[mhmt]

Dear all,

Shingal model takes into account the surface tension (σ) value when modelling the source terms Re and Rc as seen from the Fluent theory guide.



Since surface tension term is placed at the denominator of the equations, when supplying zero, the source terms is divided by zero. The solver should diverge and give a floating point exception error normally. What is strange, in a forward facing step test case, the solver is running somehow and does not diverge.

However, when I write the same model in a UDF with zero sigma, code diverges immedeately.

Is it possible a different implementation of the above formulation is coded in Fluent?