Three or more phase VOF visualization
|
In VOF problems with more than 2 phases, there is no direct variable to visualize all the three phases simultaneously. Solution no. 1197 suggested the following custom field function, which would help in visualizing all three phases at the same time in a three phase problem. Function = phase-1 + 2*phase-2 + 3*phase-3 where phase-* is the volume fraction of each phase. When the function is displayed, all the three phases would be shown in different colors with lowest in the color scale being phase-1 and highest being phase-3. But above workaround doesn't actually helps, since we see the color from one phase is intermingling with the same color of the other phase and its difficult to track each phase distinctively. Further it might even give a wrong impression about the physics thats going at interface (since actual intermingling of phases occurs at micrometer level or may not even happen at all) With 3 phases or more there is a gradient of colors to represent the VOF fraction from 1 to 0, and the problem is even more intense when 2 phases are actually mixing together. It is a inherent limitation of the VOF method and the post-processing which is grid dependent, which is not addressed so far. Apart from solution suggested in "Problem Decription" follwoing few methods worked well & resolved the issue. 1) Turn off legend, scale, axis under display-option Turn on overlay under display-scene Use a range 0.99 to 10 for say, phase-1; select clip to range; This should be blue Use a range 0.5 to 1.5 for say, phase-2; keep the clip to range on; This should be yellowish Use a range 0.95 to 1 for phase-3; keep the clip to range on; this should be red. Save the graphics; You may want to play with the ranges to get the right colors. Remeber to turn off the overlay for next time. 2) Normally just display one plot per phase and put them on one page in the report. This shows what is going on and is pretty easy to explain. Never seen a technical solution for this. As for the interface effect they will only see micron scale mixing if the mesh is sufficiently fine to detect it. 3)The problem will always be there if you try to convey the information using a single colour-range. Therefore, one approach is to separate the contour plot into more than one piece, and to display the pieces differently. A simple example of this would be to take the cross-section surface, and create a new surface from it by Iso-clip, for example clipping to high fractions of phase-3. Similarly, create the inverse of this Iso-clip. You can now plot contours of phase-1 fraction on the inverse surface, and perhaps just plot the phase-3 surface as part of Display Grid (in grey). Regions displayed in colour with very low phase-1 fraction must be high in phase-2 (since the phase-3 content must be beneath the clip value). This approach would work best if there are clear regions of one phase (phase 3 in the example above) -- and that is the best regime for the VOF model. All above methods works in one or the other situation. |
||
|