anisotropic thermal conductivity, please help!

GBLiu

Hi everybody, Here is my problem: I want to change the diffusive term of the energy equation in fluent, since i want to use the anisotropic thermal conductivity, in order to do this, i set the thermal conductivity to be zero in the materials panel, and put my own diffusive terms in the source term, so the second derivatives of temperature are needed for calculating the diffusive terms, according to the UDF HELP of Fluent, i obtain the temperature second derivatives with respect to x and y directions by setting the first-derivatives of the temperature to two UDS (e.g. dt/dx=UDS_0 and dt/dy=UDS_1)through defined Macro: DEFINE_ADJUST, then i used the first derivatives of UDS_0 and UDS_1 to get the second derivatives of temperature:d2t/dx2=dUDS_0/dx and d2t/dy2=dUDS_1/dy in my own source terms for energy equation with different thermal conductivities in x and y directions, but the calculated results is very unrealistic, i don't know where is the problem and how to revise these, please help me, thanks in advance.

best regards GBLiu.