[DWilson1997]

Hello,

I am currently trying to create a model, the purpose of this model is to measure the diameter of the bubbles produced at the outlet of this venturi type pipe. For this model water is flowing into the inlet of the pipe with air also being injected at this inlet.

I am also trying to find a balance between the time it takes to run the simulation and the quality of the data produced.


Models used:
Multiphase eulerian 2 phases
Viscous realizable k-e
Discrete phase air injection @1.045m/s 0.01045kg/s mean diam 0.0018m

Materials:
Air
Water
Nitrogen

Boundary conditions:
Water and air at inlet at 1.045m/s and 0.01045kg/s for each phase
Water and air injection enter the inlet at 1.045m/s
Gravity -9.81m/s^2

500 time steps at 0.0025 step size and 10 iterations per time step

Automatic mesh method used element size 5e-4 with sizing refinement made at outlet to 2.5e-4 as this is my area of interest

I'm using an IdeaPad S340 ryzen 7 processor 2.30GHz 8GB RAM to run the sims

Upon initially running the simulation iterations occur quickly however about 10 step sizes in messages regarding reversed flow and turbulent viscosity limits are seen as well as a very small amount of the discrete phase mass not being tracked and the iterations almost come to a stand still.

Is there anything I can do to fix this like I said I need a balance between good data and processing time of the simulation.

Any advice would be appreciated I can supply more info if you need to know more