[Gfs-devel] Concentration tracers using VOF
Jose M. López-Herrera Sánchez
Forget the previous email, I sent it accidentally before I finish the writing. Since you were going to take a look on the VOF approach meanwhile I have been working these days on the BCG scheme using T as volume fraction. The first test was promising but next tests has been very discouraging (maybe I have been doing something wrong) so I have abandoned this line for your more promising idea of a second order VOF.
direction-split VOF advection loop (this loop would become a VOF +
associated tracers advection loop). The loop I am talking about is
here vof.c:1539:gfs_tracer_vof_advection()
What is your opinion about to add a new class of variable named for example GfsVariableConcentration for the associated tracers (similar to GfsVariableCurvature).
About how to apply the SourceDiffusion to this new associated tracers:
- It should be apply in the same loop. right? The loop would become gfs_tracer_vof_advection_diffusion () instead a gfs_tracer_vof_advection ().
- In a first stage it has to be explicit.
- I am not sure how we can introduce a diffusion flux within the direction-split VOF scheme.
Maybe these points are naive but I am still trying to understand gfs_tracer_vof_advection () (I found a bit confusing the concept of effective volume). I have to study deeper the splitting scheme.
Note that having such a scheme would also help with other problems
such as phase change (i.e. evaporation etc...)
You are right, This scheme would abroad a lot the problems affordable with gerris. I am thinking for example in electrokinetic problems
Just thinking a bit more about this. A simple way to make the scheme
second-order spatially is just to weight the VOF fluxes not with the
cell concentration but with the cell face concentration which can be
obtained using the cell concentration and slope-limited concentration
gradients. I think this should work and should be quite simple to
implement.
Right, Ok.
In order to be more efficient and not overlap. Do you want I center my attention to any of the above features (or other you have in mind)?
Cheers
Jose
Stephane Popinet
> What is your opinion about to add a new class of variable named for
> example GfsVariableConcentration for the associated tracers (similar to
> GfsVariableCurvature).
Yes, I have a first implementation of something I called
GfsVariableTracerPhase. I am not sure it's a good name though, maybe
VariableVOFConcentration or VariableConcentration would be better. I
thought that "Phase" could convey the fact that it's a quantity linked
to another field (i.e. the VOF field). What do people think?
It implements more or less what I talked about and seems to be close
in accuracy (and order) to VariableTracer away from the interface
while being consistent with VOF in interfacial cells.
I finish a test case and I will push the implementation and write a
bit more about it.
cheers
Stephane
------------------------------------------------------------------------------
All the data continuously generated in your IT infrastructure contains a
definitive record of customers, application performance, security
threats, fraudulent activity and more. Splunk takes this data and makes
sense of it. Business sense. IT sense. Common sense.
http://p.sf.net/sfu/splunk-d2dcopy1
_______________________________________________
Gfs-devel mailing list
Gfs-...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/gfs-devel
Jose M. López-Herrera Sánchez
Thanks a lot for your feedback.
GfsVariableTracerPhase. I am not sure it's a good name though, maybe
VariableVOFConcentration or VariableConcentration would be better. I
thought that "Phase" could convey the fact that it's a quantity linked
to another field (i.e. the VOF field). What do people think?
It implements more or less what I talked about and seems to be close
in accuracy (and order) to VariableTracer away from the interface
while being consistent with VOF in interfacial cells.
Jose