Hi Bing

Have you tried to interpret/compile it in Fluent?, if yes did fluent give you any errors or warnings?.

Check also if you need to specify the corners in your model.

I already impose a variable temperature along the wall (x-axis) which varies linearly and the temperature reduce to 298 k after an hour.

The code is:

#include "udf.h"

DEFINE_PROFILE(temp, thread, position)

{

face_t f;

real x;

real t = CURRENT_TIME;

begin_f_loop(f, thread)

{

x = F_T(f,thread);

if (t

Hatef . Seyyedvalilu, do you need to define x? I mean if it is real, it must be then an integer, right? please correct me if I'm wrong, because I had this problem before and fluent gave an error until I put a value for real parameters.

One more thing, how would you know that your function is working good? I'm also using a UDF for a transient temperature BC on some walls of my model. I need to know if the function gives the same temperature profile that is in the code. would setting a surface monitoring help to get this profile? what do you think?

Regards

Hatef . Seyyedvalilu, thank you so much for the valuable example. May I ask that how should I fix the y and z values in the coding? In your coding, temperature is only related to x, but in my case, it is 3D-related. Hope to see your answer soon. Wish you a good day!

Oula Fatla, I am still trying to correct the coding, will let you know if I have it validated by Fluent. Thanks for the discussion! :)

Dear Oula Fatla , Hatef . Seyyedvalilu , thanks for the discussion! I successfully applied the UDF to my Fluent iteration. The coding is simple as follows, there is no need to assign y and z locations, because it is applied to a specific wall surface.

#include "udf.h"

DEFINE_PROFILE(temperature_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) = -0.0023*y*y*y*y+0.0331*y*y*y-0.1478*y*y-0.0779*y+296.17;

}

end_f_loop(f,t)

}

Another problem I encountered is that, I compiled this coding in Windows, while my Fluent is run in Linux system. Although I made the coding correctly, it showed errors in Linux and could not be read \ line 2: syntax error \. The problem is the coding compiled in Windows is not compatible with Linux, so I compiled again in Linux text. Finally, It is working now. Hope it helps, although it is a piece of cake for seniors. :)

Dear Bing Han , I'm glad that you have managed to solve your problem and thanks for sharing this valuable information.

Do you need to set up a surface monitoring to see if the function is working good?.

Regards

Oula Fatla , thanks for the response. I finished one run of computation, as I checked temperature of the wall where the BC was applied onto, unexpectedly, the temperature distribution was not like what the coding describes, instead, it is a kinda uniform distribution. But there was no error occurred during the computation. Would you please give me some hints for this? Tons of thanks in advance! Have a good day!

Kind regards

You can try Fluent 19.2, you no longer need udfs for doing that. You can specify your function in BC pannel as dependent on time or space or both at the same time. :p good luck

Hi Bing Han

I don't know I'm afraid, I also want to know the answer. If there is anything we need to set to monitor the temperature as the simulation is running and evaluate the performance of the udf, that would be good. I had a chat with a friend regarding this, he suggests running the fluent with everything off except the transient BC of interest. Another thing is to create a plane close to the wall and create a temp. contour for example to observe the performance of the UDF. If you have any information, please let me know. Good luck with your work.

Regards

Hi

I'm using a udf to set a transient boundary condition on a wall in my model. The temperature suppose to increase with time. The temperature profile as it was obtained from the actual process thermolcouples readings is shown in the attached file. The udf code is shown below. when I tried to run the function alone I have noticed that the temperature of the wall is decreasing instead of increasing. Does anyone know why? Is the code wrong?

The code:

#include "udf.h"

DEFINE_PROFILE(unsteady_wall_temperature, thread, position)

{

face_t f;

real t = CURRENT_TIME;

begin_f_loop(f, thread)

{

F_PROFILE(f, thread, position) =(-9e-23*t*t*t*t*t*t)+(8e-18*t*t*t*t*t)-(2e-13*t*t*t*t)+(3e-09*t*t*t)-(2e-05*t*t)+(0.0845*t)+282.3;

}

end_f_loop(f, thread)

}