Mesh independence for Multiphase Dispersed Flows
 

Hi folks,


I have been trying to understand if it is possible to obtain mesh-independent results using the solvers twoPhaseEulerFoam or reactingTwoPhaseEulerFoam.


My understanding is that these solvers are based on an averaged modeling methodology. As a consequence of such methodology, it would only be reasonable to employ meshes with grid spacings larger than the dispersed phase diameter.


According to the Multiphase Flow Handbook edited by Crowe (2006) (p. 13.18): “ computational volume must be much greater than the particle diameter for the Eulerian continuum approach to be reasonable"


However, I am unable to obtain mesh-independent results for my bubbly flow simulations considering this constrain.



Does anyone know if it is possible to achieve mesh independence for these solvers or if there is any other way to estimate the domain discretization error?


Thanks a lot