The UDF and FLUENT case to simulate EHD flow in the point-to-ring corona configuration
This code accompanies the paper: Experimental and numerical investigation of electrohydrodynamic flow in a point-to-ring corona discharge By: Yifei Guan, Ravi Sankar Vaddi, Alberto Aliseda, and Igor Novosselov URL: https://journals.aps.org/prfluids/abstract/10.1103/PhysRevFluids.3.043701
To implement the UDF properly:
- Open the FLUENT case file
- Compile UDF
- Setup everything properly as in the example. (The corresponding names of the UDS is listed in the UDF file) Here something to notice is to unselect the equations for other UDS except for potential and charge density. The diffusivity of charge density should be specified. The time-dependent term for potential should be unselected. The uds1_flux function is actually for uds2 which is the charge density. Remember to set the number of user-defined memory locations = 1
- Un-load the libudf
- Initialize the flow field (zero velocities should be fine)
- Run for 1 iteration (No need to run the whole time step)
- Load the libudf
- Unselect the charge equation in solution controls.
- Run it until the potential is steady-state
- Activate the charge equation solver in solution controls
- Run it at time step size 1e-7, then incrementally increase the time step size as the charge density is becoming steady state.
If apply it to other geometries, make sure these are correct:
- In the UDF file, change #define factor (2.0 * M_PI) to #define factor 1 if the geometric is 2D planar or 3D.
- #define x_ionization -3.0e-3 denotes the ionization zone boundary limits to avoid ionizations due to coarse mesh or large curvature in other places except for the corona anode. Change it if you need.
- In DEFINE_ADJUST(effective_volume,d), change t = Lookup_Thread(d, 12); to the fluid domain in your model.
- Change other Lookup_Thread, e.g., t = Lookup_Thread(d,14); to the correspondng boundaries in you model.