Expansion Loop Design

[PipeNewbie]

Guys,

Firstly congrats on an insightful and knowledgeable forum. I'm just starting a career as a mechanical engineer doing pipe stress analysis and have read a few of threads, which have been very useful.

I have a question that I'm struggling to get my head around.....

Based on the guidance in Kellogg's "Design of Piping Systems", the optimum expansion loop has a K1 value of 0.5, i.e. an expansion loop length equal to 1/2 the distance between supports. This makes sense to me as it creates equal minimum deflection at each of elbows in the loop and allows a minimum leg length (minimising steel required).

Why in practice do you tend to see expansion loops that don't correspond to a K1=0.5 and tend to be small in length relative to the total length of pipe? Sure this just results in a greater displacement at the elbows and longer leg length that would be required?

I've attached a sketch if that doesn't make sense.

Cheers

PipeNewbie

 

[PipeNewbie]

Sorry folks, presentation attached to explain what I'm on about.

 

[BigInch]

The idea is to convert axial stress into bending moment. Pipe has a significant capacity to carry bending moment that is not used in a straight length of pipe. Allowing the axial movement reduces the axial stress, but creates bending moment in the lines parallel to H. However you don't want too much bending stress, so the longer the dimension H, the more flexibility the system has resulting in less bending moment being developed. It is true that increasing your L dimension will also increase flexibility, but not enough. Without increasing lengths H, bending moments will remain quite high.

 

[pennpiper]

PipeNewbie:
I think it would be wise for you to get some field time or search online for examples of real life as it relate to your new career.
Expansion Loops do not always exist in a "one pipe rack, one pipe loop". More often there are many loops 'nested' in the same bay of a pipe rack. For a real good example of this see the following:

http://www.shutterstock.com/pic-871...t-pipes-rack-in-factory-area-aerial-view.html

Your next question will be "WHY" do they do it that way?

With the help of a real good & experienced Pipe Stress Engineer and Piping Layout Designer you should be able to gain the basic understanding of this issue.

Sometimes its possible to do all the right things and still get bad results

 

[BigInch]

Personally I don't like that horizontal loop business. It makes it real difficult to ever expand that rack and each time you try, the loops get bigger and bigger.

 

[pennpiper]

BigInch,
I agree with you completely, but I was trying to make a point and could not find any nested 2D loops at the time.

Here is some help for the OP (Rack Piping for a Piping Stress Engineer) which also shows some nested 2D loops.

Sometimes its possible to do all the right things and still get bad results

 

[bellowsmfg]

Pipenewbie,

In addition to the comments above, and apart from the math, a major reason is real estate and practicality. You're original sketch takes up a larger corridor. The loop from BigInch takes up only local real estate. Do you have access to an experienced piping desinger or engineer? A good walk around an existing facility will show you (hopefully) some good samples.

Conor

https://www.bellowsmfg.com

Bellows Manufacturing and Research, Inc.

https://bellowsmfg.com

 

[racookpe1978]

And, if the expansion loops are "big enough" and vertically oriented, then you have an accessible place to drive trucks and other pipelines and people through at ground level without increasing the need for structural steel.

 

[MJCronin]

The Piping Stress Analysis book by Sam Kannappan contains good information about piping expansion loops.

It is a good addition to a serious engineer's library.

http://www.amazon.com/Introduction-Pipe-Stress-Analysis-Kannappan/dp/1605305146

MJCronin
Sr. Process Engineer

 

[StressGuy]

PipeNewbie,

I don't have the Kellogg book handy, but I must confess I've never seen the loop shape described as your slide shows. The guidance I was given and try to operate by is that the K1 dimension you show should be about one half the depth of the loop (dimension H in BigInch's graphic) as a reasonable proportion. So, a loop that was 25 ft deep would have about a 12.5 ft K1 dimension between the legs. I think you'll find most layouts try for this, with the understanding, as has been pointed out, that loops for several adjacent lines are frequently grouped together and nested in a single bay of the pipe rack.

Doing so has the advantage of only "killing" one bay of the rack where branch lines in/out of the rack can't be run. There is also an aesthetic quality to arranging piping loops together like this. Even though efficiency is the primary goal of piping layouts, there are preferences of a visual nature that piping designers try to pay attention to.

Edward L. Klein
Pipe Stress Engineer
Houston, Texas

"All the world is a Spring"

All opinions expressed here are my own and not my company's.