The coarser the mesh, the better the results?

CellZone · September 12, 2016, 06:50

Hi,

I made some simulations in OpenFoam CFD and I am wondering: the coarser my mesh is, the better the results match the ones of the experiment. For the coarse mesh the residuum is also better, it shows a better convergence. Does this mean my mesh is a bad one?

Thank you for your help

CellZone

KateEisenhower · September 12, 2016, 11:16

Hi,

I am only making a guess here since you gave very little information. Are you using wall functions? Can you post a picture of your mesh?

Best regards,

Kate

cymbourne · September 15, 2016, 09:50

Can you give us more information on your case? RANS? URANS? What are you changing between both grids ?

Workshop Technologies · October 21, 2016, 10:48

I think your doubt is right. It is a case of bad mesh.

We can help you better if you can provide a brief description of your case. You can send enquiries to http://workshopwale.com/ or email us at: adarsh@workshopwale.com

Regards,
Workshopwale Team!!
KBAT Technologies

Quote:

Originally Posted by CellZone

Hi,

I made some simulations in OpenFoam CFD and I am wondering: the coarser my mesh is, the better the results match the ones of the experiment. For the coarse mesh the residuum is also better, it shows a better convergence. Does this mean my mesh is a bad one?

Thank you for your help

CellZone

monkeymax · October 25, 2016, 05:19

Definitely due to a poor quality mesh.
As the mesh gets coarser you are solving a more generic form of the problem and any issues with the mesh are effectively dampened/dissipated out.
As you refine the mesh the problem cells become more evident and the solution is more affected by these low quality cells.

It's most likely luck that the coarser mesh is giving results closer to experiment (well, simply that you're solving a more general form of the problem).

Also check boundary conditions and initial conditions as you refine the mesh. Perform a mesh independence study and examine what happens numerically at the cells where greatest changes occur (i.e. look at gradients).

Also bear in mind that maybe even your best solution in CFD won't match experiment that well - there could be differences between solution and experiment that will always mean an offset between the results.

September 12, 2016, 06:50	The coarser the mesh, the better the results?	#1
CellZone Member Join Date: Apr 2016 Posts: 90 Rep Power: 10	Hi, I made some simulations in OpenFoam CFD and I am wondering: the coarser my mesh is, the better the results match the ones of the experiment. For the coarse mesh the residuum is also better, it shows a better convergence. Does this mean my mesh is a bad one? Thank you for your help CellZone

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
decomposePar problem: Cell 0contains face labels out of range	vaina74	OpenFOAM Pre-Processing	37	July 20, 2020 05:38
[snappyHexMesh] SnappyHexMesh no layers and no decent mesh for complex geometry	pizzaspinate	OpenFOAM Meshing & Mesh Conversion	1	February 25, 2015 07:05
[Gmsh] 2D Mesh Generation Tutorial for GMSH	aeroslacker	OpenFOAM Meshing & Mesh Conversion	12	January 19, 2012 03:52
[snappyHexMesh] snappyHexMesh won't work - zeros everywhere!	sc298	OpenFOAM Meshing & Mesh Conversion	2	March 27, 2011 21:11
engrid: Internal volume mesh becoming coarser during boundayr layer addition	Arnoldinho	OpenFOAM	1	January 22, 2011 04:31

September 12, 2016, 11:16		#2
KateEisenhower Senior Member Join Date: Mar 2015 Posts: 250 Rep Power: 12	Hi, I am only making a guess here since you gave very little information. Are you using wall functions? Can you post a picture of your mesh? Best regards, Kate

September 15, 2016, 09:50		#3
cymbourne New Member Marion Join Date: Apr 2013 Posts: 15 Rep Power: 13	Can you give us more information on your case? RANS? URANS? What are you changing between both grids ?

October 25, 2016, 05:19		#5
monkeymax New Member Max Join Date: Dec 2010 Posts: 4 Rep Power: 15	Definitely due to a poor quality mesh. As the mesh gets coarser you are solving a more generic form of the problem and any issues with the mesh are effectively dampened/dissipated out. As you refine the mesh the problem cells become more evident and the solution is more affected by these low quality cells. It's most likely luck that the coarser mesh is giving results closer to experiment (well, simply that you're solving a more general form of the problem). Also check boundary conditions and initial conditions as you refine the mesh. Perform a mesh independence study and examine what happens numerically at the cells where greatest changes occur (i.e. look at gradients). Also bear in mind that maybe even your best solution in CFD won't match experiment that well - there could be differences between solution and experiment that will always mean an offset between the results.