Pipe connections External

[stevenc80]

Please can someone help, i have been asked, "Please give loads Fx, Fy, Fz and max torque Mx, My, Mz for all connections." I am designing a pumping system for a Russian oil/gas company. How do i calculate these figures please? I am not sure if they are asking for loads due to pressure, the weight of the flanges or max allowable loads. There is mention of max allowable loads in the API 674 standards, but a previous example of of these loads does not seem to tally. I am quite Naive in this are i have joined a pump system manufacturer but i am not experienced in the field of hydraulics. Having said that no-one else in our company knows how to do this. Thanks for reading.

Steve

 

[LittleInch]

Basically you do a stress analysis of the pipework, typically using something like Caesar II or similar software which models all your pipework and issues reports of all loads at the nodes, including the connection points for a set combination of events - operational, testing, incidental loads (snow, wind etc). Search stress analysis on this site and you'll get hundreds of posts...

Sounds like you need a piping designer to me.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[stevenc80]

really? is it that complicated?

 

[LittleInch]

Not if you know what you're doing and your pipework isn't too complex.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[BigInch]

If you are the manufacturer of the pump, I think that your clients are asking you to provide them with the maximum allowable loads that they can apply to the flanges of your pump without warping the casing, moving it out of alignment, or causing some other type of damage to your pump. These allowable loads are often quite small.

I hate Windowz 8!!!!

 

[stevenc80]

My customer has sent this example, but i am still none the wiser. Does this enlighten anybody else?

 

[XL83NL]

Does this enlighten anybody else?

Yep, seems likes he's after the allowable nozzle loads (which is a set of 3 forces in x,y,z and 3 moments in x,y,z,) for each nozzle.
If your pump design is to a certain API spec, changes are that spec will prescribe the minimum required loads a nozzle can take, depending on flange size and pound class.

 

[LittleInch]

We're none the wiser about whether you are the pump designer giving info to the piping designer or vice versa. what exactly is a "pump system manufacturer?" Are you designing skids which include pumps or what? If the latter, you need to model your skid and input a few loads to see if your skid or pump nozzle can stand them.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[racookpe1978]

Hint: The pump seal manufacturers want 0.0 forces and 0.0 moments on that shaft under all conditions.

The pipe (skid and pipeline and pipe support and pipe arrangement and pipe stress) engineers can't always deliver 0.0, and work very hard to minimize those stresses under all expected conditions.

What your customer is expecting (rightly!!) is that YOU have solved for as many conditions of pressure temperature and installation geometry as you can expect AND have the results of YOUR calculations available so HE can confirm YOU have done YOUR job to minimize stress on HIS pump and pump seals so HE won't get fined for EPA leaks and lose dollars being shutdown replacing expensive high-tolerance pump seals.

If YOU can't produce YOUR analysis results under his conditions, HE (properly) won't believe you have done your job.

 

[stevenc80]

Hi Guys,

Back from my holidays, I hope you all had a merry christmas. In reply, our company manufactures pumps, my job a design engineer is to use the pumps to build a system that suits the customers application. Designing a skid, piping and adding valves as required. The inlet connection of the pump is a rubber expansion coupling and the outlet is via flexible hose, these are 3" and 2" respectively. All the other connections to the pump (cooling and flushing) are BSP threaded connections between 3/4" to 1.5." And are intended to be connected into via flexible hose by the customer.

There is a statement in the API specifications that reads "the vendor shall specify, in the quotation, the magnitude of forces and moments which may be applied, simultaneously, to the inlet and outlet connections at the rated operating conditions.

Pumps shall be designed for satisfactory performance if subjected to the forces and moments in table 6"

This table then gives max forces and moments for particular pipe sizes. But surely these values are only applicable to hard piping?

I don't see how it is possible for me to calculate any loads that may be applied. All their hard piping connected to our flexible must be self supporting. I am unable to get this allowable load information on our hose or expansion coupling. And for the BSP's i have looked and there is not and information to stated what loads are applyable to these threads.

I'm pretty sure i must have misunderstood and I'm barking up the wrong tree.

Our customer is insisting I put some figures on the GA moments and forces in all three directions for each connection. However the customer is not the end user, it is the end user that seems to request these figures as standard. Our customer has actually stated that he does not believe the end user will take any notice of these figures and stated that i could make them up and they will not be checked, however i am not at all comfortable putting my name to this sort of work.

 

[pennpiper]

Do people today really believe that "connecting the pump suction and discharge by a rubber expansion coupling" will eliminate the movement, loading, force and moments on Pump nozzles?

Years age we had a very smart Pipe Stress Engineer debunk this myth for us. He had the Model Shop (yes it was that far back) build a test stand for the classroom to simulate the pump nozzle, some piping and a rubber connector (1 1/2" radiator hose). This was made air tight and fitted with an air pump. Anchors and guides were included but made removable so we had the option of guided or unguided tests. There was a strain gauges so we were able to see the reaction of force on the "pump nozzle".

Adding air pressure to the model simulated the pressure force of a real piping system:
- when the model was anchored and unguided the forces were low but the rubber joint warped out of shape and the piping shifted out of alignment.
- when the pipe was anchored and guided the forces were very high and the rubber joint warped out of shape but the piping stayed in alignment.
- when the pipe was anchored and modestly pressurized and unguided and the distance between the two anchors was shortened (an adjustment screw simulating thermal expansion) the rubber joint and the piping warped out of shape.
- when the pipe was anchored and modestly pressurized and guided and the distance between the two anchors was shortened (an adjustment screw simulating thermal expansion) the rubber joint warped out of shape but the piping stayed in place.

In all these cases the gauges showed strain at the "pump nozzle"

- when the anchors and guides were all removed and the pressure was increased the strain gauge showed no strain. However there would be dead loading from the weight of the piping (as would also be the case in all the other tests).



prognosis: Lead or Lag

 

[stevenc80]

I'm not saying it will eliminate any movement entirely, but it is obviously going to reduce it significantly over hard piping directly to the pump.