[chany]

Hi everybody,

I have a square geometry with imposed heat flux (1000 w/m²) at north and south faces and velocity inlet (25m/s, 300K) and outflow at est and west. I want to use UDFs to define periodic phenomenon. Actually, for every iteration, I want to take velocity and temperature profiles at outflow and imposed them to velocity inlet. But i don't know how to read velocity and temperature profiles for being able to manipulated them.

Thanks

[argeus]

I usually use this: /cell variables accessing for instance/

UDF for data access:

DEFINE_PROFILE(my_profile, thread, index)
{
real x[ND_ND];
real y;
cell_t c;
int ID = ?; //outflow ID
Thread *tf;

Domain *domain;
domain = Get_Domain(1);
tf = Lookup_Thread(domain,ID);

if (! Data_Valid_P())
return;

begin_c_loop(c, thread)
{
F_CENTROID(x, c, thread);
y = x[1];

C_U = C_T(c,tf); // example for "u" direction, check UDF manual

F_PROFILE(c, thread, index) = ?; //setting outflow
}
end_c_loop(c,thread)
}

DEFINE_PROFILE(my_velocity, thread, index) //profile for velocity inlet
{
.
.
.

[chany]

Thanks a lot argeus, I'll try this method, hope that it will solve my problem!

[chany]

Please, I want to let fluent do calculations without UDF, and after some iterations (100 iterations for example), fluent begins to use my UDF.
The program of "argeux" has wored very good for field velocity, but for temperature, I have some problems.

[argeus]

You can run the simulation without UDF, and the after a few iterations interpret UDF anyway, I guess. What kind of problems do you have with temperature? Using heat transfer you need to have a fine mesh on the walls. Use boundary layers.