Supports on Steam Pipes on a Ship 2

[samuelo]

Hi,

I'm working with an installation with two boilers and a turbine, with turboalternators and turbopumps. The steam obtained is at 62Kg/cm2 and aprox. 560ºC (superheated). The material of the pipes is an alloy of Cr-Mo, ST35.8 and others depending on the temperatures and pressures involved.

The question is that the plant is for propelling a ship, thus I have to take into account the movements of the ship. I think (correct me if I'm not right), that the ASME B31.1 code is not prepared for this movements.

Do you know any literature or can give me some approach to the problem?

Thank you

 

[hacksaw]

B31.1 has been used on ships in the past

 

[yiqunzhang]

If you are working on the boiler B31.1 does not apply, but ASME BPVC. If you are woring on external piping and joint of the boiler B31.1 is good.

 

[KANN]

How is condensate drainage of steam piping handled with ship movement? Could not the condensate flow be co-current and counter-current in the same steam pipe? Are the steam pipes oversized to a certain maximum velocity to permit condensate drainage independent of the slope of the steam pipe?

 

[22catch]

Make sure the support configuration is such that all suports and restraints are active in all conditions and are independent of ship movement.

 

[samuelo]

When the drainage is short, it's not supported, so the expansion of the big pipe doesn't cause the drainage to get tensioned. If it's long, the first support is conveniently positioned with the appropiate length.

22catch, that's the question, how do I manage to get the supports independent of the movements? With the temperatures involved, I cannot put many fixed points, because of the thermal expansion, only the connections to the equipments.

Thank you all for the answers.

 

[zeven]

The condensate flow would be a secondary issue as long as the pipes have a correct fall when the ship is level. The condensate will eventually find its way to the collection points. You may require bigger drain pockets, as it may arrive in spurts rather than smoothly.
As far as the actual supporting, due to the dynamic nature of the supporting structure (the ship), I would suggest that you would need to be carefull having too soft a system (avoid using too many constants) and try to guide the pipework more than usual.
The natural frequency (and harmonics) of the ships roll needs to be avoided.
If there are large lateral thermal movements then use sway braces instead of rigid guides. They have a spring rate and are designed to react against the pipe being pulled out of position.
Another thing to avoid would be to have rod supports all the same pivot length as the whole piping system would tend to swing in the same arc.

 

[samuelo]

Zeven,
I had thought already in the sway-braces, but I have a problem with them. I cannot model them correctly in the program I'm using (CaePipe). If I put the sway-braces as snubbers, they don't appear in the sustained or operational condition, but only in the ocasional one (dynamic with the accelerations assesed in the ship), while in the reality, they should act on the pipe with an opponent force as you explained, so I think the displacements are not going to be the expected ones. Is there another way to model the sway-braces?

Thank you

 

[zeven]

Sway braces are in effect a very soft snubber, and the trick is to get maximum effect for the dynamic cases without
overloading the thermal case.

Your right, its not easy to model sway braces, and much depends on the particular manufacturer as to how to model them.
In most cases they are non-linear springs, but a reasonable modified spring rate (for the thermal case) can be obtained through iteration.

In this case, they are essentially modelled as a rigid type support,but with a spring rate (much the same way snubbers are sometimes modelled as stiff springs). They should be included in all cases. To be conservative you should use the given manufactures spring stiffness for all cases except thermal. A modified one (taken into account the preset load) should be used for the thermal case.

As the effect of the sway brace(s) may reduce the overall thermal movement, then seveal iterations may be required to get an accurate picture.

Also be aware, on soft systems, that the preset may never be overcome. This can be checked by modelling the struts as rigids any checking the loading.

I have only experience with the LISEGA type.
The LISEGA type has a preset load which must be overcome before the any deflection is achieved.
The preset can be defined by the analyst within the range of the size of brace chosen.
The spring rate is also dependant on the size chosen.
Modified spring rate is calculated by adding the preset load with the estimated deflection x spring rate, and then dividing by the estimated deflection.
One other detail is that they can also have a gap (free stroke), so they can used as a gap type restraint as well.
Please refer to the LISEGA catalog for details. you can download from

http://www.lisega.com

in pdf format Section 2 page 2.13