Hello everyone,

Sorry I'm new to stress analysis. My problem with this pump is that it's kinda impossible to lower the seismic loads on pump nozzles with considering gaps on the supports. I Know for occasional loads, guides with gap and without friction are the worst case. but please let me know if I've modeled it wrong or if I'm missing something?

Thank you

Charlot,

More detail please,

First of all, all the input data has been input correctly??

Your pump nozzle is 230? or 1160? or else?
Which case cause problem?
Do you have allowable nozzle load? Or any specific code/standard for pump? Habe the pump nozzle load to be below the allowable load?
Can the allowable load be increased during occasional?
Are you able to add/delete support? or change the pipe route?
Engineered support such as spring support or snubber etc. can be used?

If there aren't proper x, z direction restraint(without gap), you can't prevent seismic load acting on the pump nozzle. But when you add guide/stopper at the support, if the nozzle load exceed by thermal reaction. you need to change the route or include small gap or use other support which does not limit the thermal displacement.

There are too much consideration, for given limited information.

By the way, your way of inputting the data shall be checked too. you cannot weld elbow to straight line directly, and have you check the minimum tee length?

Are you sure you have 1.49g? seems very high.

CAESAR

The Input data is correct as far as I know.
Pump Nozzles are both 1160 and 230 and I'm using API allowable(Using extrapolation for 18" 230 nozzle). and I've tried to keep the ope(W+P+T) load Under the allowable. I can delete or add supports wherever applicable but not hangers and snubbers preferably. and no pipe rout change.
the guides are only for seismic loads and without them thermal loads are fine.

My question is that how can I lower the occ loads with considering gaps on the guides which seems impossible in this case. Is there something wrong with the way I've modeled it or can I omit the gaps and if I do that, wouldn't it be unrealistic?

Dan, yes it's kinda conservative but this acceleration is what I have to use.

Thanks a lot

Charlot,

If you omit the gap, wouldn't the nozzle load exceed the limit?

You may adjust gap to meet the API allowable, but is it realistic?

As Mr. Dan pointed out, seismic factor is very high(I never seen), and if you don't want to use snubber it seems not possible to meet the allowble.

API does not give you allowable loads for occasional, so how about contact the vendor for confirmation?

Civil engineering structures are designed on the basis to two main criteria- strength and serviceability.
The strength criterion is related to damageability or ultimate limit state, usually assuming some plastic (limited) deformation. Ductility is an essential attribute of the building and serves as shock absorber for it reduces the transmitted force.
The serviceability (or rigidity) assures the structural displacements remains in some limits.

Now transposing the building structures concepts and applying strength and serviceability spectra for piping calculation rise up some difficulties, B31Ea provides some guidance.

Your conservative value seems to be one of peak on strength spectrum, corresponding probably to a 0.4g or 0.5g peak ground acceleration. If true, it means this value assumes that all vibration modes have frequencies that on spectrum correspond to maximum possible pseudo-acceleration and no credit is given to ductility and possibility to develop plastic deformations. If my guess is true, it remains to clarify yourself what you expect to have in your piping system and pump at this seismic requirement.

CAESAR, Mariog,

I meant without any guides ope loads are under allowable and the reason I'm adding lateral supports is seismic loads.

The actual PGA is 0.9g which is high too. even with that acceleration, seismic loads on nozzles exceed the allowable.

so my question is which one is more realistic or usual when checking the allowable for occ loads on pump nozzle? considering lateral supports with gap or without gap, because it makes a big difference.

Thanks everyone

In my opinion, at PGA=0.9g you may consider any gap you want because it shall be just a math software figure intended to solve an unrealistic input.

Simply said, if true as an event once from early Homo sapiens sapiens to today, at this level of PGA your body will receive a horizontal force which equals almost your weight, so will be impossible to stand up for note what happened with your pump and after that, being a happy survivor, you would try to recover the pump from debris to see more...

:)) I see. Thanks

Charlot,
the first thing I would be doing is questioning the pump vendor on the allowable nozzle loads under a seismic loading case. These should not be the same as the values for an operating condition. The API 610 allowables are based on the longevity of a pump operating 24/7, and are not appropriate for a short term event such as an earthquake.

My view is that if a design seismic event occurs, the pump should have the strength to withstand applied nozzle loads without causing failure or leakage. However it may be necessary, for example, to realign the pump after the event. The actual requirements for the pump may depend on the service. If the pump was in critical service and expected to keep operating throughout and after a seismic event then nozzle loadings much be kept low enough to allow this. However if the plant is likely to trip during a seismic event, then this would allow higher loads to be experienced by the pump.

In our experience, for large diameter nozzles it is generally difficult to get compliance with nozzle loading requirements when using friction type guides. It is common to use either low friction guides or more often to use tied links (pinned tie-rods) to avoid friction loads. The stiffness of the supporting structures becomes very important.

Hope this helps.

Thanks a lot Ross. Actually we don't have a vendor yet, but as you said based on the operation condition I decided to increase the allowable loads for the occ case.

The other issue that is really important is to get the requirements for allowable nozzle loadings under seismic action into the technical specification at the bid or tender stage, and make sure they are agreed before placing the order for the pump(s).

If the finalisation of these details is left until after order placement, it becomes much more difficult to get a satisfactory result for the stress analyst. The pump vendor is not incentivised to carry out the engineering investigation required to determine the appropriate loads.