Anchor loads for GRP lines 2

[B.L.Smith]

Hi
In a 30"(NPS),12m straight GRP line at ambient Temperature and 30 bars pressure between a nozzle and an anchor, restraint loads on nozzle and anchor are 40 times of allowable nozzle load. what is wrong here?

 

[psapiensone]

If you have two points restricting the pipe in the same direction, it can not move .... you need to change the design.

See the basic rules of piping flexibility.

 

[B.L.Smith]

But how can we change the design for 30"diameter and 12m pipe?
Expansion joint? Doesn't it increase nozzle force due to pressure thrust?

 

[psapiensone]

this is a good question...you will find the solution! In your 12m you can make many things!

 

[C2it]

You need to think this one through. Let's assume that the problem is understanding the effects of axial pressure elongation in the GRP line, between two fixed vessel nozzles.

If the pipe were closed, say with 'valves' at each nozzle, the pipe would be an independent pressure containing system. Pressurising the pipe alone would cause the restrained ends to 'see' the resisted axial load due to pressure elongation, like a long balloon being inflated.

But, if the 'valves' were open, the axial pressure force will act on the wall of the pressure vessel opposite to the nozzle. There will then be no axial force directly in the pipe wall and there should be no axial elongation of the pipe. Thus the real axial nozzle load will be small, ignoring the poisson effects of hoop stress of course.

 

[B.L.Smith]

Thank you C2it. But in Caesar and based on codes(for example BS7159 and UKOOA)nozzle load is measured according to pipe with end cap(and bourdon effect on). Should I ignore results of the software and codes in calculating restraints loads? Is this permissible?

 

[C2it]

Personally I include pressure elongation, known these days (erroneously) as Bourdon effect in all pipe stress calcs as a general rule. It's not so significant in average metal systems but is so dominant in GRP that free axial expansion will often exceed thermal effects.

What you have is a piece of geometry that is an 'exception to the rule', because it is too simple'. I hope my words above make it clear that you need to do some lateral thinking in addition to any stress calcs. No Code will advise you about this, it's just good engineering practice.

I suppose one should ask the question, why are you doing a stress calc on such a simple bit of pipe in the first place ?

 

[B.L.Smith]

1- Nonmetal lines should be analyzed.
2- Pressure and dimension are high.

As you mentioned, I think we should include pressure elongation in GRP in all conditions.

 

[C2it]

You are risking being blind to the obvious.

I said "as a general rule" include Bourdon effect in pipe stress calcs. I don't mean to repeat myself, but in the case you have described I previously tried to clarify that you should NOT include Bourdon effect. If you set up Caesar II et al that way, the axial pressure load obediently calculated simply cannot exist. If you wish to hugely over-engineer your end load resistance, that's up to you.

 

[denniskb]

Brad,

You are right to be concerned about this configuration as it certainly has the capacity to overload the nozzle.

With no temperature change and the two ends open the Poisson's effect will cause the pipe to get shorter (and NO C2it you must not ignore the Poisson's effect) and this may be sufficient to overload the nozzle. Careful though because CII typically reports the "closed end" axial stress even when it is not there because this approach is normally conservative.

Turning swithches on and off in CII can make the problem appear to go away but what you need is to be sure that it really will.

To analyse this situation you should put CII away and calculate the response of the 12 m pipe anchored (at the anchor) and free at the nozzle. Then apply this movement to the nozzle to see if it overloads. If not then you are fine. The real response is going to be a battle between the nozzle deflection snd the pipe growth and will be less than the above.

If the load is too high can you configure the anchor so it has some free movement (slack) before it takes up the load? This way the problem goes away for real.



Dennis Kirk Engineering

http://www.ozemail.com.au/~denniskb

 

[Pipingtop]

Brad,
What load case did it show nozzle overstressed?
At ambient temperature axial loads due to expansion should be very minimal.
Depending on how you define an anchor in your model. A fixed anchor is not normally used in GRP piping.
Nozzle loads can be reduced if you perform the following:
1. span length 12m support is too long for a 30" GRP pipe.
Reduce support length to 7m and this will reduce the nozzle loads.
2. Replace the anchor with limit stop + guide and allow +/-3mm gap on each side of the stop.

 

[B.L.Smith]

Dear Pipingtop,
- The huge anchor load is due to pressure thrust and not temperature.
- I have seen fixed anchor at GRP line. Do u have a handbook or document that forbids use of fixed anchors in GRP lines?

 

[Pipingtop]

Brad,
No handbook or any other design codes/documents that forbid using fixed anchor on GRP piping.
It looks like that you probably have an untied expansion bellow installed near the nozzle and this is where you get the pressure thrust generated by the expansion bellow.
Without knowing the details of your design, my suggestion is to release the anchor or provide more flexibility in your pipe run.
Also add tied rods to your expansion bellow to absord the pressure thrust otherwise you have to design the pipe support structure or reinforce your nozzle to cater for large axial loads.
Good luck