Thermal bowing 1

[Eigensolver81]

I have a piping with a commodity of LNG. This is subjected to thermal bowing. I have put an anchors in every straight lines, also hold down supports. But when I check the result, the pipe is failing. What should be the proper support for a piping subjected to thermal bowing?

I doubt the result of the analysis since it is more worst compare to the thermal stress of the piping system. I assumed that the hold down supports counters the effect of thermal bowing.

Any thoughts on these?

 

[desertfox]

Hi

What about an expansion joint which allows the pipe to expand without subjecting it to stress

 

[LittleInch]

Ok, to really get useful responses and replies on this site, you really need to give us as much info as you can. Therefore please add pipe size, wall thickness, material, operating pressure , op temp (I assume for LNG is about minus 160 C) to ambient(temp before you put in LNG) and most importantly some sort of sketch or plan of your piping and where you have problems. You seem to be trying to solve a particualr issue instead of looking at the whole layout to see why the issue occurred in the first place.

From what little info you've provided so far, a few things seem a bit strange. You say "thermal bowing" - I can only assume you mean thermal expansion causing the pipe to lift off the pipe supports. For an LNG system it's normally thermal contraction which causes the problems as LNG is introduced into the system. Putting in anchors everywhere normally causes you more problems than ou solve. When you get differential temperatures, pipe wants to expand or contract, often not very far, but if you lock it in place then you can get large forces and concentrate movement into certain locations. Preventing uplift by putting in hold down supports might stop the movement, but increases the forces instead, which is why your pipe may be "failing". Double check your inputs and make sure all the signs are correct when you operate at high negative tempereatures.

You need to look at your piping layout and see where the movements can occur which add flexibility into the system - It you make it very rigid then you get big problems - If you allow the pipe to move either by bellows, expansion loops or elbows then the forces can reduce by a factor of 10 for only a few mm of movement. Only anchor things when you really need to.

My motto: Learn something new every day

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

 

[Eigensolver81]

Hello LittleInch,

Do you understand thermal bowing very well? I presume you understand where is the hot and cold surface of the pipe when LNG introduced in the system at ambient temp. Then I presume you will know where is the contraction and expansion surface of the pipe that makes lift-off from support.

Anyway, I have a NPS 4 pipe, sch 10-s, A312-TP316 (stainless), 0.25 bar and -171 deg celcius.

The system has problem on thermal stress range.

 

[LittleInch]

Now becomes a bit clearer.

See

http://www.eng-tips.com/viewthread.cfm?qid=153799

Is the software you're using doing it correctly? This is not a common effect so may not be fully covered in the analysis.

Also how much is the pipe bowing? If this is a rare / short term event then think about leaving it if it doesn't either overstress or fall off the support or depending on the amount of lift, look at constant load pipe supports and hangers which move with the pipe but are spring loaded or use weights to allow movement by still provide supports and guidance. Look up carpenter and Paterson for starters -

http://www.cp-ltd.co.uk/

As said, once you start introducing stops, anchors and other things that stop the pipe moving, then you can "solve" one problem but create three more. Whilst it might look a bit odd, having a pipe bow and lift off the supports for a transient effect may just need horizontal guides to stop it falling off the support once it fill completely with LNG and comes back down to sit on the support.

My motto: Learn something new every day

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

 

[Eigensolver81]

LittleInch, are you joking? aren't you?

Using a spring on thermal bowing?

In my case i have no experience yet in thermal bowing analysis, "But" I know that this bowing effect is caused by an introduction of cold fluid which is partially filled in a piping system, other cases were due to rains etc.

The bowing effect creates moment on both ends with restraints, also a displacement on the middle of span with an equivalent force calculated from basic beam formulas.

Any other comments?

 

[LittleInch]

No

My motto: Learn something new every day

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

 

[davefitz]

Most of the above replies are correct, but it is not clear what teh real source of the bowing is.

If the bowing is occurring at a distributionheader which has all distribution nozzles lcoated on one side of the header, then the solution is to locate half of the nozzle on the opposite side of the header.

If the bowing is occurring due to stratified flow on a horizontal line, with the bottom (invert) liquid boiling, then a potential solution is to add a internal swirler device to esnure the liquid washes the entire circumference- but if teh swirler breaks free, then other issues ensue.

If the bowing is casued by end reactions on the piping, then adding increased flexibility plus springs is the solution- as litteinch says the worst response would be to add increased rigidity to the system- the more flexible it is ,the better.

"Nobody expects the Spanish Inquisition! "

 

[LittleInch]

Actually a couple more - Bowing effect is caused by differential temperature effect, no matter how it's caused.

And I said spring loaded supports, not springs.

The other reason you're gettng high loads is use of Stl Steel. If you're really running in to trouble you may need Invar piping, but the key is to enhance the flexibilty, not make it more rigid or prevent movements.

Also providing "witty" return comments or "funny" smileys and no real response to quite reasonable questions is guarenteed to get you a lot of responses.....

Over and out.

My motto: Learn something new every day

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

 

[Eigensolver81]

Hi,

The bowing is caused by uneven thermal distribution on pipe surface.

Okay, how will you analyse the bowing effect on piping using manual calculation? then how will you do this in computer analysis using CAESAR II or Autopipe? what would be the thermal bowing temperature?

 

[Eigensolver81]

Okay littleinch, I will try your suggestions, anyway you are right not to restrict the thermal movement. I can imagine what you are saying. Thanks.

 

[ione]

From the “Special Execution Parameters” menu insert a delta T value equal to the difference between the upper and lower pipe of the pipe.

 

[BigInch]

You might try using some constant load spring cans hangers in particularly problematic areas.

Independent events are seldomly independent.

 

[desertfox]

Hi

A couple of references on expansion joints

http://www.jameswalker.biz/en/pdf_docs/9-comflex-metallic-expansion-joints-engineering-guide

http://www.minikins.co.uk/IMG/minikindesignbook4thedition.pdf

 

[chicopee]

Isn't the pipe insulated thereby minimizing the effect of localized heating from such source as from the sun's radiant heat? It seems that insulation would be an effective way to control thermal bowing.