Learning Forum

[muhammad.sarwar]

Hi,

I am modeling two-phase with euler euler, multifluid VOF approach. The approach I am using is that
Phase 1: H2O, OH-, H2 and O2
Phase 2: H2 and O2
In species model, I defined the electrochemical reaction for phase 1 which bascially produce dissolved H2 and O2 in the solution. In multiphase tab, under phase transfer I defined mass transfer of dissolved H2 and O2 to phase 2 (H2 and O2), respectively via Henry mechanism.
The problem I am facing that as soon as I enable two-phase flow the continiuity does not converge. I attached the residual plot for kind consideration. Here I turned of the generation of O2 that's why o2-phase-2 is showing zero.

Please let me know what should I do so that continuity will also converge or guide me what should I look into my setup.

Regards,

Muhammad Asim Sarwar

[Rob]

The problems likely come from rate of phase change and/or mesh resolution as phase change occurs. The main cause being the change in volume from gas-liquid. Decreasing time step or phase change coefficients can help, but also check the mesh: high aspect ratio and/or too small a cell can make convergence more difficult. 

[muhammad.sarwar]

Hi Rob,

Thanks for the reply. 

The fluent automatically calculates mas transfer coefficient. I am not defining it manually. The time step I am using is already very low 1e-05s. Moreover, the maximum aspect ratio is 2.35. Maybe I should coarse a mesh bit as I have too much fine mesh across the domain.

About mesh, can you clear one query about it. I read that for multifluid VOF: radius of curvature is much larger than the mesh size. Is it correct. The bubble diameter I have is 0.0001m. So according to information should I make cell size smaller than 0.0001m.

Regards,

Asim

[Rob]

No need to make the cell smaller than the bubble, or at least I've not seen that part of the manual! Typically you need to resolve the free surface but if you need that level of precison you may need VOF too. How large are your channels relative to the 100micron bubble? 

[muhammad.sarwar]

Hi Rob 

The thickness of the channel is 2mm.  Isn't multifluid VOF has functionality of both euler euler and VOF. The model captures interface where possible and where is not possible it applies euler euler formulation. 

Regards,

Asim

[Rob]

More or less. But as you drop down in scale it's not as straightforward.