Running with MINC and non-MINC elements
In a post I submitted to the TOUGH2 forum ( https://tough.forumbee.com/r/m2fgt1 ), Stefan Finsterle provided me the following steps for running a simulation with both MINC and non-MINC elements in TOUGH2.
(1) Mesh generation (ENDFI at the end), with the zero-volume element separating the parts of the primary mesh that shall and shall-not be MINCed. You now have a mesh that contains MINC and non-MINC elements (and yes: the two types are correctly connected to each other).
(2) Edit the resulting mesh file (named MINC) by (re)moving the zero-volume element or rearranging elements to separate the parts of the mesh that are active (i.e., primary variables change) and inactive (i.e., primary variables do not change, having the effect of a Dirichlet boundary condition).
(3) Run the simulation (with ENDCY at the end).
(4) Include the word "MINCed" as a proper verb into your vocabulary...
I'd like to know how this is implemented in iTOUGH2. When I performed step 1 with the "-mesh" option (i.e., "tough2 -mesh FORWARDFILE EOS"), it generated a "FORWARDFILE.mes" and "FORWARDFILE.min" file. The latter I take to be what was referred to in step 2. However, when I attempted step 3, I couldn't get iTOUGH2 to recognize "FORWARDFILE.min" as the MINC file. Looking at the source code (meshm.f), it appears that it's looking in the temporary directory for a file named MINC, but I don't know how to include this ahead of time, since a new temp directory is generated each time itough2 is run. Is there an option, similar to "-m" that will tell iTOUGH2 where to look for the MINC file?
Ok, now I feel like a dummy, but I think I may have figure this out. Someone please let me know if I'm wrong, but here's what I did:
For step (3), I used the file FORWARDFILE.min as my mesh file (i.e., "tough2 -m FORWARDFILE.min FORWARDFILE EOS &"). However, I made sure that the ENDCY command came before the MESHM block (containing the MINC commands) of FORWARDFILE. I assumed that, by using FORWARDFILE.min for my ELEME and CONNE blocks, the elements had already been MINCed (step 4!).
Yes. Here in more detail:
In step (3) you want to used the MINCed mesh for a simulation. If so, provide that file with the ELEME and CONNE block (i.e., the potentially modified file FORWARDFILE.min) by either (a) copying it into the forward input file, (b) using the -m option (Linux/Unix), or (c) renaming it to MESH (on PC). In the forward input file, remove the MESHM block (otherwise your already MINCed mesh will be MINCed again!) and replace ENDFI with ENDCY.