[zjuevol]

i want to build a simple electric potential using udf for practice
and here is the simple code

#include "udf.h"

DEFINE_ADJUST(potential,domain)
{
Thread *t;
cell_t c;
face_t f;
real x[ND_ND]

thread_loop_c(t,domain)
{
begin_c_loop(c,t)
{
C_CENTROID(x,c,t);
C_UDSI(c,t,0) = 10.*x[2];
}
end_c_loop(c,t)
}
}

when i try to interpret udf the error comes

line:10 parse error
line 12 c_loop_last:undeclared variable

i can not figure out where is wrong with this two line
(thread_loop_c(t,domain) and begin_c_loop(c,t))
and i wander if there anyone can give me some help?

any advise will be a big help for me
thank you

[Bruno Machado]

you missed the ; after the real x[ND_ND]

[zjuevol]

Originally Posted by Bruno Machado

you missed the ; after the real x[ND_ND]

thanks for your help
i add the ; and it successfully interpreted into the fluent
but after calculate i check the value of udsi it is all zero!
do you have any idea about this ?

[Bruno Machado]

Originally Posted by zjuevol

thanks for your help
i add the ; and it successfully interpreted into the fluent
but after calculate i check the value of udsi it is all zero!
do you have any idea about this ?

have you hooked the UDF? did you define diffusivity and boundary conditions?

[zjuevol]

thanks for your reply
since it was the first time to use UDF and may have lots of thing i don't know.
i use this code just try to give UDSI0 some value which did not work know.
As you replied i think you mean i have to set the transport equation and set boundary condition, and the UDSI0 should be calculated by fluent rather than give it by user directly?
again thank you sincerely

[pakk]

If you want to store the potential in a memory location, I think you should have a UDM (User Defined Memory), not a UDS (User Defined Scalar).

C_UDMI(c,t,0) = 10.*x[2];

For Fluent, a UDS is used for something that is transported by your flow. The electric potential is not affected by your flow (except for very unusual situations when you are working with charged fluids), so use a UDM instead.

[Bruno Machado]

Originally Posted by zjuevol

thanks for your reply
since it was the first time to use UDF and may have lots of thing i don't know.
i use this code just try to give UDSI0 some value which did not work know.
As you replied i think you mean i have to set the transport equation and set boundary condition, and the UDSI0 should be calculated by fluent rather than give it by user directly?
again thank you sincerely

There are few things you need to consider:
If there is no source term, field is not affected by flow and also there is no need to compute gradient, you can use UDM to define it, as Pakk said.

If any of the previous mentioned considerations are false, you need to define it through UDS; If so, you need to define the domains it is solved (in the menu, Domain -> User Defined -> Scalars) and also the diffusivity of it in the materials menu.

[zjuevol]

Originally Posted by pakk

If you want to store the potential in a memory location, I think you should have a UDM (User Defined Memory), not a UDS (User Defined Scalar).

C_UDMI(c,t,0) = 10.*x[2];

For Fluent, a UDS is used for something that is transported by your flow. The electric potential is not affected by your flow (except for very unusual situations when you are working with charged fluids), so use a UDM instead.

thanks you for replay i will try UDM to do it

[zjuevol]

Originally Posted by Bruno Machado

There are few things you need to consider:
If there is no source term, field is not affected by flow and also there is no need to compute gradient, you can use UDM to define it, as Pakk said.

If any of the previous mentioned considerations are false, you need to define it through UDS; If so, you need to define the domains it is solved (in the menu, Domain -> User Defined -> Scalars) and also the diffusivity of it in the materials menu.

thanks for all the thing you teach me i will try UDM first