ABAQUS - buried pipeline modelling 1

[jaypatel23]

Hi all
Am modelling an onshore pipeline which is buried to about 2m. Objective is to assess a weld defect & confirm if stresses are within allowable. So only modelling a small section (~5m) in 3D around the girth weld. I want to include the soil load but am not sure how best to do this; the options are a) pressure load b) body force load (which has units for Force per Length^3 (rather odd...!). Any ideas on how this should be done in a 3D model? Bearing in mind I'm not so interested in the global behaviour - the pipeline has low loading & not susceptible to buckling etc.

I'm also thinking given the objective, excluding the soil load is conservative for hoop stress...? But what would be suitable boundary conditions in this case i.e. fix bottom half in vertical?
Any thoughts/ideas? Thank you.
​

 

[FEA way]

You can model both pipe and surrounding soil (using for example Mohr-Coulomb material model for soil) and define contact between these parts.

Abaqus also offers pipe-soil interaction elements but they are used with simplified pipe models (pipes represented with 1D elements - beam, pipe or elbow type).

 

[goutam_freelance]

Even if you are not following ASME B31.1 you can refer the following for good guidance.
B31.1 NONMANDATORY APPENDIX VII
PROCEDURES FOR THE DESIGN OF RESTRAINEDUNDERGROUND PIPING


Engineers, think what we have done to the environment !

https://www.linkedin.com/in/goutam-das-59743b30/

 

[1503-44]

The conventional approach is to model soil as a pressure load. Do not forget to include the surcharge from the water above.

Depending on depth, critical load may occur at minimum internal pressure causing collapses of the pipe due to external pressure. Normally an initial imperfection in the pipe's circular cross section is modeled as an initiator of a local buckling collapse.

 

[jaypatel23]

Thanks all. Given the pipeline response within the soil is not the focus, am trying to keep it as simple as I can (time being the other issue!). So the soil load as a pressure could be a good & quick way to do it. Would this apply around the whole circumference? Or top half only...
Note: Assessing buckling not part of the problem as only interested in the impact of the defect within girth weld.

Stresses in the hoop direction will be the most critical, so excluding the soil pressure (external) would be conservative. And it is an order of magnitude smaller.

 

[1503-44]

All the way around,

 

[LittleInch]

It would help if we could understand where this weld is.

Is it on a straight fully restrained portion of pipeline or coming into a bend or other location where you have bending moments and expansion forces etc?

For the former then Ok, modelling a straight potion and applying tensile or compressive forces to the pipe plus soil load etc seems ok, but not for the anything else.

I'm a little surprised abaqus doesn't already cover this in their manual or help pages.

And is this weld defect only now found after inspection and having been in service a long time?

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

 

[jaypatel23]

Hi LittleInch

Yes, its on a straight section at 12 o'clock position. Defect found after some years of operation; it doesn't pass the analytical checks & hence FEA to take a closer look. Not believed to be a serious issue but require some analysis to back it up. Also the reason why the soil modelling should be fit-for-purpose - which I think the soil pressure approach would achieve.

I've not found anything directly in the manual. Came across lots of complicated pipe-soil models etc. but not something similar.

Cheers.

 

[1503-44]

I've never seen soil modeled any other way but pressure in pipeline, foundation, retaining wall, tunneling, well, or geologic works.