CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   Questions concerning VOF (

CFDtoy October 8, 2006 15:11

Questions concerning VOF
In vof mixture properties such as rho = alpha*rhol+(1-alpha)*rhogas and same as effective viscosity mu=mul*alpha+(1-alpha)*mul is used?

Why is this mixture property used in the computations? What is its effect?

Further, note that in the mixture viscosity calculation mu = mul*alpha+(1-alpha)*mugas

now if mul = infinity ...or very high value ..what happens to the effective viscosity of the mixture?

I would really appreciate some help in understanding the basic fundamental behind the VOF calculations..



Jim_Park October 8, 2006 17:01

Re: Questions concerning VOF
I'm starring at the original VOF report [LA-8365, 1980].

For completeness, the 'alpha' in your equation is the volume fraction of fluid in the particular cell under consideration (called 'F' in the code, for 'fraction'). So the 'rho' your equation calculates is the mass density of the fluid mixture in the cell.

Clearly this gives the correct density of the fluid mixture in the surface cells.

The code listing does not show the viscosity being averaged in the same way. The input listing shows only one value for viscosity. Therefore I conclude that (at least the original version of) vof does not average the viscosity between the two fluids (in a two-fluid problem).

Of course, a lot of work has been done with the vof idea since Hirt's orginal report and papers.

CFDtoy October 9, 2006 10:52

Re: Questions concerning VOF
Hello Jim:

If you look at the present VOF schemes listed in almost all the papers, they use an effective viscosity and effective density both of them approximated in a linear fashion. Could you comment on the appropriateness of them?

Example: Fluent VOF ..etc


Jim_Park October 9, 2006 13:12

Re: Questions concerning VOF
"Could you comment on the appropriateness of them?"

Yes and no.

I think the density is an appropriate average to use. sola-vof just doesn't use an average viscosity (at least the original version from 1980).

A thought for people to shoot at:

The vof technique assumes a sharp interface between two fluids. When the interface crosses a particular mass cell,that cell is flagged as 'surface.' The momentum equations are applied to control volumes (including the surface cells) surrounding the faces of the mass cell. Nothing scary here - this is just the staggered mesh, with momentum CV's off-set a half cell width from the mass CV's.

So, for control volumes enclosing faces between full and surface cells, use the viscosity at the edges of the control volume and difference the viscous diffusion as if you're were treating a 'usual' variable viscosity. A diagram would really help here.

If you do this, it turns out that the viscosity needs to be evaluated at the center of the mass cells. And it makes sense to me to evaluate them using an 'F-weighted' average.

Please understand that I haven't actually tried this in hte context of VOF. Also be careful of the difference between "mass cells" and "control volumes" in the discussion above.

Good luck and let us know how it comes out.

CFDtoy October 12, 2006 18:54

Re: Questions concerning VOF
Hello ParK: I can understand your comments. I guess it all depends on the way the viscous effects are computed. Now, question is...different fluids respond to the flow field in a different fashion. But in VOF we use a common Pressure field ..a single phase approach ..for both the face and do an does it compare? Dont you think we are missing some physics?

For example, assume that one of the viscosity is very very high..close to inf and other a smaller number...In a cell with smaller volume fraction, still the averaged value would be far higher !!!! So, averaging does it really help in the computation or someother formulation is required?

Also, Park could you refer me to some papers where this averaging procedures are explained? What if we use colocated meshes?

Thanks for your insight..I am looking forward for more of them ..


ravi October 31, 2006 01:02

Re: Questions concerning VOF
hi, im ravi, working on vof method. i want to know whether in vof method we can give mass flow rates for both liquid and gas, when im simulating the transient stratified flow.

CFDtoy November 1, 2006 01:16

Re: Questions concerning VOF
VOF method solves single incompressible fluid. Well if the VOF is not 1 at the inlet can have dual liquid/gas combination. But the mass flow rate cant be different. You need to come up with a single velocity, pressure or some form of inlet ...taking into account both the phases !

ravi November 1, 2006 04:26

Re: Questions concerning VOF *NM*

ravi November 1, 2006 04:34

Re: Questions concerning VOF
hi, cfd toy thanks for reply, in your reply u have mentioned that, mass flow rates cant be different,can u tell clearly about this please!

ok if i have to give the mass flow rates of liquid and gas both. how can it be possible for giving same mass flow inlet for both phases. because the liquid phase will be evaporating in to gas, the vol fraction of gas will be x. suppose total mass flow rate is m, then mass flow rate of liquid will be (1-x)m, and mass flow rate of gas is mx. can u explain this clearly that how to specify the mass flow rates separately. one more, do u have any idea about taylor-raleigh instability parameter

CFDtoy November 1, 2006 12:07

Re: Questions concerning VOF
You did not mention that the liquid would evaporate which case add a source term to the mass conservation equation and that would take care of it. Look at the fluent tutorials they discuss some detals on VOF implementation with source terms in mass conservation.for mutual mass exchanging flows..

m = rh*a*c

may be u can do a rho avg using rho = rho1*x+(1-x)*rho2 c = effctive velocity etc and 'a' the area..

i have not worked in detail regarding taylor-rayleigh stuff..I can learn for sure :)


ravindra November 1, 2006 23:28

Re: Questions concerning VOF
hi cfd toy thanks for your reply, it will surely help me, can u send me any theory regarding the taylor-raleigh instability parameter, for two phase flows. if iam giving velocity inlet for liquid and vapour separately, how can i specify volume fraction of both phases? can we give volume freaction as 0.5 and 0.5 (or any between 0 and 1) for both the phases at initial condition, because i want to simulate stratified flow.(which contains both liquid and vapour separated by boundaries). waiting for your reply.

All times are GMT -4. The time now is 06:55.