Grid independence test?

MStudent · July 10, 2015, 04:51

I've seen some threads about this, but none of them actually explain how to do it...

The problem I'm working with is a double pipe counterflow heat exchanger with a laminar flow of gas through the inner pipe and water through the annulus. The model is SST k-w. I'm using workbench 15.0

I know I'm supposed to start from a lower cell count and then make the count higher till there is only a low percentage of difference between the "solutions". However I can't find how to change the mesh so that it keeps the same layout but with more cells...

Do I just use the Course/Medium/Fine options or is there another way, since this doesn't really change the amount of cells (stays around 1M).
Course mesh:

Nodes: 352021
Elements: 979392

How do I change the mesh to do the test?

Also, how do I compare the value of the "solution"? Can I just use the 'stable' value from a surface monitor?
So far I've been using 'convergence history of static temperature on Gas_Out' But this doesn't give an actual temperature so...

http://i12.photobucket.com/albums/a2...psdbzuvk1j.png
http://i12.photobucket.com/albums/a2...ps8r2w4klj.png
http://i12.photobucket.com/albums/a2...psuagl7u6g.png
http://i12.photobucket.com/albums/a2...psiv0ecwnx.png
http://i12.photobucket.com/albums/a2...psuaafxwbv.png

mehrzad · July 10, 2015, 15:34

The general way is in the meshing software to reduce the size of element to the desired value. a rule of thumb say that you reduce the element size in a way that tha number of cells increases with factor of 1.5 .
for checking the convergence of solution you should check multiple references. beside checking the residuals, compare velocity, temperature profiles.
I recommend you to read this article about the convergence of simulation.
http://support.esi-cfd.com/esi-users/convergence/

MStudent · July 11, 2015, 12:00

Thank you for your reply and the informative link about convergence, however I'm still having trouble with 'how' to get the number of cells to increase...
This is the sizing tab for the meshing in Fluent:

I realize that this is probably a dumb question, but how am I supposed to adjust these in order to get the cell count to increase?

Also, as can be seen in the previously posted images for the 'convergence history of static temperature' there is simply no change during any of the iterations..
Am i simply choosing the wrong type of variables to look at for the monitor or is there another problem?

-Thanks

July 10, 2015, 04:51	Grid independence test?	#1
MStudent New Member Join Date: Apr 2015 Posts: 11 Rep Power: 11	I've seen some threads about this, but none of them actually explain how to do it... The problem I'm working with is a double pipe counterflow heat exchanger with a laminar flow of gas through the inner pipe and water through the annulus. The model is SST k-w. I'm using workbench 15.0 I know I'm supposed to start from a lower cell count and then make the count higher till there is only a low percentage of difference between the "solutions". However I can't find how to change the mesh so that it keeps the same layout but with more cells... Do I just use the Course/Medium/Fine options or is there another way, since this doesn't really change the amount of cells (stays around 1M). Course mesh: Nodes: 352021 Elements: 979392 How do I change the mesh to do the test? Also, how do I compare the value of the "solution"? Can I just use the 'stable' value from a surface monitor? So far I've been using 'convergence history of static temperature on Gas_Out' But this doesn't give an actual temperature so... http://i12.photobucket.com/albums/a2...psdbzuvk1j.png http://i12.photobucket.com/albums/a2...ps8r2w4klj.png http://i12.photobucket.com/albums/a2...psuagl7u6g.png http://i12.photobucket.com/albums/a2...psiv0ecwnx.png http://i12.photobucket.com/albums/a2...psuaafxwbv.png

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
critical error during installation of openfoam	Fabio88	OpenFOAM Installation	21	June 2, 2010 04:01
OF 1.6 \| Ubuntu 9.10 (64bit) \| GLIBCXX_3.4.11 not found	piprus	OpenFOAM Installation	22	February 25, 2010 14:43
Problems in compiling paraview in Suse 10.3 platform	chiven	OpenFOAM Installation	3	December 1, 2009 08:21
Grid independence check	Mattis Voss	FLUENT	0	November 23, 2005 08:31
Grid independence with k-eps model	Mark Render	Main CFD Forum	3	February 21, 2002 10:54

July 10, 2015, 15:34		#2
mehrzad New Member mehrzad Join Date: Jul 2011 Posts: 23 Rep Power: 15	The general way is in the meshing software to reduce the size of element to the desired value. a rule of thumb say that you reduce the element size in a way that tha number of cells increases with factor of 1.5 . for checking the convergence of solution you should check multiple references. beside checking the residuals, compare velocity, temperature profiles. I recommend you to read this article about the convergence of simulation. http://support.esi-cfd.com/esi-users/convergence/

July 11, 2015, 12:00		#3
MStudent New Member Join Date: Apr 2015 Posts: 11 Rep Power: 11	Thank you for your reply and the informative link about convergence, however I'm still having trouble with 'how' to get the number of cells to increase... This is the sizing tab for the meshing in Fluent: I realize that this is probably a dumb question, but how am I supposed to adjust these in order to get the cell count to increase? Also, as can be seen in the previously posted images for the 'convergence history of static temperature' there is simply no change during any of the iterations.. Am i simply choosing the wrong type of variables to look at for the monitor or is there another problem? -Thanks