The question is the following: I have to determine the optimal thickness of a polyuretane layer for a small container, with an outside enclosure of Aliminum panels. Therefore, I 've thought to solve a heat transfer problem in transient state in Ansys CFX 10. I know the power source inside (700 W), the temperature of internal hot fluid and the outside temperature of cold fluid. Substantially, I suppose to treat it as a conduction problem, and it is not so difficult with an analytic approach in one dimension. ANyway, I've been requested to simulate in 3D, with the presence of natural convection. In this first step I'm neglecting the internal components (made of plastics and metals). My difficulty is in the definition of input domains and boundary conditions. How can I define the internal insulator surfaces? As walls? And the internal fluid? Is it necessary to define an inlet and outlet entity? I've not found yet a tutorial quite similar to my problem. Please help me. Thank you in advance. Bye. Igor

check multicomponent modelling approach. as for internal walls are considered, you need to sepearately define and mesh the internal part and then export to pre

Thank you, I followed your advice. At present I've decided not to model the internal fluid region, so I've defined a separate domain for the internal regions and put there energy and transfer coefficient as boundary conditions. I've run a transient analysis and it seems to work. Now, I want to plot the cooling time for the insulator and see how it changes with the variation of thickness. I've noticed that in Tools->timestep selector I can only set initial and final time step, but for the intermediate steps? I'm sure there's an easy way to do this. The next problem for the model I've to analyze in future is a bit more complicated for me. I try to explain it: imagine my container with polyurethane enclosure, recovered by external aluminum panel. Inside I've a lot of components: most of them are sources of energy; neglecting radiation phenomenon, and within the hypotheses of pure convection, I 've to consider the presence of internal fluid, but is it really necessary to mesh it, creating a volume with "Fill" tool in ansys workbench? The parts inside are suspended, as they floated, so there's no contact with the internal walls. I've already tried to define such a geometry but without success: "Fill" tool give me errors, so it is not possible to build the internal fluid region this way. Can you help me? Thank you a lot.