The following is my udf code of setting thermal diffusivity on coupling heat transfer interface. It can be worked but I think it does not affect the simulation results.Can somebody give some advice on it? Thanks! DEFINE_PROFILE(cwall_diff, tf, position) { face_t f;

real er0[ND_ND], er1[ND_ND], es[ND_ND], r0[ND_ND], r1[ND_ND], xf[ND_ND]; real ds, h0, h1, dr0,dr1;

cell_t c0 = F_C0(f,tf); /* returns ID for c0*/ cell_t c1 = F_C1(f,tf); /*returns ID for c1 */

begin_f_loop(f, tf)

{

F_CENTROID(xf, f,tf);

C_CENTROID(er0, c0,THREAD_T0(tf));

NV_VV(r0, =, xf,-,er0);

dr0 = NV_MAG(r0);

h0 = dr0/C_K_L(c0,tf);

C_CENTROID(er1, c0,THREAD_T1(tf));

NV_VV(r1, =, xf,-,er1);

dr1 = NV_MAG(r1);

h1 = dr1/C_K_L(c1,tf);

NV_VV(es, =, er1,-,er0);

ds = NV_MAG(es);

F_PROFILE(f, tf, position) = ds/(h0 + h1); /* harmonic mean of adjacent grids in different zone*/

} end_f_loop (f,tf) }

Does anybody has some opinion about it? When i solve conjugate heat transfer between fluid and solid, it will be imporatant in whole-field solution. And i find in Fluent, the coupling face interfacial thermal diffusivity is dealed with algorithmic mean, but not with harmonic mean? Is it right?