[suktan]

I created a model for heat transfer in fluid. The model (2D axisymmetric) consists of a heat source, some metal parts surrounding the heat source and fluid in the "background". They are shown in the graph below. Parts with mesh are metal parts and the heat source. This is essentially motor and coil of a loudspeaker driver.



When I used water as the fluid, transient simulation converged quickly. The results look okay and meaningful. When the fluid was air with 1/10 step size , convergence did not seem possible.


Here is my dump question. Is it typical for water vs air in cfd? Do I need smaller mesh and/or smaller step size for air?


Thanks!