How to model ship impacting pipe?

[GasGuy]

I would like your thoughts on how to model a ship impact on an over water pipeline. I want to determine the force and displacement required to rupture the pipe (i.e. the work). I was thinking of a non-linear analysis but am unsure of how to determine at what point a leak would begin. Any ideas?

 

[GBor]

Not wanting to seem rude, but this type of analysis is not a "walk in the park". How familiar are you with non-linear analyses? It appears you will be looking at impact, large displacements, and non-linear material models all of which are not simply pushing a few buttons.

Do you already have a particular FEA package that you are using?

Garland E. Borowski, PE
Borowski Engineering & Analytical Services, Inc.
Lower Alabama SolidWorks Users Group

 

[GasGuy]

Using Strand7. What criteria would you use to determine if a leak has occured?

 

[GBor]

Using Strand7, I'm not certain. Generally, I'm looking for yield, but you would be looking for net section failure and potentially a location exceeding the ultimate strength of the material. I might suggest that you find a metallurgist that can take your stress field and check for crack propogation, material attributes, etc.

As engineers, we tend to look in a book and say, "so that's the yield strength", but you may be dealing with high strain-rate effects (like a car crash for which LS-Dyna is generally the software of choice), temperature effects (depending on where your pipeline is physically located, salt water effects, welding in the pipeline...and I'm sure the materials guys would suggest another consideration or two.

I would also say that you may want to run a few small tests to insure that your analysis techniques are matching reality.

Garland

 

[Stringmaker]

GasGuy, not sure what capabilities Strand7 has but I'm hoping for you it has some sort of explicit solver. As GBor said you're talking about an impact problem with large displacements and mucho plasticity that will probably take an implicit solver a light year to solve. Do a little research on FEA impact analyses. Strand7's documentation is probably a great place to start as well as in general resources on the web or in texts if you have them.

The conservative approach would be to model the ship hull as a rigid body that impacts the pipeline. Therefore the pipe absorbs all of the impact energy. The realistic approach is perhaps something less than that. You understand the event and assumptions better than I do so I'll leave that up to you. Stuff like this is really it's own seperate branch of FEA and requires a methodical approach.

Good luck!
-Brian

 

[GBor]

I "grew up" in the military shipbuilding industry. I don't recall Strand 7 having an explicit solver. I would agree with Stringmaker, but you may be able to get away with an implicit solver...they are moving MUCH faster these days. Good thought on the ship as a rigid body. There are parts of a ship that are so incredibly stiffened that they will act pretty much rigid compared to a pipe wall...even a really large pipe.

If you switch softwares but maintain an implicit solver, you may want to make sure it has some type of kinematic element that doesn't bother calculating stresses. Use these for the ship. You will also need one that transfers inertial loads.

 

[GasGuy]

What I'm REALLY trying to figure out from this exercise is: Approximately how much kinetic energy must a boat transfer to the pipe to cause a pipe rupture?

My thinking was to apply a load to the pipe to displace it incrementally until it ruptures (not sure how I figure out where that point is). Then by plotting the force vs displacement and integrating to get the area I'll have the work done on the pipe which can be equated to the vessel KE.

This is not an exact study and clearly the pipe section in question will be replaced if a vessel ever contacted it. The context is a probability input for a risk analysis.