[C.C]

Hi all,

I want implement a new turbulence model (CHC model) and the wall boundary condition for epsilon is: 2*kinematic viscosity*k/(wall distance)^2
in which kinematic viscosity, k and wall distance have the values corresponding to the 1st node off wall. Anyone knows how can I implement this.
I tried:

/* Wall boundary condition epsilon */
DEFINE_PROFILE(wall_epsilon,t,i)
{
cell_t c;
real xw[ND_ND];
real rw;
real EpsWall;
begin_c_loop(c,t)
{
C_CENTROID(xw,c,t);
rw=xw[1];
EpsWall=2.*C_UDSI(c,t,TKE)*0.001003/998.2/SQR(0.0381-rw);
F_PROFILE(c,t,i)=EpsWall;
}
end_c_loop(c,t)
}


The value obtained is always zero, but epsilon is different from zero.
Please, anyone can help me?

[C.C]

As at wall I imposed epsilon=C_UDSI(c,t,TKE)=0, it seems that the wall boundary consition for epsilon is correctly interpreted by fluent. My main doubt is how can I impose C_UDSI(c,t,TKE) in the expression of EpsWall equal to the value of k in the 1st node off wall?

[C.C]

Please, anyone can see if the following part makes sense? Thank you.

/* Wall boundary condition epsilon */
DEFINE_PROFILE(wall_epsilon,t,i)
{
cell_t c;
real xw[ND_ND];
real rw;
real EpsWall;
face_t f;
Thread *t0;
cell_t c0;
c0 = F_C0(f,t);
t0 = THREAD_T0(t);
begin_c_loop(c,t)
{
C_CENTROID(xw,c0,t0);
rw = 0.0381-xw[1];
EpsWall=2.*C_UDSI(c0,t0,TKE)*C_MU_L(c0,t0)/C_R(c0,t0)/SQR(rw);
F_PROFILE(c,t,i)=EpsWall;
}
end_c_loop(c,t)
}