EXTENT OF BURIED PIPE MODEL

[GPB005]

My experience says that the buried pipe should be modelled till virtual anchor lenght in a software... some times we model twice of the virtual anchor length. Virtual anchor length can be calculatd by hand calculations.
However client is asking us to model 10 km long pipeline in CAESAR II. When we put forward the above statement and denied modelling the entire run client asked us to give reference document where it is stated that beyond virtual ancor line need not be modelled and generic calculations for equivalnt stress are acceptable. Further how to ensure the safety of cold bends?
any inputs

 

[LittleInch]

That approach is Ok only if the intervening points are very straight and very flat.

Otherwise you are not analysing the pipeline. Often the road crossing in the middle of the line is the one that causes a problem...

I don't understand your comment about the cold bends. You analyse them in the software to see if they are below the allowable stress limit. Please expand your question.

My motto: Learn something new every day

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

 

[BigInch]

You need to model any point which you cannot prove is fully restrained. Since bends between anchor points cause a misalignment of axial forces, which can be quite high, the misaligned axial forces might very easily move the bend out and away in the lateral direction, allowing some axial movement as well, if the soils within the trench are not very tightly compacted. Hence bends, even though inside virtual anchors at the ends of pipeline, would need their very own additional anchors, real or virtual, one on each side of the bend as well.

In other words, any point in a pipeline that moves and eventually comes to rest, also requires anchors on each side, real or virtual.

The codes themselves say that fully restrained segments do not, by definition of a restrained section itself, require flexibility strain analysis. Why model the flexibility and strain of something that has no strain and flexibility.

you must get smarter than the software you're using.

 

[GPB005]

thanks all,
Is it an acceptable approach that we make separate models like:
1. Pig Trap to little further ofvirtual anchor
2. model for each bend with pipe modelled till virtual anchor length on either side of bends?

Means we need not model a continuous 10 km long model. Software takes time to handle it.
Further is it required to model the elastic bens with radius 40D or 80D or shall we consider them restrained? I belive the software using beam elements are not effective for bend analysis and put us under trouble.

regards

 

[BigInch]

1 & 2 Yes.
The same thing can happen to a vertical bend (overbend) as a horizontal bend. Vertical uplift at the apex of the overbend is possible. if sufficient soil cover is not provided. Furthermore, buckling in the vertical direction may occur if axial compression loads are high enough.

CAUTION
40-80D are Cold Bends and exceed the elastic limit of steel, but when released from the bending machine, only relatively small residual stresses remain.
Hot bends are similar, made to much smaller radaii and residual stresses are reduced.
Elastic bends create stresses lower than elastic limit as their bend radaii are usually greater than several hundred meters, but lock in the elastic bending stress forever until the pipe is straightened out again.

you must get smarter than the software you're using.

 

[LittleInch]

you could always greatly reduce the number of nodes in the straight section (say one every 500m) and then the client will be happier that you have analysed the whole system...

My motto: Learn something new every day

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

 

[BigInch]

Then they'll find that there are 100 V&H bends in their 10km and will want to model each one and when they get a 1000 km pipeline, what will you do with them then?

you must get smarter than the software you're using.