[TomM]

Hey guys,

I'm simulating natural convection of a melting phase change material in a vertical cylinder. A hot water stream heats up the solid PCM around.

My problem is, if I leave the residuals at default settings (10e-3 and 10e-6 for energy) for a time step calculated by the rule of thumb of Ansys, the solution converges after a while for the first Iteration and as a result there is no big progress in the simulation. If I force Fluent to iterate each time step with 20 Iterations the result is totally different, the melting goes much faster.

Do I have to change the residuals to a lower number or what would you guys suggust? I'm a bit lost here