[davide91]

Hi, I have implemented several UDFs in my simulations.
In particular, I have a boundary condition that depends on a parameter which I call FRAC.

#include "udf.h"

...
...

DEFINE_PROFILE(myBC,th,i)
{
    real FRAC = Get_Input_Parameter("my_parameter");
    begin_f_loop(f,th)
    {
        F_PROFILE(f,th,i) = FRAC*...;
    }
    end_f_loop(f,th);
}

...
...

Initially, I did not know what the value of FRAC would have to be, so I was solving the simulation using direct optimization and optimizing FRAC so that I would get the correct value of a different variable at the system outlet.
Without going too much into detail, this is based on mass balances so it is physically correct.

When seeing the optimized values of FRAC, I realised it depends on another value which can be obtained as an integral of a certain expression over a specific cell thread of my domain.
In order to avoid having to optimize FRAC for each Design Point of my simulation, I'd rather calculate it that way.
On the UDF User Manual, I found this info:

DEFINE_ADJUST is a general-purpose macro that can be used to adjust or modify ANSYS Fluent
variables that are not passed as arguments. For example, you can use DEFINE_ADJUST to modify flow
variables (for example, velocities, pressure) and compute integrals. You can also use it to integrate a
scalar quantity over a domain and adjust a boundary condition based on the result.

The last sentence seems to describe exactly what I need to do.
Therefore, I have modified by code in the following way

#include "udf.h"

real FRAC;
...
...
DEFINE_ADJUST(my_adjust,d)
{
    Thread *t = Lookup_Thread(d,7);
    real my_integral=0.;
    cell_t c;

    begin_c_loop(c,t)
    {
        my_integral += (UDS dependent expression)*C_VOLUME(c,t);
     }
     end_c_loop(c,t)
     FRAC = my_integral/... ;
}

DEFINE_PROFILE(myBC,th,i)
{
    begin_f_loop(f,th)
    {
        F_PROFILE(f,th,i) = FRAC*...;
    }
    end_f_loop(f,th);
}

...
...

After the modification, the code still runs and converges but it does not yield the same value of FRAC that is obtained via optimization. Therefore, the mass balance I am checking is also not respected.

If anyone as any idea/suggestion of why this happens I'd really appreciate it.

Have a good day.

Davide