Dynamic Mesh with Boundary Layer
 

Hi Everybody, I'm trying to simulate flow around a rotating body. In order to evaluate exact separation over the surfaces I'd like to use a mesh with boundary layer (BL) applied. As far as I know, the layering meshing method has been developed for translational motions only (i.e cylinder motion), so I use the following settings:

My mesh consists of two separate domains of fluid one for the bulk (tetrahedron cells) and an other represents the BL (structured hexahedra). These volume meshes are created in ICEM and they feature a mutual surface with identical mesh (different types of mesh domains are merged with the aid of piramid elements). Once this mesh is opened in Fluent, the mutual surface presented as a wall and a Shadow surface of it is also generated automatically. If I set one if them to interior type, these two zones are sewn together. The theory is simple, while I rotate the whole BL mesh as a rigid body, the other domain is being remeshed.

Now, after I compiled my UDF of motion and applied it for both my solid surface and the BL (int_boundary-layer) surrounding it, remeshing does not occur in the cells adjacent to the BL domain. I use remeshing and smoothing and no mather how I vary spring factor the result is always the same: negative elements in the deforming domain once the first row of cells deminishes due to the lack of remeshing. I followed the concerning instructions of a 2D case tutorial ('Solving a 2D Box Falling into Water'- Advanced Tutorial on Fluentusers.com).

So my questions are:

- does this way of mehing work proper for such 3D cases? - if so, how should the interface treated (dynamic zone settings: rigid body?)