CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM (http://www.cfd-online.com/Forums/openfoam/)
-   -   Piso / B-Piso and linearisation (http://www.cfd-online.com/Forums/openfoam/79789-piso-b-piso-linearisation.html)

noramat September 3, 2010 13:42

Piso / B-Piso and linearisation
 
Hello all. Currently i work with mhdFoam and combined it with an boussinesq-approximation like in boussinesqPisoFoam. My question is: how the piso-algorithm works in the first step in detail (the momentum predictor step). I thought in that step you take the old values of temperature, pressure and velocity to predict the new velocity and correct the values in the next steps. But how to handle the magnetic value B!? I think it's also an unknown value, which also has to be calculated implicit, like u, right? Or do i have to use also old values of B to predict the new velocity and later update it with the induction equation and the calculated velocity u? Also i have a question to linearisation.. In some articles i already found how to linearised the nonlinear convection term (use old value of velocity). I have to do the same with the terms in fvc::div(phiB, 2.0*DBU*B) and in fvc::grad(DBU*magSqr(B)) , but not in the energy-equation in fvm::div(phi, T), right? I need to know it, because i try to write out the finite-volume-discretisation using in OpenFOAM. For temporal discretisation i use the implicit Euler-Scheme and try to understand several steps. Thanks a lot! noramat


All times are GMT -4. The time now is 00:43.