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/)
-   -   Adding source term involving derivative in y direction (http://www.cfd-online.com/Forums/openfoam-programming-development/113594-adding-source-term-involving-derivative-y-direction.html)

iqbalsk8 February 22, 2013 13:51

Adding source term involving derivative in y direction
 
Hi everyone,

I want to add a new transport equation to simpleFoam.
The source term looks like Sw = nu/pow(y,2) * dw/dy
where y is the y-coordinate of the cell and
dw/dy is the derivative in the y-direction.

fvScalarMatrix UwEqn
(

fvm::div(phi, w)
- fvm::laplacian(nu, w)
- Sw
);

My confusion is to get the derivative in a one direction not to calculate the
whole gradient.

Can anybody give me an idea how to achieve this.

regards,

Sohail

anishtain4 February 22, 2013 15:10

If i'm correct you may get the gradient only in the explicit way which would be:
vectorField gradU=fvc::grad(w)
then you can extract the desired component by:
scalarField dwdy=gradU.component(vector::Y)
I may not be correct in spelling or own the best way, but this should work

iqbalsk8 February 24, 2013 09:31

Dear Hosseinali,

Thankyou for giving time to answer. I also thought the same and
it works fine.
But now I have another problem. I have a term
Sw = (Uy * w)/y

where Uy is the y component of velocity
and y coordinate.

Can this term be put in UwEqn as,

fvm::Sp(Uy/y, w)

where Uy=U.component(vector::Y)

Best regards,

Sohail

anishtain4 February 25, 2013 05:20

Just from the top of my head, when you are using the SIMPLE every equation is being solved separately, also in source term of your equation for UwEqn there is no Uw (Uy is being considered known from previous step) so I think there would be no difference if you use fvm or mention it explicitly

iqbalsk8 February 25, 2013 10:30

1 Attachment(s)
Dear Hosseinali,

I tried to put the term

Sw = (Uy * w)/y

with fvm::Sp and explicitely. but in both cases the value of Sw tends to infinity
after 500 iterations approximately.

I am using OpenFOAM-1.6-ext. I am simulating a 2D axisymmetric case.
Where this new transport equation calculates the swirl-component
of velocity.

Please have a look at the file.
If you want the whole case file, I can also send you that...

regards,

Sohail

anishtain4 March 2, 2013 11:39

I'm not sure how this affects your stability, All I'm saying is that with the solver you chose this term would have same effect explicitly or implicitly on the solution.


All times are GMT -4. The time now is 23:28.