Schneer Sauer Bubble Number Density
Hi everybody,
I'm using Schner Sauer model for modeling cavitation. I want to simulate fuel oil liquid and fuel oil vapor in a pump (I'm using mixture multiphase model). I have some convergence problems and i think the problem is about cavitation model. I don't know how correctly set number bubble density and I'm using the default value: 10e13. Has anyone any guidelines? Thank you all for your replays!!!!!!! :) |
Hi,
cavitation problems are usually difficult to converge because of their implicit physics. The Schnerr and Sauer model takes into account the bubbles number density as the custom parameter: this is the number of bubbles per cubic meter. This parameter is a function of the liquid quality, i.e. the quantity of dissolved gases. Usually the default value is ok for not degassed water. But, since it is an input parameter, this should be tuned by comparing simulation results with experimental tests. To solve your convergence problem I suggest to switch to unsteady solver and use a small time step (it can be 1e-7 s, 1e-8 s). Cavitation problems are in general very computational expensive problems... |
Besides bubble number density, are we supposed to assign an initial nuclei size? Because if we want to assign an initial vapor volume fraction, based on the Schnerr-Sauer formulation, we also need the initial nuclei size. How do you choose the value for that?
|
Hi,
In Fluent the only parameter for Schnerr-Sauer model is bubble number density (n_0). Do I also need to assign the size of bubble nuclei (R_0)? If yes, how can I do it? Also, do I need to calculate the initial vapor fraction using n_0 and R_0 (using the equation in Fluent Theory Guide) and use that value as initial vapor volume fraction when I want to initialize the solution in Ansys? Or I can simply initialize the simulation with initial vapor fraction set to be zero? |
Hi Navid,
Usually in cavitation, the second phase is created from the first phase only when local pressure drops below vapour pressure. This means that the initial vapour fraction can be safely set to zero. This is the way I used to do for cavitation simulations. This will work for normal fluids like water. You may have solved it already; if not you can use this. But for highly thermally sensitive fluids, though the physics direct us to use zero initial vap vol fraction, setting a very small value may help in getting better convergence. This is pure guess. You may try it. |
All times are GMT -4. The time now is 19:58. |