CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Programming & Development (http://www.cfd-online.com/Forums/openfoam-programming-development/)
-   -   Non-dimensionzalized icoFoam (http://www.cfd-online.com/Forums/openfoam-programming-development/118132-non-dimensionzalized-icofoam.html)

umar82088 May 21, 2013 11:00

Non-dimensionzalized icoFoam
 
1 Attachment(s)
Hi All,
I am trying to perform non-dimensionalization of icoFoam. Please have a look of attached file which is modification which I am having now.
I have turned-off dimensional consistency check as solver always complains about it. Calculations are performed but in my view they are not physically meaningful.
1 ) I am wondering whether solver is not itself not correct, having some mistake??
2) Non-dimensionaliztion of geometry is also needed, as far as I know, then how it could be implemented??
Any help regarding this, would be highly appreciated.
Many Thanks
Umar

Cyp May 21, 2013 16:58

Quote:

Originally Posted by umar82088 (Post 428931)
Hi All,
I am trying to perform non-dimensionalization of icoFoam. Please have a look of attached file which is modification which I am having now.
I have turned-off dimensional consistency check as solver always complains about it. Calculations are performed but in my view they are not physically meaningful.
1 ) I am wondering whether solver is not itself not correct, having some mistake??
2) Non-dimensionaliztion of geometry is also needed, as far as I know, then how it could be implemented??
Any help regarding this, would be highly appreciated.
Many Thanks
Umar


Hello Umar,

I had a glance at your enclosed file. First of all, you can not get rid of the temporal term. Indeed, you attempt to modify a PISO loop and this latter is created to solve an unsteady problem ! If you really need a steady solver, have a look at simpleFoam.

Second point, instead of turning-off the dimensions check, I will have defined new dimensionedScalar (for example unit1) that is equal to 1 and have suitable dimensions and write : unit1*fvm::div(phi,U).

Regards,
Cyp

Tushar@cfd May 22, 2013 05:12

Quote:

Originally Posted by umar82088 (Post 428931)
Hi All,
I am trying to perform non-dimensionalization of icoFoam. Please have a look of attached file which is modification which I am having now.
I have turned-off dimensional consistency check as solver always complains about it. Calculations are performed but in my view they are not physically meaningful.
1 ) I am wondering whether solver is not itself not correct, having some mistake??
2) Non-dimensionaliztion of geometry is also needed, as far as I know, then how it could be implemented??
Any help regarding this, would be highly appreciated.
Many Thanks
Umar

Hi Umar,

I would really like to appreciate you for working with non-dimensional form of equation.
I totally agree with Cyp, it appears that you are not using the transient term in your equation. So, try with simpleFoam.

I would rather write the equation like this...

(
fvm::div(phi, U)
- fvm::laplacian(1/Re, U)
);

solve (UEqn == -fvc::grad(p));

Anyways, you need to be careful about the extra dimensions, see the weblink for better explanation:

http://www.cfd-online.com/Forums/ope...variables.html

umar82088 May 22, 2013 13:25

Hi Tushar,
First of all thanks for replying on the post. I made a mistake that I commented over the temporal term, which I didn't intend to do. My Problem needs to be solved in transient way. I have modified the code as you mentioned, now problem is arising with the discritization part. As we have to modify that part accordingly. Any suggestion to trick with that issue. And secondly how to non-dimensionalize geometry??

Thanks
Umar
Quote:

Originally Posted by Tushar@cfd (Post 429137)
Hi Umar,

I would really like to appreciate you for working with non-dimensional form of equation.
I totally agree with Cyp, it appears that you are not using the transient term in your equation. So, try with simpleFoam.

I would rather write the equation like this...

(
fvm::div(phi, U)
- fvm::laplacian(1/Re, U)
);

solve (UEqn == -fvc::grad(p));

Anyways, you need to be careful about the extra dimensions, see the weblink for better explanation:

http://www.cfd-online.com/Forums/ope...variables.html


Tushar@cfd May 23, 2013 00:57

Quote:

Originally Posted by umar82088 (Post 429314)
Hi Tushar,
First of all thanks for replying on the post. I made a mistake that I commented over the temporal term, which I didn't intend to do. My Problem needs to be solved in transient way. I have modified the code as you mentioned, now problem is arising with the discritization part. As we have to modify that part accordingly. Any suggestion to trick with that issue. And secondly how to non-dimensionalize geometry??

Thanks
Umar

Hi,

I think you need to follow either of the procedure described on the forum:

(1) http://www.cfd-online.com/Forums/ope...variables.html
(2) http://www.cfd-online.com/Forums/ope...-mesh-etc.html

I haven't tried any of these but it seems both the procedure will do the job.
:)

Mehrez June 5, 2013 12:12

Hi Umar

Did you resolve your problem ?

In order to non-dimensionalize your geometry you need to divide all the dimensions by a characteristic length that you will choose and this characteristic length will appear in your Re number.
Then, you can enter this dimensions in your OpenFoam file (MeshDict).


Mehrez


All times are GMT -4. The time now is 19:31.