Hi!

i have a solid, and i would like to define an initial temperature profile for cooling calculation in x direction (a polynom) - in y direction, the temp. should be constant.

I have created follwing UDF, but somehow, it does not work...

Maybe there is something with the units? my mesh is created in mm. the solid is 100 mm long and there should be a temp. drop at the end to 960K.

Any idea? Ralf

#include "udf.h"

DEFINE_PROFILE(start_temperature,thread,index)

{

real y[ND_ND];

real x;

face_t f;

begin_f_loop(f,thread)

{

F_CENTROID(y,f,thread);

x = y[1];

F_PROFILE(f,thread,index) = 277103.*x*x*x -64574.*x*x + 2416.*x + 1099.;

} end_f_loop(f,thread)

}

I guess there should be should be 0 instead of 1. I mean this: x = y[0];

yes!! that's it!! now, it works!!

thanks a lot!

For me, without c+ knowledge, it is rather difficult to understand these udf, also the udf manual is not very helpful for these problems...

Ralf

I think u must do 2 UDF´s and not only one. One for x and other for y. Then, u can input the direction of your temperature field, as u do with velocity components.

Why did u use "x = y[1];"?

Hi,

the udf works for my 2d. model. Attached for completeness the udf.

T = x*x*x*277103. -x*x*64574. + x*2416. + 1099. is the polynome, that describes the Temp. distribution (x in m !!!)

Thanks you for your help!!!

Ralf

-----------------------

#include "udf.h"

DEFINE_PROFILE(start_temperature,thread,index)

{

real y[ND_ND];

real x;

cell_t c;

begin_c_loop(c,thread)

{

C_CENTROID(y,c,thread);

x = y[0];

F_PROFILE(c,thread,index) = x*x*x*277103. -x*x*64574. + x*2416. + 1099.;

}

end_c_loop(c,thread)

}

Ralf, your UDF "DEFINE_PROFILE(start_temperature,thread,index ) "for initialize temperature,not for boundary condition? lucy