[John mayer]

Hello everyone,
I wrote a UDF that calculates the heat flux in DEFINE_AT_END and stores it in UDM(or C_YI(c,t,i)) in a Eulerian model. Then gets the UDM(or C_YI(c,t,i)) in DEFINFE_PROFILE and calculates.
However, the DENINFE_PROFILE doesn't get the data from UDM, and receives a fatal signal while the C_YI(c,t,i) was used to stores datas.
I had a test in a single-phase model with C_YI(c,t,i), and it works well.

Thanks for any suggestion,
John mayer

[John mayer]

Originally Posted by John mayer

Hello everyone,
I wrote a UDF that calculates the heat flux in DEFINE_AT_END and stores it in UDM(or C_YI(c,t,i)) in a Eulerian model. Then gets the UDM(or C_YI(c,t,i)) in DEFINFE_PROFILE and calculates.
However, the DENINFE_PROFILE doesn't get the data from UDM, and receives a fatal signal while the C_YI(c,t,i) was used to stores datas.
I had a test in a single-phase model with C_YI(c,t,i), and it works well.

Thanks for any suggestion,
John mayer

This is my UDF.
#include "udf.h"
#include "mem.h"
int ID=15;
int sum=0.;
int cc=0.;
DEFINE_EXECUTE_AT_END(execute_at_end)
{
Domain *d;
Thread *t;
face_t f;
cell_t c;
real h;
real q;
real dt=1.;
real k=1.2;
real cc2;
Thread *t_steel;
d = Get_Domain(2);

thread_loop_c (t,d)
{
t_steel=THREAD_SUB_THREAD(t,S_PHASE);
begin_c_loop (c,t)
{
C_YI(c,t,0)=1/C_UDMI(c,t,0);
}
end_c_loop (c,t)
}
}

DEFINE_PROFILE(up_flux_profile,t,i)
{
Domain *d;
face_t f;
cell_t c;
real h;
real x[ND_ND];
real k=1.2;
real q;
real dt=1.;

Thread *t_steel;
t_steel=THREAD_SUB_THREAD(t,P_PHASE);
begin_c_loop(c,t)
{
C_PROFILE(c,t,i)=C_YI(c,t,0);
}
end_c_loop(c,t)
}

[AlexanderZ]

you code can't work

do you calculate heat flux (flux through boundary) or heat source (heat, applied to volume) ?

C_YI(c,t,0)=1/C_UDMI(c,t,0);

C_UDMI(c,t,0) is not defined, most likely you are dividing by 0. But C_UDMI(c,t,0) could contain ANY value inside

why do you use DEFINE_EXECUTE_AT_END but not DEFINE_ADJUST ?

[John mayer]

Originally Posted by AlexanderZ

you code can't work

do you calculate heat flux (flux through boundary) or heat source (heat, applied to volume) ?

C_YI(c,t,0)=1/C_UDMI(c,t,0);

C_UDMI(c,t,0) is not defined, most likely you are dividing by 0. But C_UDMI(c,t,0) could contain ANY value inside

why do you use DEFINE_EXECUTE_AT_END but not DEFINE_ADJUST ?

Thank you for your reply! I had patched the value to the C_UDMI(c,t,0). And the value of C_YI(c,t,0) can be acquired in the contours. The PROFILE would report the error while the C_YI(c,t,0) was used in the boundary condition(bu the C_T(c,t) would not report error).

[AlexanderZ]

how many questions can you see in my previous message?

[vinerm]

That's incorrect usage of C_YI. C_YI is not a generic memory like C_UDM. It stores mass fractions of species and is available only if species transport is enabled. And if that is the case, then Fluent will overwrite C_YI values based on the solution of species transport. So, do not use C_YI the way you are using it. You can use more than one UDM though.

And then, you need to store something in UDM before using it. The code does not store any value in UDM. Furthermore, you don't need two functions. You can do all the calculation within DEFINE_PROFILE.