VRANS vs OpenFOAM Sink Term
 

Hello everyone,

I guess this is not a specific thread only in regard to waves2Foam, but also in regard to OlaFLOW or any other solver using VRANS.
One thing that's been bugging me for the past few weeks and I can't understand is the difference between VRANS and specifying a local sink term for the momentum equation, such as OpenFOAM's explicitPorositySource. Since most models use the Darcy-Forchheimer equation, I will assume the same for the sink terms for my questions below. I think it shouldn't matter how you quantify the pressure drop, as long as you quantify it :D.

Now when there is a pressure drop, e.g. a liquid flowing through a porous medium in a pipe, then one can often quantify the pressure drop by a polynomial of some order. Often second order is enough, fulfilling a version of the Darcy-Forchheimer equation. Comparing the coefficients of the polynomial and the Darcy-Forchheimer equation, one can calculate the laminar and turbulent pressure drop and can then apply the sink term to model the porous medium.

Now, why do we need to include the porosity in the momentum equation leading to the VRANS equations. Is the sink term a problem if one wants to model the actual flow inside the porous medium, as the acceleration is wrong otherwise? What problem arises by modelling the pressure drop inside a porous medium by just a sink term. Because now I cannot simply use the polynomial equation as basis for the sink term, as the porosity in the momentum equation is changing the acceleration of the fluid inside the porous zone.

If anyone has a good read comparing both methods, I'd be happy to look around the literature, as I haven't found anything along those lines.

Thanks for your help in advance :)
Felix

Hi Felix,


It is simply the way the momentum equation comes out, when you do the correct averaging over the pore space (see e.g. https://www.sciencedirect.com/scienc...78383913001816). Furthermore, for the free surface models, you need to account for the solids in the computational cells, as your free surface does not move fast enough otherwise (see again Jensen et al. above and https://journal.openfoam.com/index.p...rticle/view/72).


Kind regards


Niels

Hey Niels,

thank you for your reply. I guess I understand, that it is the correct mathematical derivation. However, if this is the case, when can you apply the simple source term integrated into OpenFOAM (https://www.openfoam.com/documentati...orosity.html)? Or would such an application always be non-physical, as one only models the pressure drop and does not account for the blocking solids inside the porous medium?

Thanks again for your help!
Felix

Hi Felix,


I would say that it is formally non-physical, but it is for you to assess, whether the bias is acceptable for your application. Personally, I would always apply the porousWaveFoam in waves2Foam.


Kind regards


Niels