interFoam alphaEqn: TVD+MULES - why both?
 

Hello Foamers,

I have a question concerning the alphaEqn of interFoam solver (version 2.2.1):

Why does interFoam first use TVD schemes for the fluxes, which guarantees a boundedness, and then additionally use the MULES limiting, which is made for ensuring a bounded solution?

Looking at Jasak's and Damian's thesis, I read that TVD limits the face value whereas MULES (=FCT) limits the face flux. The methods work differently, however, at the end both are having the same objective: a bounded solution.

... I don't exactly understand why we need both? In which case would there be an unbounded flux that needs to be clipped with the MULES limiter?
Or is MULES only needed, when non-TVD schemes are used?

Can anyone maybe help me understanding this?:o
Any hints are much appreciated!

Regards,
Lydia

Hi Lydia, MULES is Flux Corrected Transport. This method blends a Hi-Resolution flux with a Low-Resolution one in order not to exceed local and global maximum. The low resolution (and ever bounded) flux is given by the Upwind scheme, otherwise the Hi-Res flux is given by any scheme you want (central-difference or TVD). Then, the lambda factors for blending are calculated in order to give a bounded solution.

With FCT (MULES) you have better chances to have a bounded solution, specially in multidimensions (I don't exactly why, but it's clear in practice).

Regards.