[vinaynitrkl]

What should be done to show that the column to be used is already filled with water?

Is this what is called patching?

Please explain the step by step procedure.

[-mAx-]

patching is more usefull if you want to initialize a subdomain, ie: zone (not all the domain) with a value.
There are some options, but basically you need to specifiy the zone you will patch (for instance in gambit create a volume which describes your column, and define a separated fluid_domain on this volume). Then when you want to patch, just select this fluid domain
Check help, especially Chapter 25.14.2 "Patching values in selected cells"

[vinaynitrkl]

thanx..
So i assume you are saying if the column is fully filled with water(or say any fluid) nothing needs to be done?

If u define certain material does fluent assume it to be fully filled in the domain?

[-mAx-]

if your domain consits in the column, and if the entire column is filled with a fluid, then you don't need to use patch.
If you are working with VOF, then set the right VOF-value in the initialization

[vinaynitrkl]

thanx for the clarification!

i have some more trivial doubts if you would like to answer..

Is 2D rectangular mesh apt for solving gas-liquid (O2 species) mass transfer?
I am using segregated-unsteady solver.
Or should i go with 3D mesh and steady state solver?

i posted the question a few days back..
http://www.cfd-online.com/Forums/flu...as-liq-mt.html

Also now i am defining a source term for mass in liquid and the gas phase.
But i am unsure of the values i should put in their.(signs and the conventions)
Please elucidate!
If my approach is wrong please correct me.

Any help is thoroughly appreciated.

[-mAx-]

I have no experience, and no idea about mass transfer

[vinaynitrkl]

Anyone else please?

[CFDtoy]

using source / sink terms, running mass transfer calculations can be done using any mesh any dimension (1D, 2D, 3D).


Signs / convention

Linearise your source term..if negative elements exist (good for your stability), push it to the left side of the transport eqn.

What is + and what is - in the source ..only you can answer.

For example, lets say liquid is converted to vapor - boiling, in this case, mass transfer for the liquid is - and for the vapor ..its + ..so, designing the source/sink terms is "modeling" and its a unique challenge for CFDer's


Good luck.

Originally Posted by vinaynitrkl

Anyone else please?

[vinaynitrkl]

thanx CFD toy

that was helpful...

One more pls...
I defined liquid as primary phase and air as secondary phase..
But when in species model i define phase properties of the two phases as mixture-template to activate species in the two phases, both primary and secondary phase turn to mixture-template. Is it ok?
Since i want only liquid phase to be present in column initially.

[CFDtoy]

Any model, mixture (VOF etc), eulerian multi-fluid can be used. When you patch, you indicate the vf of a given phase and appropriately the properties are calculated. For example, vf=0.5 each phase => viscosity = alpha1*primary_phase_visc1+alpha2*secondary_phase_ visc2 etc

same applied for other properties (cp, rho, etc).

If you patch liquid to the domain, only liquid properties will be present and basically, if you do not want to solve for species in liquid, then you should put a condition in your udf that if vf_liquid > 0.5 etc..no mass transfer etc.

Hope this helps.

Originally Posted by vinaynitrkl

thanx CFD toy

that was helpful...

One more pls...
I defined liquid as primary phase and air as secondary phase..
But when in species model i define phase properties of the two phases as mixture-template to activate species in the two phases, both primary and secondary phase turn to mixture-template. Is it ok?
Since i want only liquid phase to be present in column initially.