Meshing thin and extremely long insulated pipe

Y27 · September 16, 2020, 18:48

Hi, I wish you all well

What method of meshing is suitable for a geothermal well which is, for example;

a circular, insulated pipe which is buried starting from the earths surface unto 1000 m deep (length), it has a diameter of 0.3 m, and both pipe and insulation possess a wall thickness of 0.01 m. The computational domain (earth/rock) would be slightly larger than 1 km^3 .

I would obviously like to capture the thermal interactions between the solid-solid interfaces as well as the solid-fluid. I may be incorrect, but I think that the thin mesher would be suited to the thin pipe, and then i should use the polygonal mesher for the large domain and fluid.

I guess the question I am really asking is how do I scale the mesh appropriately in the axial and radial directions so that I capture near wall effects without creating a mesh which is extremely dense in the axial direction and thus computationally taxing? I am currently working in 2D but will eventually change to 3D.

Is it possible to do this in both 2D and 3D?

Should I use the automated mesher?

Should i use and unconformal mesh?

Is STAR-CCM+ designed to handle this type of simulation, if so,

What system of controls can be used to stretch a mesh which is less than 1 cm thick over a distance of 1 km?

Any help is much appreciated, and apologies for any ignorance displayed in my message, as usual, Thank you!

cwl · September 17, 2020, 07:57

Quote:

I may be incorrect, but I think that the thin mesher would be suited to the thin pipe, and then i should use the polygonal mesher for the large domain and fluid.

I'd say - go with Directed Mesher, for a long pipe swept mesh would be the best.
As for the pipe and insluation: instead of meshing - 3D conducting shell might be of use.

Quote:

Is it possible to do this in both 2D and 3D?

You could do it in axisymmetric approach, unless boundary conditions in earth/rock are strongly non-axisymmetric.

ping · September 17, 2020, 08:16

star has a special extruder meshing which is ideal for long pipes - read about it here in the help: Pre-Processing > Meshing > Volume Meshers > Generalized Cylinder Mesher

as i've tried to explain to you or someone else in a previous post recently, for many solid thin and some thick walled pipes with or without insulation and multilayers these can be vastly simplified and speed up solving by modelling the pipe detail solids as a thermal barrier with the correct thermal resistance and so ignore thin meshing etc.

use the mapped contact interface to connect the fluid to the earth because the earth will probably have a different mesh and so the interface is non-conformal.

but even with this simplification this case will be challenging so it would help to know exactly what the aim of the analysis is.

consider doing the whole case in 2d using the axisymmetric tool and i would do the first full setup and runs this way - unless there is a real reason for having 3d earth. the mesh size and solve might be manageable this way.

you might be able combine a 2d axisymmetric fluid pipe with a 3d earth using the mixing plane interface but it is not what it is designed for and it does not have a thermal resistance property to all for thermal resistance in the interface as suggested above.

September 16, 2020, 18:48	Meshing thin and extremely long insulated pipe	#1
Y27 New Member Yunus Aysan Join Date: Aug 2020 Posts: 12 Rep Power: 6	Hi, I wish you all well What method of meshing is suitable for a geothermal well which is, for example; a circular, insulated pipe which is buried starting from the earths surface unto 1000 m deep (length), it has a diameter of 0.3 m, and both pipe and insulation possess a wall thickness of 0.01 m. The computational domain (earth/rock) would be slightly larger than 1 km^3 . I would obviously like to capture the thermal interactions between the solid-solid interfaces as well as the solid-fluid. I may be incorrect, but I think that the thin mesher would be suited to the thin pipe, and then i should use the polygonal mesher for the large domain and fluid. I guess the question I am really asking is how do I scale the mesh appropriately in the axial and radial directions so that I capture near wall effects without creating a mesh which is extremely dense in the axial direction and thus computationally taxing? I am currently working in 2D but will eventually change to 3D. Is it possible to do this in both 2D and 3D? Should I use the automated mesher? Should i use and unconformal mesh? Is STAR-CCM+ designed to handle this type of simulation, if so, What system of controls can be used to stretch a mesh which is less than 1 cm thick over a distance of 1 km? Any help is much appreciated, and apologies for any ignorance displayed in my message, as usual, Thank you!

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problem with meshing long, thin faces in CFX	Martin	CFX	3	January 8, 2009 21:51

September 17, 2020, 08:16		#3
ping Senior Member Ping Join Date: Mar 2009 Posts: 556 Rep Power: 20	star has a special extruder meshing which is ideal for long pipes - read about it here in the help: Pre-Processing > Meshing > Volume Meshers > Generalized Cylinder Mesher as i've tried to explain to you or someone else in a previous post recently, for many solid thin and some thick walled pipes with or without insulation and multilayers these can be vastly simplified and speed up solving by modelling the pipe detail solids as a thermal barrier with the correct thermal resistance and so ignore thin meshing etc. use the mapped contact interface to connect the fluid to the earth because the earth will probably have a different mesh and so the interface is non-conformal. but even with this simplification this case will be challenging so it would help to know exactly what the aim of the analysis is. consider doing the whole case in 2d using the axisymmetric tool and i would do the first full setup and runs this way - unless there is a real reason for having 3d earth. the mesh size and solve might be manageable this way. you might be able combine a 2d axisymmetric fluid pipe with a 3d earth using the mixing plane interface but it is not what it is designed for and it does not have a thermal resistance property to all for thermal resistance in the interface as suggested above.