i try to simulate the shape of a 100 microliter olive oil droplet on water. looking at the interfacial tensions of the system (approx. values from varoius papers: water-air 72mN/m, oil-air 32mN/m, oil-water 25mN/m) and using youngs equation the droplet should spread.
performing the simple experiment using water and olive oil, the droplet does not completely spread: it forms a circular disk. (two pictures attached: top view and side view with two droplets)
performing a simulation using openfoam the droplet does not spread at all...
any hint on a paper describing the shape of an oildroplet in water would be helpful!
What do you mean with that "it does not spread at all"? Not very clear your scenario in OpenFoam. Essentially you could start in OpenFoam with a 100% submerged oil droplet in water. If gravity and buoyancy and surface tension force is included in the simulation, than the oil droplet should first rise to the water surface and than start to spreaden over the surface, since it cannot stay spherical on the surface. It would be a 3-phase Eulerian-Eulerian simulation.
Regards,
Dr. Th. Frank.
hi,
spreading - to my understanding - would mean that the droplet vanishes and i get a solution with three layers (air-oil-water).
this does not happen in reality -
as the experiment shows - the droplet forms a rather flat disk. which already shows that there is a mismatch between my understanding and reality (maybe some 3d effect or wrong interfacial tensions)
in my simulation using multiphaseInterFoam i get a shape which is more spherical. not completely spherical: due to the different surface tensions and gravity it is flatter on the top, but it is about as thick as wide and not a flat disk... with the surface tensions i use i would even expect the droplet to completely spread...
i will try the rising droplet case - but i think the problem is rather an issue with wrong interfacial tensions or missing physics: why does an oil droplet with a spreading coefficient > 0 not spread in the simulation? is that an issue with the VoF apporoach?
Your explanation makes it more clear to me - what you expect, what theory would forecast and what can be observed experimentally.
Regarding VOF numerics: as far as I know, surface tension in a 3-phase Eulerian VOF approach is a tricky thing to get predicted numerically accurate. The forde is by physics only acting on the phasic interface. But in VOF there is no interface tracking algorithm involved. Instead the interfacial force is translated into a volumetric force term, which is just zero in the volume of a continuous phase and non-zero on the interface between two phases. Rather very high grid resolutions close to the interface are required to get the involved gradients predicted accurately, so that the resulting surface tension force term is close to its reality. Some implementations tend to the prediction of what is called "spourious currents" - a small-amplitude artificial velocity field which arises from an imbalance between discretized (interfacial) forces - as well. So I would not be surprized, if in such a scenario like yours, e.g. the surface tension force is getting overpredicted due to its discretization or implementation....
Best regards,
Dr. Th. Frank.
- Badges
- Science topic
- Similar topics
- Engineering
- Petroleum Engineering
- Oil