Hi all,

I am doing stress analysis on a system that contains a significant portion of buried piping. It has lead me to some questions that I cannot find answers for in previous forum posts.

1. If I have a circumferential protrusion on buried pipe (lets say a flange), what is the best way to model this protrusion to capture the drag effect in the soil. I'm thinking of testing the below options:
a. Add pipe branches to the header which protrude from the header with an equal cross sectional area to the circumferential protrusion, and classify these as Zone 1 in buried model.
b. Add a pair of reducers on the header that increases the OD of the pipe to the same OD as the thickness of the flange/protrusion.
c. Add insulation locally in place of the flange/protrusion and set to Zone 1 in buried model, and classify them as Zone 1 in buried model.
I am running tests at the moment but the queries below leave me to doubt my results.

2. Is the insulation thickness used to increase the effective OD of the buried pipe and hence impact the lateral push on pipe (zone 1) or the axial friction (zone 3)?

3. If I have run the 'Underground pipe modeler', and then want to make a change to the buried piping (e.g. increase insulation thickness or add a buried reducer), can I just rerun the B file in the 'Underground pipe modeler' again, or do I need to make the changes in the 'pre-underground modeler' stress model, then put through the 'Underground pipe modeler' again?

4. Is there a way to clear all the soil restraints quickly?

5. If the restraints are removed automatically by the 'Underground modeler', why does it not remove all the soil restraints when I run the B file again with no buried elements selected?

6. Is the Zone 1 (lateral push zone) only important for non-circumferential geometry (e.g. tees, bends) or does it also need to be applied to circumferential geometry (e.g. flanges), reducers.

7. If I am to set a Zone 1 region, do I need to calculate the 'lateral bearing length' myself using Lb = 0.75(π) [4EI/Ktr] 0.25, and then add nodes at these positions in the model, or does the 'Underground Pipe Modeler' do this?

8. In the training notes it suggests to use WNC instead of W+P1 as the baseline for the stress range loadcase. Why is this? Is it something to do with the underground modeler removing pipe and fluid density? I have read that the WW overrides the zero fluid density.

9. If I add Zone 1 myself, does the 'underground modeler' automatically calculate the Zone 2 lengths.

Hi, anyone able to respond to this? Richard Ay, your help would be very welcome.
Kind regards.

1) I'm assuming the protrusion you mention adds stiffness to the pipe. Therefore I would suggest your "option b".

2) The answer here depends on the insulation/cover used.

3) Yes you can just adjust the "B" model, BUT there potential problems with this method. If you make multiple changes to the "B" model, and then the routing must change, the only proper way to make that adjustment is to make the changes in the original model. There is a potential to "lose" changes and refinements made in the "B" model if a larger modification is then needed in the original.

4) Not really. You can try to do this graphically, but you may end up deleting non-soil restraints also.

5) The buried pipe modeler removes restraints from the "buried" elements before remeshing the system and computing the new soil restraints.

6) Unless you have a really sensitive system, and very loose soil, the Zone 1 setting shouldn't be necessary for flanges or reducers.

7) The buried pipe modeler will do this for you.

8) The pipeline codes are interested in the thermal stresses in the pipeline. The assumption is that there is adequate support for the pipe, hence there will be no weight stresses. This is why CAESAR II zeros the densities. The soil restraint stiffness values are computed assuming there is no weight.

9) Yes.

Many thanks Richard for concise reply.

Richard,

The mesh applied to the buried piping seems very dependent on the element lengths used. For example, a straight buried pipe of length 30m has a completely different mesh applied if the element lengths are consistent vs element lengths being varying lengths. This in turn produces different loads (say 0-50% change in my testing).

I did not know the buried piping was so sensitive to the element lengths.

How can this be explained?

Thanks

See if this helps. There are a number of slides in this (2013) presentation that address your question(s).