Significant change in results with changing either mesh size or time step size.

nettpanic · October 19, 2014, 05:42

Dear All,

I need your help. I am trying to simulate the adsorption process of CO2 in a tank filled with activated carbon (adsorbent). This is a transient case. I have successfully simulated this process, but the problem is with changing the mesh size results change significantly, similarly the same thing happens when I change the time step size. Please need your suggestions to to get out of this situation. Axisymmetric geometric model of the storage system is attached.

Regards,

CeesH · October 21, 2014, 02:01

My suggestion: decrease your mesh and timestep size to a size where they do not significantly influence the solution. Clearly, your results are not grid/stepsize independent. What else do you think you could do about that?

nettpanic · October 21, 2014, 02:12

Dear Ceesh
Thanks for your valuable suggestion. I have tried many different settings including grid refinement. Initially My mesh comprises of 12240 cells. I have refined my mesh till 92000. But facing same problem. I have tried different pressure-velocity coupling schemes, and also change the spatial discretization to second order upwind. But of no use. I haven't tried coarse mesh below 12240 cells.

CeesH · October 21, 2014, 03:17

Hi,

Some questions: Is the case 2 or 3-D?

Do your residuals converge (and do other monitors you may be using converge) during a simulation?

If you plot your results versus mesh density or timestep size, do you see any trend?

nettpanic · October 21, 2014, 13:12

Dear Ceesh

Here are my observations:

1) It is a 2D case

2) Residuals are converging well

3) I am monitoring Temperature at different points within the storage system, once I plot these temperature agains mesh density and time step size follwoing trend has been observed;

a) Keeping time step size constant (t=3sec), increasing the mesh density from 12240 cells to 91000 cells, temperature at the center point of the storage tank decreases from 350K to 318K.

b) Keeping the mesh density constant (12240 cells) increasing the time step size decreases the temperature at above mentioned location from 350K to 325K.

CeesH · October 24, 2014, 02:37

What did you base the timestep size on? It seems like it may be too high, causing the problems. A converging solution doesn't mean a mesh independent solution of course, so maybe you simply need to reduce the time and mesh size until you do find a solution which is independent of those parameters..

October 21, 2014, 02:01		#2
CeesH Senior Member Cees Haringa Join Date: May 2013 Location: Delft Posts: 607 Rep Power: 0	My suggestion: decrease your mesh and timestep size to a size where they do not significantly influence the solution. Clearly, your results are not grid/stepsize independent. What else do you think you could do about that? nettpanic likes this.

October 21, 2014, 03:17		#4
CeesH Senior Member Cees Haringa Join Date: May 2013 Location: Delft Posts: 607 Rep Power: 0	Hi, Some questions: Is the case 2 or 3-D? Do your residuals converge (and do other monitors you may be using converge) during a simulation? If you plot your results versus mesh density or timestep size, do you see any trend? nettpanic likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
simpleFoam error - "Floating point exception"	mbcx4jc2	OpenFOAM Running, Solving & CFD	12	August 4, 2015 02:20
AMI interDyMFoam for mixer nu problem	danny123	OpenFOAM Programming & Development	8	September 6, 2013 02:34
same geometry,structured and unstructured mesh,different behaviour.	sharonyue	OpenFOAM Running, Solving & CFD	13	January 2, 2013 22:40
pisoFoam with k-epsilon turb blows up - Some questions	Heroic	OpenFOAM Running, Solving & CFD	26	December 17, 2012 03:34
fluent add additional zones for the mesh file	SSL	FLUENT	2	January 26, 2008 11:55

October 21, 2014, 02:12		#3
nettpanic New Member Ali Join Date: Oct 2014 Posts: 4 Rep Power: 11	Dear Ceesh Thanks for your valuable suggestion. I have tried many different settings including grid refinement. Initially My mesh comprises of 12240 cells. I have refined my mesh till 92000. But facing same problem. I have tried different pressure-velocity coupling schemes, and also change the spatial discretization to second order upwind. But of no use. I haven't tried coarse mesh below 12240 cells.

October 21, 2014, 13:12		#5
nettpanic New Member Ali Join Date: Oct 2014 Posts: 4 Rep Power: 11	Dear Ceesh Here are my observations: 1) It is a 2D case 2) Residuals are converging well 3) I am monitoring Temperature at different points within the storage system, once I plot these temperature agains mesh density and time step size follwoing trend has been observed; a) Keeping time step size constant (t=3sec), increasing the mesh density from 12240 cells to 91000 cells, temperature at the center point of the storage tank decreases from 350K to 318K. b) Keeping the mesh density constant (12240 cells) increasing the time step size decreases the temperature at above mentioned location from 350K to 325K.

October 24, 2014, 02:37		#6
CeesH Senior Member Cees Haringa Join Date: May 2013 Location: Delft Posts: 607 Rep Power: 0	What did you base the timestep size on? It seems like it may be too high, causing the problems. A converging solution doesn't mean a mesh independent solution of course, so maybe you simply need to reduce the time and mesh size until you do find a solution which is independent of those parameters..