VOF for single phase?
This post is lengthy- with hopes that it is as complete and thorough as possible. Please bear with me and thank you in advance.
My experience with FLUENT has been pretty one-dimensional (though not in the geometrical sense). My modeling is simply steady-state applications of flow perfusion throughout various devices/designs...basic velocity inlets, pressure outlets, and occasional porous media domains.
But now I have a need to visualize the time-dependent progression of the 3D volumetric flow through my devices. I need to be able to define various 2D planes (at different perspectives/orientations) and look at the fluid as it fills the cavity in a discrete fashion. I can step through the fluid flow using DPM particle injections, but this only shows pathlines- and adding rakes/surfaces to the inlet still is limited by the grid resolution of my mesh. Not to mention I haven't found a way to view pathlines of a 3D model on an 2D surface...
Any thoughts on how to visualize the data I have in the fashion described above without the need of any discrete/multiphase modeling?
I tried a VOF model, with phase 1 and 2 as the same liquid (logic being that phase 2 replaces phase 1) but am having difficulties with BCs (possibly errors due to patching?) and post-processing. The parameters in my model are:
-Pressure-based, unsteady solver
-Implicit, 2-phase (identical material properties) VOF
-BCs: Inlet: phase 2 volume fraction = 1; mixture = absolute [constant] velocity
Outlet: phase 2 volume fraction = 0; mixture = pressure outlet
-PISO P-V coupling
-Discretization: Green-Gauss Node Based, PRESTO!, momentum and volume fraction are both Second Order Upwind
-Second Order Implicit Transient Formulation
-Non-Iterative Time Advancement is turned OFF
-Default under-relaxation factors
-(no surface or volume monitors are created - should they??)
-Initialization: phase2 volume fraction has been set at both 0 and 1 (neither yield usable data)
-(I could not patch any surfaces as was done in the FLUENT tutorial. Some other forum posts said this could be done with the same scheme employed in an adaptive mesh- but I don't want to have an adaptive mesh - any work-around??)
-Fixed time step size of 1e-5, 2000 time steps, 20 iterations/time step, report interval = 1
Misc.: laminar, incompressible fluid
Results are not consistent with steady-state data (though the solver says convergence is met (1e-4)). NOTE: steady-state data has been validated and presumed accurate.
Moreover, I cannot define discrete time points in my contours [viz. contours-> phases -> phase # -> volume fraction]. All graphs are at time == 0 and phase 1 = uniformly 1, phase 2 = uniformly 0. I can't define vf = .5 (e.g.).
I know this was lengthy, but hopefully it was thorough enough to assist the reader in their support. Thank you kindly.
|All times are GMT -4. The time now is 10:21.|