Anistropic thermal conduction in Fluent 6.2
 

Hi fluent users

I have a long list of questions in making the anisotropic thermal conductivity work in fluent 6.2. I'd greatly appreciate your input into this.

Problem Set-up: This is a case study with pure heat diffusion from a source in a circular domain. The source is located in the center of the domain. A UDF imparts anisotropic heat diffusion coefficients in some volumes near the source. This is an unsteady problem. Over time, the species diffuses NOT asymmetrically and no matter what the extent of anisotropy is, the species distribution is symmetric. This is not what one would expect. In addition, the inbuilt panel for specifying the anisotropic tensor components (XX, YY, XY, YX) yields the same result (homogenous circular contours as against anisotropic elliptical contours). The question I have is: Is the anisotropic diffusivity really taken into calculation through a single UDF I specify for the single UDS in enable to make the thermal conductivity anisotropic.

1. What is the typical range of values for Kxx, Kyy, Kxy and Kyx in the anisotropic thermal conductivity tensor matrix? Are the units of the tensor components given in W/m-k?

2. Is this an Eigenvalue matrix or the regular thermal conductivity tensor? 3. Can it always take off-diagonal elements or the tensor is assumed to lie in its principal axis so that only the pivots really influence the problem though the user inputs the off-diagonal elements? 4. Can the off-diagonal elements be negative? 5. Can the thermal conductivity tensor be non positive-definite (I don't think so)? 6. What is the typical range of value of the pre-multiplier? Is the pre-multiplier a product of density and thermal conductivity diffusivity? Or, what is the physical significance of the pre-multiplier and in what way it differs from the thermal diffusivity in the panel? 7. How does the tensor relate mathematically with the thermal diffusivity (K) value of the species in the material property panel apart from the UDS diffusivity? 8. It seems that the anisotropic panel (default) is not working at all or I still do not see the big picture in making the tool work to see ELLIPTICAL contours instead of CIRCULAR contours for a domain with thermal conductivity anisotropy.