Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations SE2607 on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Pump nozzle load evaluation due to expansion joint thrust load 1

Skybubble

Mechanical
Oct 31, 2024
1
Hello, experts,

I am designing a pump line for a high-pressure water supply system, and the pump has an expansion joint connected directly to its nozzle. The issue is that this pump provides an extremely low allowable nozzle load. The expansion joint is designed with tie-rods to resist thrust; however, due to thermal expansion issues with the piping, the inner nuts on the tie-rods need to be loosened. As a result, the thrust is no longer resisted by the tie-rods and instead enters the connected equipment nozzle — the pump — exceeding the allowable nozzle load.

I have heard from several other engineers that the actual net thrust acting on the pump nozzle should be evaluated as follows:
(Of course, the pump foundation (including anchor bolts) must resist the full thrust force.)

P = P1 - P2
P = Net thrust acting on the nozzle
P1 = Thrust generated due to the bellows' mean diameter
P2 = Thrust generated due to the pump nozzle’s internal diameter

I find this design approach reasonable and quite persuasive. However, I haven’t come across any design codes that support it.

Experts, is this approach correct? If so, is there a design standard available that could justify this approach to help persuade the client?

I would appreciate your opinions.
 
Replies continue below

Recommended for you

Can you draw or sketch this?

I don't understand your set up. The piping needs to be restrained / anchored pretty close to the pump to reduce the required movement of the expansion joint.

tie rods are only there to limit maximum movements or bending.

The nozzle load should only be the external forces acting on the nozzle, normally thermal expansion forces, not the pressure thrust forces as this is dependant on pressure, for which the pump is designed.

P is normally pressure, not force.
 
Force will be equal to :

F=P*Aexj

P=internal pressure at discharge
A
exj=internal area of expansion joint

So if you loosen the tie rods, pump casing must be designed for thrust loads due to EXJ, which may not be acceptable.

You may try to change the type of expansion joint, e.g pressure balance type expansion joints which can absorb piping thermal expansion.

For more information refer to the manual published by EJMA.
 
the inner nuts on the tie-rods need to be loosened
If your outer nuts are in place the tie rods are still resisting pressure thrust.

And no, you can't take credit for the ID of a rigid member attached to an untied EJ.
 
Pressure thrust is determined based on the effective diameter of the expansion joint which takes into account the convolutions. The thrust area is based on this diameter. You cannot subtract the area of the nozzle ID. The pressure acts on the entire projected area formed by the exp joint effective diameter on the pump.

Any axial compression of the joint will cause the nuts to disengage and put the thrust load on the pump, unless the piping is so flexible downstream that the pressure thrust acting in the opposite direction downstream will cause the inner nuts to reingage by pulling the pipe back in the opposite direction.

If you really need to absorb axial movement towards pump I would install the exp joint in a pipe segment at the pump that is perpendicular to the pipe movement so that the movement is taken up by lateral deflection of the expansion joint.
 
I can see what you fellow engineer is saying with the equation you show above. If you take the entire pressure thrust on the pump based on the effective diameter of the expansion joint and then subtract the pressure thrust based on the diameter of the pump nozzle, the result is the external thrust in the reducer section at the pump nozzle. This is really the only thrust on the pump external to the pump on the nozzle. The thrust of the pump nozzle ID is really acting on the pumps internal parts and casing, and does not result in the same moments as when the force is acing at the nozzle. I see some validity in this but I would check with the pump manufacturer how they account for this in the pump design.
 
Last edited:
Pump Nozzle loads are limited to stop external forces acting on the pump base and effecting the motor/pump alignment. The allowable forces limits are not there to limit the stress in the nozzle steel.

As above, if the bellow's tie rods are not engaged then the pressure force does act through the pump body down to the base.
 
OP,
There are two parts to your problem.
Expansion joints are placed in a piping typically to take care of the thermal expansion/contraction and reduce anchor reaction loads.
1. With the removal of the tie-rods, the bellow is actually exposed to high tension load which may eventually rupture if not provided by a reaction force at the anchors (bellows wall thk is about a quarter of the pipe wall thk).
2. The pressure thrust. The pressure thrust can be imagined by two end caps at the two ends of the pipe. In this case, one end is the pump (also an anchor point) and the other a bend (an axial anchor should be provided at the bend). If the piping is of the same diameter at both sides of the bellow, the pressure thrusts will be opposite and balance out. There will be no movement of the pipe. However, anchors will be required to stop the bellows from straightening out.

If pipe diameters are different, you will have a net thrust force.

For corrugated bellows, the pressure force can be calculated by the internal pressure x mean (pitch) diameter of the bellow

Because it's a bellow, anchors must stop the axial movement of the bellow to stop it from rupture. You need to figure out how much will be the total expansion and how much you want to allow for the expansion, then accordingly size the anchors. You will need to account max expansion of the bellow so that the anchor reaction forces are minimum to match the pump flange. You need to play with it.

Be careful, the thermal loads could be larger than the pressure thrust. And that's the reason, you let the tie-rods go.

if the above doesn't work, you might have to go for a longer bellow with more convolutions.




.
 
there are no direct ASME or EN codes that explicitly support this “net thrust” approach in the way you described.
However, ASME B31.3 (Process Piping) and ASME B31.1 (Power Piping) offer guidance on allowable nozzle loads and expansion joint considerations. While these standards do not detail net thrust calculations for cases with tie rods that are not engaged, they provide the basis for allowable stresses and loads that can inform the design.
EJMA (Expansion Joint Manufacturers Association) Standards offer detailed guidance on expansion joint design, including thrust force considerations, and would likely recommend using tied or restrained joints for applications with low allowable nozzle loads.


Consult with the expansion joint manufacturer. Many manufacturers endorse this net thrust calculation under similar conditions, and their endorsement, coupled with proper analysis, can help validate your approach to the client.
 
For an unrestrained expansion joint (which is your case after loosening the tie-rod nuts), the anchor force on each anchor (pump is considered one anchor) will be:
1/2 of the piping (in between the anchors) friction force + total bellows Spring force (total piping expansion x spring rate) + total bellows pressure thrust (internal pressure x mean (pitch) diameter of the bellow).

From above, you can reduce the anchor reaction force by the following options (process conditions - pressure, temperature, pipe size are fixed):
  • reduced spring rate
  • reduced bellow mean diameter
Out of the above two, the only option you have will be to select a different bellow with reduced spring rate. the bellow mean diameter being fixed corresponding to the pipe size.
 
Last edited:
Attached is a paper by Peng which discusses rotating equipment nozzle loads and expansion joint installations at equipment.
 

Attachments

  • Eqp-Reli.pdf
    319.9 KB · Views: 15
Here is another paper by Peng discussing loads on pump nozzles.
 

Attachments

  • Rotating-Eqp.pdf
    295.2 KB · Views: 10
Last edited:
As indicated in Peng's paper, it may be troublesome to have expansion joints in pipelines. This is especially true for high-pressure lines.

A similar application is boiler feed pump suction and discharge lines in power plants. No expansion joints are used here. Piping loops are provided for flexibility such that allowable nozzle loads are not exceeded. This may require extensive modifications in the piping layout.

Another option is to use anchors or restraints near pump discharges. If feasible, stress analysis needs to be done for this condition. Pump thermal movements are also to be adequately taken care of.
 

Part and Inventory Search

Sponsor