can OpenFoam model a pole in Cylindrical mesh??

Dan1788 · October 26, 2014, 20:03

I am making a 3D cylindrical mesh for LES computations of premixed flames.

I would like to know if OpenFoam can correctly model singularities or poles at the origin (as shown in the attached figure) correctly, or would I have to mesh my geometry in a different manner.

Would really appreciate any help on this topic. Thanks!

ChristoBeers · October 27, 2014, 16:17

I had previously tested this with rhoPimpleFoam, and did notice abnormal action along the cylinder axis (center of the "pie slices") where the cell skewness gets really large. I then tried rounding out one square hex, but this introduced abnormalities where the corners were stretched out (probably for the same reason). I then meshed a square in the center of the circle which simulated properly.

I honestly didn't spend too much effort trying to see if there were more advanced solver options to account for skewedness, I just changed the mesh. Somebody else with more experience may speak to your exact question.

Good luck!

RodriguezFatz · October 28, 2014, 04:05

I am wondering: why..? I mean, people use o-grids for this normally. I didn't really question this up to now, because I thought this is pretty common. Can you explain, why you prefer this way of meshing?

ChristoBeers · October 28, 2014, 12:11

You are absolutely right, people use o-grids. It may have been my specific mesh/configuration/BC, but there were false temp/density iso's which protruded down the cylinder axis which should not have been the case. I was merely warning the OP that it is possible if he's not careful. As I said, I didn't play with it too much because it was simple enough for me to change the grid.

To keep an o-grid in my system I had calculated an excessively dense mesh in the center to keep the skewness down (increasing #/meter of "onion shells" around the axis where the "pie slice" tips narrow to zero @ the center). I have very little experience in CFD, but in my simulation, the denser mesh & running >3 nonOrthogonalCorrectors will slow me down, and I guessed a nice cubic in the center may help the solver converge more rapidly...? I wanted to spend the mesh density near the walls where there's something more interesting going on. Am I way off?

Dan1788 · October 28, 2014, 13:14

Even I am modeling my mesh using an O-grid but was wondering if OpenFoam could model poles because for an axisymmetric grid it is able to take into account pole formation.

Anyways thanks for your reply Chris!

chun · May 30, 2017, 22:03

Although it might be late for you, but I do have what you want.

You can change the length and radius from the .m4 file.
Using the following command, you can get the blockMeshDict from the .m4 file.

Code:

m4 circle.m4 > blockMeshDict

October 28, 2014, 04:05		#3
RodriguezFatz Senior Member Philipp Join Date: Jun 2011 Location: Germany Posts: 1,297 Rep Power: 27	I am wondering: why..? I mean, people use o-grids for this normally. I didn't really question this up to now, because I thought this is pretty common. Can you explain, why you prefer this way of meshing? __________________ The skeleton ran out of shampoo in the shower.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[snappyHexMesh] No layers in a small gap	bobburnquist	OpenFOAM Meshing & Mesh Conversion	6	August 26, 2015 09:38
Frozen Rotor 1:1 Mesh Connection	pharley	CFX	5	January 31, 2013 16:15
New OpenFOAM Forum Structure	jola	OpenFOAM	2	October 19, 2011 06:55
Converting Starccm+ mesh	Ladnam	OpenFOAM	0	September 14, 2011 06:30
salome, openfoam and moving mesh	prhlava	OpenFOAM Running, Solving & CFD	8	November 9, 2009 08:59

October 27, 2014, 16:17		#2
ChristoBeers New Member Chris Sears Join Date: Oct 2014 Location: Boston Posts: 4 Rep Power: 12	I had previously tested this with rhoPimpleFoam, and did notice abnormal action along the cylinder axis (center of the "pie slices") where the cell skewness gets really large. I then tried rounding out one square hex, but this introduced abnormalities where the corners were stretched out (probably for the same reason). I then meshed a square in the center of the circle which simulated properly. I honestly didn't spend too much effort trying to see if there were more advanced solver options to account for skewedness, I just changed the mesh. Somebody else with more experience may speak to your exact question. Good luck!

October 28, 2014, 12:11		#4
ChristoBeers New Member Chris Sears Join Date: Oct 2014 Location: Boston Posts: 4 Rep Power: 12	You are absolutely right, people use o-grids. It may have been my specific mesh/configuration/BC, but there were false temp/density iso's which protruded down the cylinder axis which should not have been the case. I was merely warning the OP that it is possible if he's not careful. As I said, I didn't play with it too much because it was simple enough for me to change the grid. To keep an o-grid in my system I had calculated an excessively dense mesh in the center to keep the skewness down (increasing #/meter of "onion shells" around the axis where the "pie slice" tips narrow to zero @ the center). I have very little experience in CFD, but in my simulation, the denser mesh & running >3 nonOrthogonalCorrectors will slow me down, and I guessed a nice cubic in the center may help the solver converge more rapidly...? I wanted to spend the mesh density near the walls where there's something more interesting going on. Am I way off?

October 28, 2014, 13:14		#5
Dan1788 Member Daniel Join Date: Jun 2014 Posts: 60 Rep Power: 12	Even I am modeling my mesh using an O-grid but was wondering if OpenFoam could model poles because for an axisymmetric grid it is able to take into account pole formation. Anyways thanks for your reply Chris!