[A_Pete]

I was just trying to do a loop over all the cells in a cellZone. I am pretty sure that this worked some while ago with 2.2.x. The compiler doesn't complain, but the solver does. I have three examples below which could all be compiled, but the solver stops when he gets to the specific points that I marked. The solver doesn't have a problem getting the cellZone's idea nor getting a reference to the whole cellZoneMesh. It stops when trying to define one single cellZone though. In the 3rd example it stops when getting to the loop, which is basically the same problem for me.

Version 1:

const cellZoneMesh& czm(mesh.cellZones());    // this works
const label& zoneID(mesh.cellZones().findZoneID(faceName));    // this works

const cellZone& cZone2(czm[zoneID]);    // this doesn't work -> solver stops here

Version 2:

const cellZoneMesh& czm(mesh.cellZones());    // this works
const label& zoneID(mesh.cellZones().findZoneID(faceName));    // this works

const labelList& cZone2(czm[zoneID]);    // this doesn't work -> solver stops here

Version 3:

const cellZoneMesh& czm(mesh.cellZones());    // this works
const label& zoneID(mesh.cellZones().findZoneID(faceName));    // this works

forAll(czm[zoneID], i);    // this doesn't work -> solver stops here

I am pretty sure that it should be possible to loop over all the cells in a cellZone, since I can find these code snippets in the repository's sources as well. I am also quite confident that I used this kind of syntax in 2.2.x before.

Does anybody have an idea what I am doing wrong here?

[floquation]

I'm not an expert, but I do have 2 general suggestions in analogy with problems I faced with OF23x's cellID.

From the looks of it,

czm[zoneID]

is failing. Consequently I assume that zoneID is out of bounds.

Suggestions:

1. Add a print statement between the working line and the failing line, displaying zoneID. I suspect that it is out of bounds (e.g., OF might return -1 if "faceName" is not recognized).
   

   Sout << "[" << Pstream::myProcNo() << "] zoneID = " << zoneID<< nl;

   (Note that this prints for every processor. In case you are running parallel (are you?) and a zone is not known to that specific processor, that specific processor may return -1 and consequently crash your simulation, even though "faceName" is known to one processor.)
2. Post your (runtime) error on the forums, such that someone might recognize it.

[A_Pete]

Ah yeah that "Sout" output helped me! I was looking for the wrong zoneID! Thanks man!