implicit - (fvc::grad(U) & fvc::grad(muEff))

Kareem Abdelshafy · May 13, 2016, 19:15

Hello Foamers

I am trying to model two phase flow (water and high vicious fluid ) inside a micro-channel. After dimensionless, the reynolds number becomes very low. as far as I know, if the diffusion term is solved fully explicit , a new time step constrain for the diffusion term becomes dx^2 *Re which is very small in my case.

In most multiphase solvers in openfoam , the diffusion term are treated semi-explicit ( - (fvc::grad(U) & fvc::grad(muEff)) is explicit) which constrain my time step to unpractical time step. Otherwise, if I increased the time step with even Co number < 0.1 , the solver diverge. The solver works fine, if I removed this term, but the result becomes unphysical.

My question. Is there any suggestions to increase the time step or make this term - (fvc::grad(U) & fvc::grad(muEff)) implicit in openFoam ?

Thanks a lot

Kareem Abdelshafy · August 20, 2016, 21:57

I managed to solve the problem. Adding these two terms in the UEqn enhanced the solution stability.

- fvm::laplacian(max(MuEff), U)
+ fvc::laplacian(max(MuEff), U)

May 13, 2016, 19:15	implicit - (fvc::grad(U) & fvc::grad(muEff))	#1
Kareem Abdelshafy New Member Kareem Abdelshafy Join Date: Feb 2016 Location: Boston MA USA Posts: 4 Rep Power: 10	Hello Foamers I am trying to model two phase flow (water and high vicious fluid ) inside a micro-channel. After dimensionless, the reynolds number becomes very low. as far as I know, if the diffusion term is solved fully explicit , a new time step constrain for the diffusion term becomes dx^2 *Re which is very small in my case. In most multiphase solvers in openfoam , the diffusion term are treated semi-explicit ( - (fvc::grad(U) & fvc::grad(muEff)) is explicit) which constrain my time step to unpractical time step. Otherwise, if I increased the time step with even Co number < 0.1 , the solver diverge. The solver works fine, if I removed this term, but the result becomes unphysical. My question. Is there any suggestions to increase the time step or make this term - (fvc::grad(U) & fvc::grad(muEff)) implicit in openFoam ? Thanks a lot

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Can anybody give a introduction of the term "point implicit" and 'line implicit' ?	Tommy Chen	Main CFD Forum	2	August 3, 2015 04:30
Implicit source term and porousInterFoam	Andrea_85	OpenFOAM Programming & Development	0	August 30, 2013 05:30
Help. implicit discritization	dearboy	Main CFD Forum	0	November 25, 2010 22:46
how implicit is implicit?	Joe cool	Main CFD Forum	7	July 16, 2004 12:10
UDF : Chemical Reaction in Catalytic Converter	adhimac	FLUENT	1	March 27, 2001 15:04

August 20, 2016, 21:57		#2
Kareem Abdelshafy New Member Kareem Abdelshafy Join Date: Feb 2016 Location: Boston MA USA Posts: 4 Rep Power: 10	I managed to solve the problem. Adding these two terms in the UEqn enhanced the solution stability. - fvm::laplacian(max(MuEff), U) + fvc::laplacian(max(MuEff), U)