about multiple domain and face region workflow

[mustafa uygun]

Hi, I has mentioned before my works at head modeling. I saw your page about MMC and read script. I has some problem again about face labeling. Firstly I want to explain what i try to doing. 

I have head MRI masks which roughly segmented by me, as muscle mask, skull mask and brain mask. They have not holes for both internal failure and other masks intersection volume. I want to create volume mesh and import to the COMSOL for thermal analysis. 

But i need spesific face labeling. For brain and skull, i need no spesific boundary condition. maybe bottom face of head, for brain and skull overflow to muscle tissue due to MRI sequence has cutted from brain stem level,

 there will be spesific -37 C- temperature B.  and hence i need no face number more than one. Minimum face partitioning is sufficient. 

For head surface, I will add colling cap for upper head volume. For bottom head, I will add 37 C temperature boundary condition and other remaining face surface will be exposed to convection heat transfer. 

I try two seperated ways. First one is demonstration three at samples directory. I obtained geometry but has too much node, too much face. When I looked at node and face data, I saw that it is seems legit due to face labels are 1, 2 or 3. But in COMSOL there are thousands of faces, it kills my work. And I noticed that there are more duplicated faces. 

Other way, like in MMC head model page. I load my mask into matlab. I smoothed the volume and I convert them surface via v2s routine with cgalmesh for each mask seperately, and kept region point datas. I edited face labeling for brain 3, skull 2 and head 1. And I merged them into with mergemesh into one node, face data. After I used the surface2mesh but I always get PLC error that two subfaces are intersecting. And roughly I take a glance final corrupted face data but I saw that face labeling coloumn is unfortunately always "0".

My opt values is: same for each tissue radbound = 4 and distbound: 2. and maxvol = 100.


If this way will work, after getting surface from volume section, I will edit face labeling with basic algorith. After I will merge surfaces and getting mesh with s2m program. It is my workflow which i think.

Have you any offering about my workflow and problems that i encountered. And finally, is it good choice that masks are completely full region with intersection volume by other tissues? 

Thank you.