A SECOND METHOD TO MESH THE VAT  The meshing technic

Considering the problem of dealing with high number of faces to impose boundary conditions, we tried a "direct strategy" : we built directly the 3-D vat, without building quarters first. Therefore we have a limited number of faces and can easily impose the boundary conditions. We define the mesh on each face separately, and then we ask Gambit to mesh the volume, taking in account the defined faces' mesh. There is no choice for the volume meshing technic : only an unstructured mesh is avalaible. The following picture shows this new geometry : Direct building of the vat.

This way to build the geometry is easier than the one previously developped, however it does not allow to control the meshing technic as well : it 's not possible to choose the volume mesh type, and it's not easy to refine the mesh where it is necessary. The user has to find a compromise between the two technics.
As the mesh generator Gambit is developped by Fluent, the mesh can be directly used to make simulations with Fluent.

However it's important to be careful when defining the reference frame : in fact, the fact that the agitator is rotating must be taken in account. There should be possible to define a moving mesh for this agitator, and to solve the equations in the galilean laboratory frame; nevertheless, as the exterior wall present a rotation symmetry, it's better to solve the equations in the rotating frame. Therefore, the blade are fix in this frame, and the wall is rotating in the opposite direction of the agitator movement in the laboratory frame. Fluent allows to process this case by selectionning Motion Type "Moving reference Frame" in Define -> Boundary conditions -> Fluid :  Fluent simulations results

Below are few results we got when we made simulations with FLUENT.
This results were obtained with glycerin (very high viscosity) and for a Reynolds number equal to 10.
This is the closest situation we could imagine to simulate a visco-elastic fluid which is the kind of fluid that will be mixed in this type of vat. Velocity magnitude in the laboratory frame (velocity is zero at the wall).

We can see on this plot that in the lab frame, the velocity on the wall is zero and the velocity at the edge of the blades follows a linear law: V ~ w*r. Velocity magnitude in the moving frame (velocity is zero on the blade).

In the blade frame, the velocity  is zero on the blades and  equal to w*R on the wall. Velocity magnitude in the blades plan. Tangential velocity in the moving frame.

This picture shows the tangential velocity in the blades' frame. We can clearly see that there is a recirculation : the red area has a positive tangential velocity, whereas the other areas have a negative tangential velocity. This recirculation phenomena can be observed  on the following picture, showing the velocity vectors in a cross-section : Velocity vectors in a cross-section of the reactor.

This final picture shows the pressure field in a cross-section. The blades are rotating with a positive angle in the lab frame, we can verify that the pressure is higher in front of the blades (red areas) and lower behind them (blue areas). Pressure field in a cross-section of the reactor. 