[fivos]

Hi to everyone,

I am trying to calculate the material derivative of a quantity in the near wall vicinity (i.e. at the 1st cell from the wall) and, for post processing reasons, I want to plot it on the wall surface. The simulation is mixture multiphase, with 2 phases.

So I built the following UDF:

 DEFINE_ADJUST(calculate_dadt,d)
 {
 real time_step, dvofdt_e, dvofdt_m;
 int n;
  
 Thread *t, *tf;
 cell_t c;
 face_t f;
 d = Get_Domain(1); /* mixture domain if multiphase */
 time_step = CURRENT_TIMESTEP;
 　
 thread_loop_c (t,d)
 {
 begin_c_loop(c,t)
 {
 C_UDSI(c,t,2) = 0;
 dvofdt_e = (C_UDSI(c,t,0)-C_UDSI_M1(c,t,0))/time_step; 
 dvofdt_m = dvofdt_e 
           +C_U(c,t)*C_UDSI_G(c,t,0)[0] 
           +C_V(c,t)*C_UDSI_G(c,t,0)[1] 
           +C_W(c,t)*C_UDSI_G(c,t,0)[2]; 
  
 C_UDSI(c,t,2) = dvofdt_m; 
 　
/* set UDS on the wall */
 c_face_loop(c,t,n)
 {
 tf = C_FACE_THREAD(c,t,n);
 f = C_FACE(c,t,n);
 if (THREAD_ID(tf) == 8)
 {
 F_UDSI(f,tf,2) = C_UDSI(f,tf,2); 
  
 }
 }
 }
 end_c_loop(c,t)
 }
 　
 }
 

As you can see in the UDF, I calculate the material derivative of the UDS with index 0 and store it to a new UDS with index 2. Then, I try to put the values of the UDS 2 on the faces of a specific boundary, with ID 8. The problem is that even if I compile successfully my UDS, load it, run fluent without any problem, the UDS returns a zero value on the wall, while the UDS value at the adjacent cell is different from zero; the aforementioned representation is done by nodal and cell values - nodal values on the boundary are zero, whereas cell values are different from zero..

I have to mention that, since I do not want to solve for the UDS, but only to calculate gradients, I exclude it from Solution Controls -> Equations, however I hook the above UDF as a DEFINE_adjust UDF.

Does anyone have any idea how can I instruct Fluent to copy the cell values of the desired UDS on selected wall boundaries?

Thanks in advance..

[blackmask]

First of all,

F_UDSI(f,tf,2) = C_UDSI(f,tf,2); 

is equivalent to "x = x". It seems that you meant

F_UDSI(f, tf, 2) = C_UDSI(c, t, 2);

It could be more efficient if you use a combination of

Lookup_Thread, F_C0, THREAD_T0

outside of cell loop instead of doing face loop inside of cell loop.

[fivos]

Blackmask you are absolutely right!
I was confused with the thread/cell/face indexes that's way it did not work. Now it works as it should!

I will think about implementing it the way you proposed, instead of looping all cells...

Thanks a lot.