[hasib61]

Hi,

I am trying to add extra mass source through UDF but getting this error message:

Node 0: Process 15088: Received signal SIGSEGV.
Node 1: Process 16164: Received signal SIGSEGV.
MPI Application rank 0 exited before MPI_Finalize() with status 2
The fl process could not be started.

The UDF code is as below:

DEFINE_SOURCE(mass_source, c, t, dS, eqn)
{
real source = 0.0;
real mass_source = 0.0;
real area_vec[ND_ND];
real area = 0.0;
int n;

real v=0.0;
real rho=0.0;

c_face_loop (c, t, n)
{
Thread *tf = C_FACE_THREAD (c, t, n);
real SURFACE_ID = 6;
if(SURFACE_ID == THREAD_ID(tf))
{
face_t f = C_FACE (c, t, n);
F_AREA(area_vec, f, tf);
area = NV_MAG(area_vec);
rho=C_R(c,t);
v=C_VOLUME(c,t);

mass_source = (-(C_YI_RG(c,t,0)[1]*C_DIFF_EFF(c,t,0))* ((1/C_YI(c,t,0) )-1))*rho*area/v;
source += mass_source;
}
}


C_UDMI(c, t, 0) = source;

dS[eqn] = 0.0;
return source;
}

Can you suggest anything on this?

[CeesH]

Simple question; did you specify the user defined memory?

[hasib61]

No, I did not specify any location. This is basically my 1st udf code. Not very familiar with udf coding.

[CeesH]

Ok, so you have to specify your user defined memory under:
define > user-defined > memory, then increase "number of user defined memory locations" to the number you need (in this case, only one per cell).

Now, you are referring to this memory location in your code, so FLUENT will try to store something there, but since the location doesn't exist it will crash.

[hasib61]

Thank you very much.... , it works.....can you help me on how to get the mass source value that i have put into through udf? I want to know if its giving the correct value.....again thanks a lot....

[CeesH]

In this case, the mass source is stored in the UDM. This means you can take a look at a contour plot of the user-defined memory to get a qualitative grasp. If you want to quantify it, set up a solution export (calculation activities > automatic export in transient if you want to write every X timesteps, or just use export > solution data after a steady state run).

With this, you can output your data to text files for excel/matlab, or in the format of your favorite post-processing software. Other checks you can do are reports > volume integrals, where you can for example take the sum of all source terms over the whole domain; handy if you have integral data available from experiments.

Best,
Cees

[hasib61]

Hi,

I am having a small problem with the udf. "invalid command [(rpsetvar]"

I am using C_YI_RG in my udf to get the gradient, but this can only be used in density based solver and it takes a lot of time to converge.

Instead I am willing to use C_YI_G in pressure based solver. It requires the following "you will need to set the rpvar 'species/save-gradients? to #t". To do this I typed (rpsetvar 'species/save-gradients? #t) but its giving "invalid command [(rpsetvar]".

Do I have to initiate anything before typing rpsetvar?

[mep20108@tezu.ac.in]

Thanks, CeesH
your method works for me.