[Domagoj]

Hello,

i need to integrate current density in my axisymmetric Fluent model in order to obtain self-induced azimuthal magnetic field (Ampere circuital law) - see attached picture.

Z is symmetry axis, r is radial distance.

How to write UDF for this problem?

Is it possible to loop cells for various z=const from r=0 to the right boundary while replacing integral with sum of finite values calculated in cell center?

Any suggestions?

Thanks.

[AlexanderZ]

there is build in electric potential model in fluent
did you try to use it?

[Domagoj]

Originally Posted by AlexanderZ

there is build in electric potential model in fluent
did you try to use it?

Current density values are obtained using electric potential model, but it does not solve magnetic field values.

Also, there is advanced add-on module called MHD but it solves problems with externally-imposed magnetic field. In my case magnetic field is self-induced and I do not see how to apply it?

My final goal is to obtain Lorentz force from current density and magnetic field and use it as momentum source. Than I could solve fluid stirring problem.

[AlexanderZ]

I'm not sure how to do it step by step.
But it seems you should use user-defined scalar to define your magnetic field

read Ansys Fluent Customization manual for more information regarding UDS

[AlexanderZ]

does current field pre-defined? or you should solve some equation to get current distribution?

[Domagoj]

Originally Posted by AlexanderZ

does current field pre-defined? or you should solve some equation to get current distribution?

Current field is solved using Electric potential model with appropriate boundary conditions at the external walls. I have saved all necessary cell data in UDM's and should solve Ampere equation to obtain azimuthal magnetic field.

It should be relatively simple numerical integration but I am not sure how to locate and loop cells from some constant z-coordinate (and r=0) to the right boundary end perform summation instead of integral.

[AlexanderZ]

inside cell loop you may check if the cell is inside region (z-> your_boundary_coord) and make summation if True.
using C_CENTROID macro you can get coordinate of cell

detail information in Ansys Customization Manual

[Domagoj]

Originally Posted by AlexanderZ

inside cell loop you may check if the cell is inside region (z-> your_boundary_coord) and make summation if True.
using C_CENTROID macro you can get coordinate of cell

detail information in Ansys Customization Manual

Thanks, I will try to write some code and post it here.
I thought that someone already has something similar that could be modified for my case.

[rodrigovime]

I have the very same problem but in ANSYS CFX. It seems that I cannot integrate using the NLOC array. Is there a way to perform numerical integration using User Fortran?

[Domagoj]

Hello,

In the end I obtained Lorentz force using magnetic vector potential method.

I can't remember the implementation details (UDS were used), it was long time ago, but read this paper carefully.

https://iopscience.iop.org/article/1...-3727/37/6/013