[raunakbardia]

Hi All,

I have been working with the isoAdvector code in OpenFOAM v1706+.

I was trying out a basic canonical problem in 1D where the interface is in the center of the domain and simply moves due to a prescribed constant flow. There is no velocity or pressure solution and I use the frozenFlow capability to enforce that. I have attached the case files. In this simple system, the vapor is set as the left half and liquid is set as the right half. Entire domain is set to move towards the right side.

With reference to the original isoAdvector code description(PAPER), this simple case cannot be correctly solved. If the interface is perfectly parallel as is the case in this 1D system, here is what happens in the interfacial cell:
1. For the downwind face, all points are equidistant from the interface. In the function timeIntegratedFaceFlux, this results in the evaluation of the variable nShifts to zero.
2. In the implementation in v1706, the timeIntegratedArea is not evaluated and the liquid flux passing through that cell face is set to zero.
3. For the upwind face, dVf is set to zero as .
4. Hence, the volume fraction of the interfacial cell will never change.

My results show the same thing. Please let me know if I am making some gaping mistake. If not, has anyone worked on a solution yet?
As all the geometric details are quite robustly available, can we not initialize the liquid volume face flux using the face volume fraction() instead of ?
This seems to be a serious issue in a confident application of this method in general problems.