[josp]

Dear all,

I have a problem with the orientation of the faces in a set of control surfaces that I created for diagnostic purposes (for integrating mass flow in parts of the domain that are of particular interest to me).

I create the surface as an internal surface in Tgrid, if I look at the normals in Tgrid they look correct (i.e. all normals points in the same direction if the surface happens to be a plane).

I then import the mesh into OpenFOAM by:

fluentMeshToFoam -writeSets -writeZones

Everything seems fine with the conversion and checkMesh is ok.

What seems strange is that about half of the normals of the faces in the control zone have "flipped" direction. I tested that by taking the scalar product between the face normal and the distance vector from a control point (using a simple enough control surface). Almost 50-50 in normal direction.

Has this something to do with which cell that owns the face? Do I need to find a flip map somewhere in the conversion process?

If I look at the normals of the control zone in paraview after typing foamToVTK they look correct again..

Any help greatly appreciated!

Regards

Johan Spång