks,
D

 

[btrueblood]

Rubber is more flexible, cheaper. But has limitations on chemical compatibility, service temperature (both high and low).

Metal bellows are more expensive to produce, and have more limitations on the amount of flex allowed, cycles to failure (fatigue life).

 

[pennpiper]

dso88,
Do you have a specific purpose or use for an expansion joint?

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

 

[simon17]

To relieve nozzle loads due to thermal expansion of my piping system. I'm doing stress analysis. The system is low pressure and fairly low temperature (<200F) with brine service.

Thank you btrueblood for the response. This was the type of information I was looking for.

 

[XL83NL]

What line size and configuration/routing do you have? Thats not a fairly high temp.

 

[simon17]

Line size is 30", 15psig & 160F. Line routing travels down out of a flat bottom tank and bends 90deg to pump suction. Line size reduces to 18" prior to pump nozzle. There is plenty of space for an expansion joint in either the horizontal or vertical portion of the line. It just depends which nozzle I choose to relieve using the expansion joint. The flat roof/bottom tank nozzles are especially sensitive to loads, but so are pump nozzles.

 

[XL83NL]

Can you solve the problem with an expansion loop? Expansion joints are more prone to failure and sensitive to damage, and should be used as a last resort.

 

[pennpiper]

dso88,
What is your plan for when your expansion joint springs a leak?

Both rubber and metallic expansion joints can and will fail. The best insurance against a catastrophic expansion joint leak is don't use expansion joints.

OH! you say but I need to reduce the loading on the nozzles. Expansion joints do not reduce nozzle loading. In order to work an expansion joint needs an anchor on each side of the joint & pipe run. The nozzles just become the anchors, so they are taking the loading and force of any expansion.

Whats the answer? Place a fixed anchor a minimum distance from the Pump Nozzle (at or near the suction reducer). Place another anchor a minimum distance from the elbow for the vertical drop from the tank nozzle. Allow vertical growth down but no horizontal movement. Place the expansion joint any place along the 30" pipe run. The long horizontal leg of 30" pipe with the expansion joint needs to be guided so the expansion joint does not 'snake' on you. Place one guide a minimum distance upstream and down stream of the expansion joint (2 guides required).

You did not define any lengths for the two legs of the pipe run so we cannot tell you where to place the expansion joint (or Loop). Place it in a location that is easily accessible for visual inspection and maintenance.


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

 

[StressGuy]

Of course expansion joints reduce loads on nozzles - provided they are properly designed along with the piping system. You don't have a lot of thermal expansion given that temperature, but the large line size is going to make for a stiff system.

You may get some relief considering the local flexibility of the tank nozzle in your piping system, but probably not enough to qualify. Based on the fluid conditions you are describing, a non-metallic joint will probably be suitable for your application. The joint will most likely need to have control rods to restrain pressure thrust - meaning that it will not give you any flexibility in the axial direction of the joint, only in some lateral bending. A pair of joints acting as a "universal" would be needed if you had a large amount of lateral deflection to take across the joints, but I don't expect that will be your case.

Based on your description, I'm assuming that both the vertical leg out of the tank and the horizontal leg to the pump are fairly short, but that the vertical leg is much shorter than the horizontal. In such a case, I would put the joint in the vertical leg so that the growth of the horizontal leg deflects the joint and filtering it's impact on the tank nozzle. This joint won't directly help your flexibility with regard to the pump nozzle, however, you would discount any thermal growth across the joint itself as the control rods (also known as tie-rods, particularly in metal bellows speak) force any expansion across the joint into compressing the bellows slightly. That reduction in expansion in the vertical will reduce the loads imparted on the pump nozzle.



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.

 

[simon17]

Edward,

Thank you for your reply. Very well written and thought out. I very much appreciate it. It is pretty much along my line of thinking as well. I haven't been able to get back to that system and run the analysis to see whether or not one expansion joint will be enough to rectify the nozzle loads.

Your assumptions on the routing (vertical length < horizontal length) is correct, but both are still substantially long enough to place tied single OR universal joints. You recommended placing the joint in the vertical leg. Maybe it was a typo, but you stated that this would NOT reduce the pump nozzle loads. Certainly, the vertical joint will decrease some of the bending moments and shear loads on the vertical tank nozzle, but under the same principles, it should also reduce some of the axial loads on the pump nozzle.

I'm hoping to be able to get away with just one, but I'm fearing with this size and stiffness of pipe, I may need tied single expansion joints on both the vertical AND horizontal runs to relieve both nozzles.

D

 

[ZDR1985]

I would probably put the rubber bellow (restrained with tie rods up to max expansion) on the horizontal run on the pump side to protect the pipes from pump vibration. Your piping will be anchored by supports and your tank is static so I don't see any reason for additional rubber bellow on the vertical run or any other location away from pump. Just my opinion.

 

[bellowsmfg]

Edward hit on a lot of good points and I'd like to add to them.

First of all, you need to be aware that a tank has some different characteristics then we normally see in a vessel. The nozzle at the bottom of tank will rotate downwards. This is due the shell deflecting out when the tank is full. For this reason it is not a good idea to anchor such a nozzle per pennpipers comments.

If can can avoid using an expansion joint, then great. If you have tried all other options then here is your approach: Your expansion joint selection should be done in conjunction with your pipe support design to resolve the forces on both your pump and tank nozzle. These go hand in hand. You will want to obtain the properties (axial, lateral, rotation and torsional stiffness's) of the joint you intent on using, or close too. This will allow you to determine the actual loads you will see on the pump nozzle. Your supports will protect the expansion joint from undesired deflections. Locating near the pump suction nozzle makes it easier the resolve nozzle loading as you are not dealing with friction from supports prior to the nozzle (if the joint was located somewhere else).

Your pressure is low so you won't have too much of an issue with pressure thrust but you must understand it, both now and moving forward.

I would not recommend using an expansion joint simply because of vibration unless it is the result of some previous concern or modal analysis. The pump should be on a solid foundation, base plate/frame and well balanced. The piping should be properly aligned prior to final bolt up, etc.

Expansion loops are rarely desirable in a pump suction lines. You have some straight lengths as you have mentioned. Carefully go through your design, minimize where you can see if you can solve this with one expansion joint. If your flow rates are low enough remember an 18" expansion joint will be cheaper that a 30" one. As I mentioned, make the supports work with your expansion joint to solve the problem after you have exhausted all other possibilities.

Conor Walshe

http://www.bellowsmfg.com

Bellows Manufacturing and Research, Inc.

https://bellowsmfg.com

 

[11echo]

I've had some experience here, I worked for a major oil company that wanted us to use a "well known" companies rubber expansion joints at pump suctions to reduce stress on the pump nozzles. In theory it sounded good but these rubber joints would only last 6 to 9 months before there was some kind of issue. We finally ended up using braided stainless steel hose with flanged ends, never had an issue after that.