[Kamu]

Dear all,

How can I use a non-uniform heat flux boundary condition along the circumference of a tube?

Regards

[Kamu]

Originally Posted by Kamu

Dear all,

How can I use a non-uniform heat flux boundary condition along the circumference of a tube?

Regards

Is there any way I can write a udf for such in 3d?

[flotus1]

With the define_profile macro you should be able to specify such a boundary condition.

  #include "udf.h"   
 DEFINE_PROFILE(heatflux,t,i)
 {    
  real x[ND_ND];       
  face_t f;    
   
  begin_f_loop(f,t)
        {
          F_CENTROID(x,f,t);
        
          F_PROFILE(f,t,i) = 1.0;
          *instead of 1.0, define whatever spatial function you want
          *x[1] is the x-coordinate, x[2] the y-coordinate and x[3] the z-coordinate
        }
       end_f_loop(f,t)
}

[Kamu]

Dear Alex,

Thanks a lot for this. I will try it out.
so i do not need to worry about converting the cartesian coordinates to cylindrical coordinates?
The heat flux varies only in the circumferential direction and not in the radial or axial direction.

Regards
Kamu

[flotus1]

If your simulation is real 3D then there should be no problem.

[ravi1005]

above program is not running and what is the meaning of x[1],y[1] and z[1].

[flotus1]

"program not running" is not an error description to work with.

There is no x[1], y[1] and z[1] in the code.
As the comment lines say, the array x holds the position vector. Thanks to my fortran background I shifted the indices. The comment line should read:

*x[0] is the x-coordinate, x[1] the y-coordinate and x[2] the z-coordinate

I would edit this error, but the post is too old.

[m.f.arefi]

hi can you tell me what's the heat flux function you've defined mathematically as non-uniform heat flux?

[AlexanderZ]

  #include "udf.h"   
 DEFINE_PROFILE(heatflux,t,i)
 {    
  real x[ND_ND];       
  face_t f;    
   
  begin_f_loop(f,t)
        {
          F_CENTROID(x,f,t);
        
          F_PROFILE(f,t,i) = 1.0;
          *instead of 1.0, define whatever spatial function you want
          *x[0] is the x-coordinate, x[1] the y-coordinate and x[2] the z-coordinate
        }
       end_f_loop(f,t)
}

for instance

F_PROFILE(f,t,i) = x[0]*1000.0;

is heat source which increases along x-axis 1000 times with each meter
heat source could be time dependent as well

best regards

[m.f.arefi]

Originally Posted by AlexanderZ

  #include "udf.h"   
 DEFINE_PROFILE(heatflux,t,i)
 {    
  real x[ND_ND];       
  face_t f;    
   
  begin_f_loop(f,t)
        {
          F_CENTROID(x,f,t);
        
          F_PROFILE(f,t,i) = 1.0;
          *instead of 1.0, define whatever spatial function you want
          *x[0] is the x-coordinate, x[1] the y-coordinate and x[2] the z-coordinate
        }
       end_f_loop(f,t)
}

for instance

F_PROFILE(f,t,i) = x[0]*1000.0;

is heat source which increases along x-axis 1000 times with each meter
heat source could be time dependent as well

best regards

hi again thank you for your reply
I m working on a pipe in pipe volumetric solar collector and I want to choose best radiation model in ansys fluent because my multiphase model is mixture for simulating nanofluid and my solver is pressure based you know when I read papers related to this topic I found that monte carlo method (density based) is a proper model and using udf file to consider non uniform radiation is applied so what should I do ?I cant change to monte carlo method because mixture model is important for me...

[AlexanderZ]

from my point of view monte carlo method is just a slover, use other one
write new paper

best regards