CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   UDF Data access problems (http://www.cfd-online.com/Forums/fluent/89581-udf-data-access-problems.html)

Mosheraa June 16, 2011 18:45

UDF Data access problems
 
I'm fairly new to Fluent, and as part of teaching myself the UDF structure I made a small UDF to read area and flow at the boundary.

However, when I printed the data to the screen it gave me nothing but nonsense values. Also, when I directly printed the "M_PI" function, the output was somewhere above 1.6 trillion.

Has anyone encountered similar problems, or have any thoughts about what might be going wrong?

For reference, here is the small UDF I wrote. I applied it as a pressure outlet boundary condition for laminar flow in an axi-symmetric pipe.

/************************************************** **************************************
Test function to debug Fluent/UDF operations
************************************************** **************************************/

#include "udf.h"


/*Test Axisymmetric Floow and Area*/


DEFINE_PROFILE( area_test, t, i )
{
/*Variables*/

real r = 0;
real ri[ND_ND];
real q = 0;

face_t f;


/*Loop Over Faces*/

begin_f_loop( f, t )
{
q += 2*M_PI*F_FLUX( f, t );

F_AREA(ri,f,t);
r += NV_MAG(ri);

F_PROFILE(f,t,i) = 0;

printf( "Inside Loop area: %d\n", F_AREA(ri,f,t) );
printf( "INside LOop flux: %d\n", F_FLUX(f,t) );
}
end_f_loop( f, t )


/*Print out data*/

printf( "Outlet Area is: %d\n", r );
printf( "Outlet flow is: %d\n", q );
printf( "Fluent thinks pi is %d\n", M_PI);
}

coglione June 17, 2011 03:37

printf("%d",x) outputs data as integer numbers. Your variables are defined as real = floating numbers! The correct format is %g.

cheers

p4lm0r3 June 17, 2011 09:52

One thing to add. I have had problems with UDFs that use FLEUNT builtin functions. Many of the values are not defined until after you initialize your solution. Make sure you initialize your solution before hooking in your UDF. Not sure if this is your error, but if you do this it will help you out in the future.

Mosheraa June 17, 2011 11:35

Issue Resolved
 
Thanks everyone for the suggestions, switching the printf statements from %d to %g fixed the issue. I thought the output values looked like min or max integer values, and I was very confused.

As a separate question, does anyone have any suggestions for tracking and updating a user defined variable between iterations? I'm trying to model an autoregulation response in the arteries and then adjust a pressure boundary condition accordingly. The closest thing I have found so far is the F_UDMI function.

~Aaron

pranab_jha June 17, 2011 13:16

Quote:

Originally Posted by Mosheraa (Post 312492)
As a separate question, does anyone have any suggestions for tracking and updating a user defined variable between iterations? I'm trying to model an autoregulation response in the arteries and then adjust a pressure boundary condition accordingly. The closest thing I have found so far is the F_UDMI function.

~Aaron

Yes, you can use the F_UDMI macro. I have done it.
You can use the value stored at the memory location in the next time step (supposing you are doing this way) and then store the new value back in the memory.
Always, start with a steady state solver, read in the UDF and then run the steady state solver for 1/2 iterations atleast, even if you dont want a steady solution at that instant. Then load the UDF on your boundary or wherever you want it. Then switch to unsteady mode and start your calculation.
This method worked for me. Earlier I used to start my job in batch mode right after reading in the UDFs and used to get memory errors.


All times are GMT -4. The time now is 22:50.