as mentioned in api 676, the allowable load for pump nozzle is
fx = 13D
fy = 13D
fz = 13D
mx = 7D
mY = 7D
mZ = 7D
these allowable is very low and hard to pass as to API 610 which can multiply 2 to check nozzle.
is there any other appendix like API 610 to check these nozzle loads.

Hi miangun
I think there is no provision in API 676 for increase allowable like API 610. You still can talk to vendor for more allowable.
Gear pump allowable is less than API 610.
Gear pump I have worked with is generally one pump connected to one tank, not like 3 or more connected in parallel & with operating /stand by situation which generally create more problem in nozzle load for centrifugal pumps.
Check your situation, if only one pump in each stress system I think still it is manageable.

Regards
Habib

Hi shr,

Thanks very much!
In offshore platform, spring is not recommended. In our project, the elbow connected to suction and discharge is supported with dummy leg.
I refer to the previous practice, the temp is defined as installing temp. But I think the dummy leg is not insulated, it is exposed to environment, and the ambient temp is vary from time to time(max 38, mini -18). The pump nozzle check will not pass when max or mini temp is used. The reason is that the dummy leg will expanion or contract when high temp difference occur and the length of straight pipe between nozzle and dummy leg is very short(about 300mm), it cause large nozzle load in Y direction,
my question is when model the dummy leg how to define the temp? Is installing temp right, or should i consider both max and mini temp.

many thanks!

Hi miangun

If you not like to ignore dummey temperature difference because of ambient temperature change nor like to use spring then you can have following option to pass nozzle load
1) Increase length between nozzle & elbow.
2) Use dummey in horizontal position instead of vertical at elbow.

Regards

Habib

In case there is low allowable nozzle/flange loads on the pumps the best thing to reduce the effect is the introduction of double guide ( moment restraint) on suction/discharge piping of the pump at reasonable distance to the pump flanges. In addition to that you need to introduce L shape to the suction/discharge piping beyon these guides to control the axial forces and torsion.

Of course the pump model in the piping stress analysis has to be reflecting the actual pump support(s) with proper thermal expansion if available. This will help to the accuracy of the loads on the pump flanges against thermal loads.

If there is an uplift due to the thermal loads on the pump body (flanges), you need to select the distance of the vertical rigid supports to the pump flanges carefully to raduce the hot gravity loads to minimum.

This approach can be used in any pump application to reduce loads on flanges.

Hope this helps.

Ibrahim Demir

Is it proper to set dummey temperature as installing temp to ignore dummey temperature difference?

thanks!

Hi Ibrahim
Can you please clarify what is double guide ( moment restraint), you are suggesting. Is it two guide support nearby or two directional restraint ?


Hi miangun

1) If you already have existing procedure of ignoring dummy ambient temperature variation why you do not follow the same.
2) If dummy temperature variation consideration really make trouble, you can ask insulation for dummy also & close the issue.


Regards

Habib

shr,

Yes, it is two guide supports in horizontal. You can adjust the distance between guides by watching the allowable guide forces.

I would start with the support arrangement in my previous post and see the loading on the pump flanges and at guides. If the loads are smaller than allowables I would start eliminating supports one by one by going through the checking process of the loads.

Ibrahim Demir

I think I need to clarify that I considered the pumps with the suction and discharge in horizontal in my previous posts.

In case the suction and discharge in vertical the you may need different style of support arrangement. This case you need to make the piping arrangement as flexible as possible, this will eliminate the excessive loads on flanges, however you still need to comprimise with seismic and wind loads if they are available in addition to thermal loads. Probably, you may need to use spring hangers to take thermal loads. If the wind and seismis loads are not the issue and there is no dynamic loads in the system you may probable use vertical rigid/ rod hangers instead.

Kind regards,

Ibrahim Demir

Hi Ibrahim
Thank you so much for your clarification.
When pump nozzle is in horizontal position, I put (rest + guide)support at trunion to first elbow (horizontal)closer to pump. Sometime this arrangement offer very good result in terms of nozzle load.

I agree with your approach of problem solving for different kind of pump nozzle arrangement & I also follow same style.

Regarding two guide support, I understand it will be beneficiary for reducing moment but I am worried about nozzle & support force.

I would appreciate if you can send me any sketch / file you use with two guide. I have not used those kind of arrangement.
My mail Id --- habibur21@gmail.com

Regards

Habib

I would highly discourage trying to create a moment restraint. Anything can work in the computer, but I an doubtful that a pair of guides are going to give enough real world control. Typical guide details are going to have too much gap and the relative stiffnesses of the steel structures are likely to be of limited value. Trying to put a guide close to a pump is also going to be of little value in the real world. With most details, there is a small gap in the installation to make sure the pipe is free to slide through without binding. For the small displacements involved close to a pump nozzle, by the time your "guide" has become engaged, the pump will already be loaded up.

Now, getting back to the original poster - you are making a problem in Caesar that isn't going to mean anything in the real world. You are imposing thermal loads on the pump due to a temperature change in the support leg from changes in ambient temperature. Did you include this ambient effect on the pump case? on the other support steel? Probably not. Changes in ambient temperature are going to impact everything.

Try this also - take the length of your dummy leg and determine how much it is expanding/contracting at the various ambient temperatures. I'd bet you find something on the order of .01" or less. You are probably looking at an expansion/contraction that is going to be less than the installation tolerance, even with a top flight millwright doing the alignment.

Always remember that Caesar is just a giant calculator. It is a simplified representation of what you will have in the field. It is going to tend to be stiffer than the actual piping system and will tend to greatly amplify loads when you have a closely coupled system such as yours. There's a spring rate at the dummy leg/pipe interface that you haven't accounted for, the bolted flange connection isn't perfectly rigid, the steel that the dummy leg sits on isn't perfectly rigid, either. All the world is a spring.

In your case, I would ignore temperature changes in the dummy leg as it is going to lead you to search for solutions to problems that don't exist except in the computer.

Shr,

There is no magic or special support drawing for the support I mentioned. They are the ordinary support you use in every piping. The guide support could be a structural member either side in contact with pipe if the loads are acceptable, or shoes in horizontal in case the loads are higher.

My approach above is for the pump flanges/nozzles with very low allawable loads. It is not a must for all pump appliaction with reasonable allowable loads.

In case there is a differantial vertical thermal motion between pump flange centreline and vertically supported pipe centreline there always will be additional shear load and bending moment in addition to gravity actions on the nozzle/flange of pump. So reducing this additional action is your job by selecting the right location to vertical support.

I trust this clarifies all.

Kind regards,

Ibrahim Demir