CFD Online Discussion Forums - [ANSYS Meshing] Solid inside Fluid

Hi all,
I am an Ansys Meshing beginner, so I ask for a little help.
I want to obtain a structured grid, at least in the vicinity of the boundary walls, for a problem constituted by a large box of fluid and a void completely inside him (cilindric). I obtained the fluid geometry by means of a boolean subtraction in the DM sofwtare and the void geometry by non preserving the tool body. It's the case reported in this image.

http://img208.imageshack.us/img208/8181/casestudy.jpg

Uploaded with ImageShack.us

The boundary wall in my intention are the ground (the base of the big fluid domain) and the walls of the void shape.
As shown in the figure, the internal void shape can be considered as a vase which is here modeled without the solid part, so there is a zone of fluid contained inside the vase while the major fluid domain resides in the external part of the vase.
The problem is that if I select the fluid body and insert in a Multizone method with Hexa "mapped mesh type" and a not allowed "free mesh type" the resulting mesh doesn't make is very coarse in the fluid domain, with very large grids on the domain surfaces. So I tried to specified an Inflation boundary on the base of the larger box (representing the ground) and on the walls of the void but here I got a new problem. How can I select the void surfaces for the inflaction specification if they are completely immersed in the fluid?
I know that this would be an easy task for most of you so please give me some suggestion about how to obtain a good grid for the fluid domain with structured mesh at least around the solid and the ground boundary. This is necessary for the application of a LES simulation.
I found some help in the forum for similar problem but only for ICEM CFD by blocking the domain, but nothing about a case like this with ANSYS Meshing and multizone method.

Thank you for any tips!