CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   FLUENT (https://www.cfd-online.com/Forums/fluent/)
-   -   Using UDF to define boundary conditions (https://www.cfd-online.com/Forums/fluent/36452-using-udf-define-boundary-conditions.html)

Forest April 21, 2005 17:24

Using UDF to define boundary conditions
 
Hi all, I was wondering how Fluent defines Neumann boundary condition(first derivative is zero)? I am thinking it should be done by UDF, is there an example in Fluent help? thanks a lot. best,F

P April 22, 2005 03:26

Re: Using UDF to define boundary conditions
 
Could you please explain more for which equation you want newmann boundary condition?

Forest April 22, 2005 11:30

Re: Using UDF to define boundary conditions
 
Actually I intended to set 3 comp of velocities's 1st derivatives as zero at a velocity_outlet boundary. I tried to use "DEFINE_ADJUST" to take out the velocities at the cells next to the boundary cells and use "PROFILE_DEFINE" to set the boundary velocities same as those I took. Good news was it worked! the derivatives were almost zero. Bad news was that results were not convergent due to continuity residue. Probabley my model was not that good. Any better way to do the Neumann BC for velocities? thanks!

Saad April 22, 2005 17:58

Re: Using UDF to define boundary conditions
 
If you are using the Neumann B.Cs to simulate a fully developed flow then you should know that a fully developed flow boundary condition (which means all normal derivatives of all orders are zero at a boundary) is not mathematically consistent with the N-S equations. You should use the 'Outflow' Boundary Condition in the FLUENT instead of writing a udf, which in addition to putting the normal derivatives equal to zero does some other things to ensure a consistent solution.

Sincerely,

Saad


All times are GMT -4. The time now is 13:11.