Hi,

Dave, I saw a post which you gave the following explanation:

"However, I don't think John directly addressed your query -- "Do I need to worry about the pressure thrust load on the pump nozzle which exceeds the load allowed by API 610?" I see only minor pressure thrust loads on the nozzle.

The way I handle pressure thrust loads is to get inside the pipe (figuratively speaking) and look up and down the line. The surfaces I see are the surfaces loaded with pressure and therefore, the location of the pressure loads. In the joint, looking away from the pump I see the inside surface of the elbow. The pressure thrust here is taken directly by your anchor. Turning around and looking the other way, I see the pump internals. These pressure loads go right into the pump shaft, casing and then supports. So far, no pressure load on the pump nozzle. That is, except for the exposed area of the expansion joint convolutions. The difference between the large ID of the bellows and the small ID of the bellows will provide an area upon which the pessure will develop a compressive load on the nozzle. It's a little more complicated than that if you want to consider all the stiffnesses involved -- the joint stiffness, support stiffness and pipe stiffness but I don't think the numbers will change much."

However I read in the book "Pipe Stress Engineering" of Peng that this kind of instalation is wrong and might crippled the operation of the plant (see annex). I agree and understand that there is only a minor pressure thrust loads on the nozzle, but the external forces and moments provided by API rules are to prevent basically shaft misalignment. Since there is a weak link beetween the pump body and the pedestal/base plate the pressure thrust acting on the pump internals might produce some misalignment.

I would like to hear your opinion about that.

Thank you in advanced.

My opinion is ...it's better to quote also the question of that old post.

"I am working on a cooling water pump with a 14" diameter discharge nozzle that has a rubber expansion joint connected to it. The expansion joint does not have any control rods to absorb the pressure thrust. I can anchor the piping downstream of the joint to hold the pressure thrust. The pump will act as an anchor for the other side of the expansion joint. My question is: Do I need to worry about the pressure thrust load on the pump nozzle which exceeds the load allowed by API 610?".

I guess Mr.Peng rule refers to other case: don't lose the rods advantage. Your opinion?

In this case doesn´t matter if there is rods or not since the bellows will contract (thermal effect) and the rods will not act (see figure in the annex).

In "EJ no rods" case or in case the rods are not working, the EJ "breaks" the piping system in two parts and the pressure thrust is acting on both of them. It is supposed that you’ll analyze the consequences. Mr. Peng made a suggestion on the piping part.

And yes, a part of "pressure thrust" is acting on the pump internals. I would consider the Manufacturer is well aware of this fact.

I totally agree with you.

I think that API610 and it's allowable's were based mainly in this weak link between the pump body and the pedestal/baseplate. The pump body is designed to be as strong, if not stronger, than the piping system so this part (pump case) is not the problem. Of couse the pump body will resist the trust load, what about the link?!

The doubt still remain. Which portion of pressure trhust load do I have do consider (adding the others loads - thermal, etc..) to compare with the API610 allowable? According do Mr. Dave Diehl only the exposed area of the expansion joint convolutions and the others stiffnesses involved, exactly what i have been doing before i saw the Peng's book; which left me in doubt.

Mr. Peng offers a detail that I did not consider in that earlier post.

Dear Rliberato,

To explain the problem in a very simple way: when you have an untied bellow in front of a pump, minimum load should go to the pump nozzle and maximum to the pump foundation ( this is where the danger lies).This is explainable from the simple law of structural mechanics which says that stiffer members take higher loads.Just draw a free body digram and see why.

A tied bellow or a rigid pipe will exert higher loads on the pump nozzle than when untied bellow is there is front of the same ( many stress engineers I have come across say the opposite)and hence lesser on the foundation.To the question as to whether this pressure thrust load be considered on the nozzle ( again many stress engineers use the PXA force on the nozzle in case there is an " untied bellow" in front of the pump), , ask yourself this question as to why this issue comes up when there is a bellow, specially an untied one? As I have mentioned that for tied bellow, the effect is like that of having a rigid pipe in front of the pump.So if you do not have a bellow i.e if you have rigid pipe in front of the pump,do you add the PXA on the nozzle? I don't do that and I don't know anybody else who does.Of course, remember that using an untied bellow the biggest risk to the pump is the huge loading that is going to the pump foundation.

My understanding is , based on discussions with many pump engineers , pressure loading is taken into consideration in the design of the pump casing and internals, hence providing PXA on the nozzle will be like doubling the effect. Of course I mean the case with tied bellow or rigid pipe.With tied bellow also, if the nuts are not properly tightened you have risk of higher load going to pump foundation , in a way like that of the untied one.

Regards

Hi All

Currently I am working for a hydrocarbon plant as site engineer after finishing detail engineering. I found our engineering partner use untied rubber bellow for both discharge & suction side of big pump handling cooling water from cooling tower. I was surely expecting something wrong for discharge side of pump ( 9 bar pressure) & warn them not to use untied rubber bellow. However pump is running well.
One thing they done correctly they use anchor support after bellow.
I observe suction side bellow look fine but discharge side bellow convolution become straight & stiff because of pressure thrust as expected.
I was expecting pump vibration or unusual behavior but that don't appear.
I am remember sometime back Mr Richard share some experience of pump vibration due to use of untied bellow.
May be in long term pump will behave wrongly in our plant or may be pump foundation is rigid enough to hold pressure thrust inside casing.

Regards

Habib

Thank you very much everybody for the comments.

I will try to get in touch with some pump manufactor to see what they say about this matter.

If one is modelling an expansion joint but does not specify an effective ID for it, CAESAR issues a warning, not an error message. The warning states that non-conservative results may be produced. This means to me that there are instances where it may be correct to model an expansion joint w/o accounting for pressure thrust. Anyone know what these instances are? When the expansion joint has tie rods to contain the thrust?

Yes.
There are several approaches modeling XJs. If tie rods contain pressure thrust, then, you need not model that thrust. To not model thrust, don't specify the effective ID of the XJ.
We throw a warning because the user may not intend to exclude this thrust load.

Thanks Dave.

If the thermal load compresses the joint even a small amount I would say that the thrust force would have to be considered since the tie rods would not be taking load. In the case where temp is low then I could see not considering the thrust force but in this case you might not even need an XJ.

You can not ignore pressure thrust.

Is your thermal expansion real or is only due to a value on the line list?
Did you consider nozzle flexibilities?

If your temperature is ambient, then better have a run without temperature and use tie-rods. (I know this trick from John, thank's again)

Regards,

DEAR anindya stress,


I THINK YOU ARE MAKING MISTAKE. SEE BELLOW FILE(PAGES 8-10).

DEAR DAVE DIEHL,

ARE YOU AGREE WITH ANINIDYA STRESS OR NOT?

Arian,

If you agree to be more specific on what you think is wrong in that post, I would try to "defend" Anindya's explanation.
In this moment I cannot say it is a contradiction between what Anindya wrote and the document you've posted.
Maybe Anindya tried to explain more than is necessary, but I cannot blame it... just because I use to do the same!

I would add something.
For tied EJ and for rigid pipe, the balance of pressure gives "Fp"="PxAn" as tension in nozzle.

IMO, it is wrong to add this figure as a supplementary force acting on nozzle and to consider coincidentally there a pressure longitudinal stress.

I think it is not well understood that Fp should be the same as longitudinal pressure stress multiply by pipe metal area. In practice there is a difference when we work with the formula pD/4t which is an approximation.

Best regards.

DEAR MARIOG,

THANKS FOR UR ILLUSTRATION. I HAVE A QUESTION?
IN AN UNTIED EJ,IS THE FORCE ON PUMP ANCHOR(FOUNDATION) LESSER THAN TIED ONE?

For an tied EJ properly working (with no nuts gap), you are in the case of pag 9 in your attachment: net force on equipment=0

For an untired EJ you are in the case of pag 10 in your attachment: net force on equipment=F

Dear Arian,

I will try to give you a more detailed explanation.

Let’s consider a equipment under pressure having a nozzle.
So a pump, vessel, etc.

Considering only the pressure action on the internal surface of vessel, there is always an imbalanced pressure force equals to pressure multiplied by nozzles’ "hole" area.
"Hole" means a missing internal surface; equivalently you may say that on the opposite equipment "side" appears to be an "additional" area.
This general fact has nothing to do with "tied" EJ, "untied EJ" or "no EJ". That force (I would prefer to note it as "Fp" instead pAn in your reference attached) is there because equipment has a nozzle.

The story is not ending here; the final depends on what is happening with the nozzle.
The nozzle is connected to piping, that’s why we put a nozzle there!

1. In case piping is without EJ, pressure exerts a tensile force to the nozzle. This is a consequence of the fact pressure is acting also on the first elbow in front of nozzle. See page 9 of your attachment.

In this case, the net pressure force on equipment is zero because the tension Fp in nozzle compensates the opposite force (that exists because in internal equipment surface there is a "hole" = missing internal surface). See again the page 9.

What is happening with the tension in nozzle? It would be considered as an "external load" for nozzle?

If Vendor already counted the existence of a longitudinal pressure tensile stress in nozzle, the answer is negative. IMO, you need to consider supplementary this force acting on nozzle ONLY if you are advised that the longitudinal pressure stress has been ignored in establishing the criteria of "allowable loads".

2. When you have piping with "tied EJ", this is similar to point 1. The only concern is to be sure that pressure in your piping is able to provide that tension in nozzle. If you have a particular reason for which you have a gap at nuts, this is not happening, because you are in case 3 (this is likely to be particularly true when you put an anchor immediately after an tied EJ)

3. In case you have an "untied EJ" or "tied EJ with gap at nuts" on your piping, you just succeeded to destroy the "metal" continuity in your pipe and you are not able to transmit that tension to the nozzle. That pressure force on the "first elbow" exists, of course but is not longer giving a tension in piping. That;s because your piping is now "broken" into two "independent" (against "pressure effect" criterion!) parts: one part is "restricted" by the nozzle, the second one must be controlled in other way (see that "anchor" in page 10 of your attachment).

What is happening now with the piping attached to nozzle? It’s clear is nothing to put in tension the nozzle, in fact pressure gives now a compressive force to that nozzle. Looking to the piping part attached to nozzle, thsi compressive force appears because the pressure is acting on bellows (EJ convolution) and gives a supplementary compressive force equal to pressure multiplied by an annular area "around" the sectional nozzle area (let’s say around the "hole" area). See page 10 of your attachment, but there F correspond to the sum of Fp and the force on annular, i.e. an total F acting on the "big" area).

In practice, for the nozzle point of view, this would be "better" than Vendor expected because the module of actual force may be less than Fp (of course I exclude the theoretical case when that annular area is huge…).
However, I think here things are more delicate...
As magnitude, this compressive force may be less than Fp.
However, it is likely that establishing "allowable loads on nozzle" criteria, Vendor already considered a tensile Fp (or equivalent, a longitudinal stress pressure), at least is what we considered in our discussion for case 1.
So, in my understanding, Vendor expected to have on nozzle Fp as tension and you have now a compression force; a possible approach to solve it is to consider the vector difference as "external load" (I mean it appears as an external compressive force is acting there modifying tensile force Fp to an compressive one= p* annular area).
This is my opinion, and I think the right answer depends on "what considered Vendor" in establishing nozzles loads criteria.
Anyway, it appears as a correct approach if you think is correct to ignore Fp on nozzle in case 1 .

For the equipment point of view the missing tension in nozzle is bed news because now that force (applied on annular area) adds to the Fp (which is due to "hole existence"); "adds" because is a compressive force, so for equipment has the same sign as Fp.

Please note that I preferred to use other notations than in your reference, Fp is instead of pAn and their "F" is Fp+ pressure force on annular area.

Just hope this will help you.

Best regards.

DEAR MARIOG,

THANK YOU FOR COMPLEMENTARY EXPLANATION. YOUR EXPLANATION IS SO CLEAR.

BEST REGARDS,