[Pedron]

Hello guys. I am simulating the terminal velocity of particles in power-law fluid. For this, I'm trying to implement a drag UDF, as you can see attached. I've been using the UDF in fluent help to write mine, using the same language, changing only the drag model itself, however, I try to use it, either by interpreting it, or by compiling it (opening fluent at the VS 2012 prompt) and I get this error message "divergence amg solver pressure correction" at firts iteraction. Could someone help me? I've already tried to chance my relaxation factors but with no success.

This is the message error:


My UDF:
#include <udf.h>


DEFINE_DPM_DRAG(particle_drag_force,Re,p)
{
real w, drag_force;
w = log10(Re);
if (0.01 < Re <= 20){
drag_force = ((24/Re)*(1+pow(10,-0.881+0.82*w-0.06*pow(w,2))));
return (drag_force);
}
else if (20 < Re <= 260){
drag_force = ((24/Re)*(1+pow(10,-0.7133+0.6305*w)));
return (drag_force);
}
else if (260 < Re <=1500){
drag_force = pow(10,1.6435-1.1242*w+0.1558*w*w);
return (drag_force);
}
else if (1500 < Re <= 12000){
drag_force = pow(10,-2.4571+2.5558*w-0.9295*w*w+0.1049*w*w*w);
return (drag_force);
}
else if (12000 < Re < 44000){
drag_force = pow(10,-1.9181+0.6370*w-0.0636*w*w);
return (drag_force);
}
else if (44000 < Re < 338000){
drag_force = pow(10,-4.3390+1.5809*w-0.1546*w*w);
return (drag_force);
}
}

[LuckyTran]

Check initial conditions and boundary conditions.

[Pedron]

Originally Posted by LuckyTran

Check initial conditions and boundary conditions.

Hello LuckyTran, thanks for your reply.
My simulation basically consists of injecting a particle at some point in the mesh and letting it flow to its speed terminal, so I've not configured in an initial or boundary condition. I've already did this test with a newtonian liquid and the results were very good, averaging 2% difference with data obtained experimentally. So, what boundary condition should I set?