Collapse of Subsea Pipelines

[desiry]

Good day, I am trying to carry out a finite element analysis on collapse of subsea pipelines under different loading conditions, ie straing hardening and combined loads.

In the presence of external pressure, I entered my parameters and I was expecting to get a result with a oval shaped pipe or a collapsed pipe but what I am getting is:

(i) With one end of the model fixed or constrained (ie BC - boundary condition), a pipe with reduced diameter (shrinking) with the length extending towards the unconstrained end. Sometimes it comes out like a conical geometrical shape.

(ii) with both ends of the pipe fixed, there is a gradual reduction of the diameter from both ends and looks like two conical figures attached to each other at their top ends.

What can I do to get an oval shaped collapse? Can you please help? Thanks!

 

[corus]

Not sure why it should be oval, but have you included self weight, and a varying external pressure with depth?

corus

 

[mrgoldthorpe]

Given your description of the results, I presume the model is 'perfect' and the external pressure distribution is uniform.

You perhaps need an initial geometric imperfection and NLGEOM to cause a non-symmetric buckling type response.

 

[desiry]

mrgoldthorpe, thanks for you response. corus, I am much obliged.
I can toggle on NLGEOM. But how do I introduce an initial geometric imperfection?

 

[desiry]

I have introduced a NLGEOM but after running the analysis, there is no deformation.

If I toggle off the NLGEOM, I get the results which I described at the beginning of this thread.

Maybe I am not doing something right but I am not sure. I have checked my parameters, including the external preesure but can't figure out what is wrong...

 

[mrgoldthorpe]

edecyri,

Firstly, it seems like you are fully restraining the end(s) of the pipe rather than just restraining the axial movement of nodes on the ends. Change that and you should see that the 'conical' shape disappears.

Secondly, to introduce a geometrical imperfection just reduce the pipe thickness over a particular region by slightly changing the positions of some nodes. What will happen is that the deformation will 'localize' in that area.

I suspect that the model and the loading of it is completely uniform; so without an imperfection the pipe will just uniformly shrink in diameter and extend in length due to the Poisson effect.

 

[corus]

I'd apply continuity type restraints at the ends of the pipe so you restrain the ends rotationally and planar.
You can apply a geometric imperfection but an alternative might be to simply apply non symmetric loading to give you that geometric imperfection, as would apply in real life. I imagine that a pipe on the sea bed under self weight alone would adopt an oval shape with a simple restraint underneath. An external pressure added to that would give you the buckling mode you require.

corus

 

[kellnerp]

When a pipe collapses from uniformly distributed external pressure that is generally a buckling problem and the deformed shape will be kind of wrinkled looking. Unless you are looking at an extremely large pipe (>100ft??) where there is a significant hydrostatic pressure differential it probably won't be an oval. If you are looking at a large "rock" on top of the pipe then you might get an oval.

The boundary conditions on the ends of the pipe should be those suitable for a body of revolution, i.e., circumferential and not radial using a cylindrical coordinate system unless the pipe is fixed to a flat plate or other substantial structure.

TOP
CSWP, BSSE

http://www.engtran.com

http://www.niswug.org

"Node news is good news."