Learning Forum

[wuxuetianzi]

Dear,

I want to use UDF to modify the diffusion when active species transport, melting and VOF model. There are three zones, zone 1 is the surrounding environment, zone 2 is the porous particle, zone 3 is the ash particle. also there is reaction happened on the surface of the porous particle to increase the temperature. ash particle will be melted once temperature is hight enough. phase 1 is the mixture of gases, phase 2 is the material for ash particle.

I do not active any phase interactions option. the problem is how should I define zero diffusion in the UDF? I can set a function of diffusion when the thread is on zone 1 and zone 2, but for zone 3, I can not set zero otherwise there will warning says the material property can not be zero. if I dont even mention zone 3 in the UDF, the simulation can not even be started.

Hope anyone can help me with that, thanks!

[Rob]

If you've got a species mixture assigned to a phase you set the diffusion coefficients within the species mixture. You can't specify zero diffusion (numerically it's not a good idea) but a value of 1e-15 should do much the same thing. 

[wuxuetianzi]

Thanks for the reply!

I used ‘report definitions’ and found there is a negative mass flow rate on the interface between zone 1 (fluid) and zone 2 (Ash) after I used very small diffusion coefficient for zone 2. since there is no mass transfer between phases, should the mass flux rate be zero? my domain is getting from Get_domian(2) and thread defined as thread *t_ash = lookup_thread(d, Ash_zone_ID) and similar to other zones. also if the thread pointing Ash zone. it will return a very small value. Am I using the correct function? Thank you!

[Rob]

If it's multiphase you can very easily have movement of the phase over an interior surface.  Note, interior surfaces don't hold a surface normal so you may need to check you're not seeing odd results from material moving in either direction.  Not sure about the macro, please review the UDF manual. 

Very small values can be a result of mesh & convergence. If it doesn't effect the solution you need to decide whether you need to further refine the mesh and potentially converge to a much lower level (e-5 for continuity). 

[wuxuetianzi]

As you mentioned ' not seeing odd results from material moving in either direction.' ,when I used 'contour--volume fraction' to track Ash phase, the ash phase started to float above the porous particle after many steps. But it should always stay where it was waiting for the melting. I also used vector-velocity to check at the same place,  there are velocity arrows going through the Ash phase. Even it should be solid at this time because the temperature is not hight to melt.

Please advise, thanks!

[Rob]

You'll always see some numerical diffusion: have you fully converged each time step, is the mesh resolution sufficient? Also when you display the contours are you showing node values? The contours tend to be smoothed by default, which generally gives a better image but isn't exactly what Fluent is calculating (with a fine mesh it is).