Lateral deflection in pipeline

[Stress Engg]

I am performing stress analysis of above ground surface laid flowline with buried portion only at crossings (road, existing facilities etc.). the MOC is API 5L Gr-B and the design code we are using is ASME B31.8. In one of the section of the flowline, we are having an IP (change in direction) angle 12° and with crossings at each ends at a distance of 70m and 20m on each side of bend. The flowline temperature is 120°C and pressure is 100 bar. However, the lateral deflection at IP point (only at far point, mid point and near point of bend) suddenly shoots up to approx. 500mm. Beyond this 3 nodes, the displacement is in the range of 200mm only. I am not able to understand why this much high lateral deflection is happening at bend even though the expansion length is only 90m?

 

[1503-44]

Need a diagram. Can't tell what you mean by "lateral direction" when you have a change in direction.

An above ground pipeline with buried road crossing is usually modeled with an anchor at each end of the crossing.

You should not be using B31.8 for such high temperatures.

Are you are laying 120°C bare pipe on ground surface ?
Is the line insulated?
Is there casing?

What coatings will you try to use at 120°C?

 

[Stress Engg]

The client has asked to use design temp120°C for stress analysis purpose and line is laying directly on ground without any coatng

 

[1503-44]

Let's get the full info please.

Are you only modeling 86m?
Are the Xs at each end perfect anchors? Are they real anchor blocks, or just buried pipe?
Where are the road crossing boundaries?
Where do you show the deflection?
Pipe Wall thickness?
How deep is the pipe buried under the road?
Are you using elbows to go up/down underground.


Maybe you have a mistake in your expansion coefficient.
Maybe wrong units.

 

[Stress Engg]

Its a 3Km long surface laid flowline.
X at each end represents crossings which can be considered as anchor point as buried depth is approx 3 m
the lateral deflection is at the bend in horizontal direction.
pipe is SCH 120.
Expansion coefficient is calculated by Caesar software, we just have to input material. The problem is not the expansion coefficient. If the issue is to be wrong expansion coefficient then displacement should have been increased gradually from anchor point to bend. But in this case, displacement is only at the bend suggesting there is some force which is pushing the bend away in lateral direction

 

[1503-44]

You have 10x expected deflections. That suggests decimal error. Check all your input data.

This is not a guessing game. I can't waste all day of this.
You have 2 road crossings??
Correct diagram?
Wall thickness is ???

 

[Stress Engg]

Can you share your email id? I will send you the Caesar-II file for better understanding...

 

[KevinNZ]

Don't see what the problem is. You will get this sort of expansion if the pipe is on friction less supports.

less supports.

 

[Stress Engg]

Thanks KevinNZ.
in your snapshot, the displacement in y-direction is 450mm, even though the expansion length is only 70 m on one side and 20 m on other side. Can you check why we are encountering this high displacement? The displacement should not be more than 100mm considering CS pipe expansion at temperature 120°C.

 

[1503-44]

Much of the deflection is probably caused by bending moments, but without plot of deflection locations, it's hard to say. Friction of a 129C 4" pipe on ground surface is nothing. You have a strand spaghetti there.

 

[LittleInch]

I would say you're feeding in a lot of the movement into that bend. so the bend will move a lot and then there will be a curve back from the bend to the straight line.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Absolutely. A09 is at "mid span", so that's where bending deflection is greatest.
My guess is -Dy is towards the left direction, but that's on KevinNZ's dwg.