Convergence Problem with Cavitation Model
 

I'm working at tilting pad journal bearings flow analysis with very good results for big eccentricity, that means rotor wall and pad are very close.

If I use single phase model and i reduce eccentricity, some problems occur.
Contourn plot of pressure on surface shows that there is a zone where pressure is negative. This means that pressure in that zone goes under operating pressure (i've set operating pressure at 1 bar).
This is caused by divergent region or by starvation (air is called in the domain).
In Hydrodynamic Lubrication theory this problem is resolved by adding a BC called "Half Sommerfeld" that set to zero all negative pressure.

I've tried to use mixture model with Cavitation model "Schnerr and Sauer" but convergence seems to be unreachable! I've continuity residual near to 0 and the integral of pressure on the fixed surface decrease untill zero, that is nonsense.


I prefer to find a solution with Cavitation model because Half Sommerfeld condition not garantee continuity. I've tried with SIMPLE, PRESTO and Second Order and change URF values.


I've also tried to write an UDF to apply HS BC but it gave a parse error at line 1.


#include "udf.h"
DEFINE_EXECUTE_AT_END(HALF SOMMERFELD)
{
Domain *d;
Thread *t;
cell_t c;
face_t f;
d = Get_Domain(1);
thread_loop_c(t,d)
{
begin_c_loop(c,t)
{
if (C_P(c,t)<=0)
{
C_P(c,t)=0;
}
}
end_c_loop(c,t)
}
thread_loop_f(t,d)
{
begin_f_loop(c,t)
{
if (C_P(c,t)<=0)
{
C_P(c,t)=0;
}
}
end_f_loop(c,t)
}
}

I've tried with Singhal cavitation model. Although the Fluent Guide says it's more instable then "Schnerr and Sauer" and the other one, this time it converge and pressure results seem to be correct.
The problem is that it has a very slow convergence.

My real doubt is that Singhal can be used to modelling dissolved gas cavitation but how can i modelling air coming in the domain by underpressure?