Learning Forum

[Andrés Páliz]

Hello everyone, I am doing a steady state simulation of a three-phase separator using the VOF model but I have experienced some difficulties. To put it in context, the model is made with 500k elements, with a maximum obliquity of 0.5 and a minimum orthogonality of 0.6, I am using discretization models for pressure and VOF from PRESTO! and BGM respectively, in addition to the default relaxation factors. The problem that I have is in the difficulty of reaching convergence in the model. Maybe someone has some suggestion that I could do to improve this?

[SRP]

Hi,

Its difficult to comment directly on convergence. Can you please share more details on boundary condition, model setting, solver setting you used, and residual plot?

Thank you.

[Andrés Páliz]

Hi,

The separator has one inlet and three outlets, the inlet has a bc with the mass flow of each face (oil, gas and water), and the outlets also have a mass-flow bc but only with the phase that corresponds (oil, gas or water). The VOF is enabled with the implicit formulation and the implicit body force, the primary phase is gas. The operating conditions are 50psig and I've put it on the gas phase (0;1.2;0). The density operation is the density of the gas phase. I've used a Coupled scheme with the PRESTO! solver for pressure and BGM solver for VOF. The problem is that the fluid starts to float, and the residuals start to increase. These are some pictures.

[Andrés Páliz]

[SRP]

Hi,

For VOF you can't have a mixed phase inlet. So, choice is Eulerian or Mixture.  The outlets are going to cause problems as it is over constrained the problem. I suggest to try pressure outlet boundary condition.

Thank you