[ravisalgar]

hello friends!!!

can any body help me in writing an UDF for the below boundary conditions,as m very new to fluent and that to i never come across writing an UDF while using fluent till this time.

my boundary conditions are as below :

1) dT/dy = 0
where T stands for Temperature and y is the direction of flow.

2) u = o

where u is the x component of velocity,

3) v not equal to zero.

i mean v component of velocity is not zero.

help will me very much appreciated

[ali hemmati]

Hi,

First of all, you should take a glance at Fluent 'UDF Manual'. There is a great number of examples for assigning parameters.

I tell you some key words: DEFINE_PROFILE , thread_loop_c , begin_c_loop , C_U , F_U , C_V , F_V , C_T , F_T.

Here is an example from manual, which I hope help you:

DEFINE_PROFILE(pressure_profile,t,i)
{
real  x[ND_ND]; /*  this  will  hold  the  position  vector  */
real  y;
face_t  f;
begin_f_loop(f,t)
{
F_CENTROID(x,f,t);
y  =  x[1];
F_PROFILE(f,t,i)  =  1.1e5  -  y*y/(.0745*.0745)*0.1e5;
}
end_f_loop(f,t)
}

DEFINE_PROFILE(fixed_v,t,i)
{
cell_t  c;
real  x[ND_ND];
real  r;
begin_c_loop(c,t)
{
/*  centroid  is  defined  to  specify  position  dependent  profiles*/
C_CENTROID(x,c,t);
r  =x[1];
F_PROFILE(c,t,i)  =
va1+(va2*r)+(va3*r*r)+(va4*r*r*r)+(va5*r*r*r*r);
}
end_c_loop(c,t)
}

[SolarX]

Hi,

First of all your problem - at least your boundary conditions (BC's) - seems like you don't need UDFs since you can insert those BC's directly in the UserGraphicalInterface. Perhaps it is a matter of choosing proper BCs for your inlet, or making a more flexible geometry by splitting up your inlet in 2 boundary regions!?

Secondly, I think Ansys Fluent really needs a stronger manual for using UDF, for people who can't write C-scripts. It is very weak and jumps to complex code-writing while it doesn't explained well how to structure C-language script. They made already 522 pages of UDF-manual, why not a more comprehensive introduction into C first?! It should be an independent manual.

Anyway I hope you are on your way to solve this model.

Regards