CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Phase change with VOF (http://www.cfd-online.com/Forums/openfoam-solving/57849-phase-change-vof.html)

shu June 13, 2005 07:16

Hi! I want to set up a comp
 
Hi!

I want to set up a computation region with two domains, one with the conductive heat transfer in the wall under the liquid, and the other with the conductive as well convective heat transfer in the liquid above the wall. My questions:

1. How can I set up these two domains? In the tutorial case "scalarTransportFoam/pitzDaily", it was only one domain.

2. What does the parameter "DT" mean in the following line
DT DT [0 2 -1 0 0 0 0] 0.01;
from the file ~/OpenFOAM/OpenFOAM-1.1/tutorials/scalarTransportFoam/pitzDaily/constant/transpo rtProperties? If it is kinematic viscosity (with the unit [m^2/s]), how can I introduce two different thermal conductivities?

3. Shall I create a new solver for the heat transfer equation with conductive and convective terms? The flow is laminar and unsteady.

4. Further, I will have a bubble on the wall in the liquid, which grows and detachs because of evaporation (pool boiling). To simulate such a problem, the source terms are needed in the mass-conservation equation, in the heat transfer equation, and in the VOF-equation too. Is it possible to do it with OpenFoam? (two phase, laminar flow)

Many thanks,

Bitan

henry June 13, 2005 16:26

It is possible to do all you r
 
It is possible to do all you require with OpenFOAM but it will require a substantial effort and you might want to consider arranging a support contract particularly if you would like OpenCFD to do some/all of the implementation for you.

hjasak June 13, 2005 22:47

Here's some help: 1) conjug
 
Here's some help:

1) conjugate heat transfer simulations in FOAM (as it stands) are not completely straightforward. There are several possible approaches, but you need to know what you are doing. There is no tutorial for this.

2) scalarTransportFoam only solves the standard transport equation for a scalar, given a prescribed velocity field. Thus, kinematic viscosity is not needed as the flow equations are not solved. DT that you mention is the diffusivity (conductivity) for the scalar you are solving for.

3) I would advise studying the tutorials and the code until you get a better grasp of your problem. You will then need to write your own solver, tailored to the problem you are trying to do.

4) Two-phase laminar solver with the VOF-like approach already exists in FOAM. It is called interFoam and there is a tutorial for it. In order to do the evaporation, you will need to add various terms to the original equations + solve for the heat transfer; in order to do that successfully, you will first need to understand the current implementation. That's where I would start.

Good luck,

Hrv

shu June 17, 2005 07:09

Thank you very much for the an
 
Thank you very much for the answers and the advices!

To Henry:
I am not sure, if we have budget to give such a contract. But I will discuss it in my team.

To Hrvoje:
I have created a solver for the heat transfer equation with the convective and conductive terms. It works fine.
As the next step I will try to combine it with interFoam.

Has anyone set up two or more domains with the different properties, e.g. the thermal conductivities, or densities?

Bitan

khleitz May 25, 2007 03:13

Hallo Bitan, I am working on
 
Hallo Bitan,
I am working on a similar problem like you. I want so simulate the melting and evaporation of a solid. Therefore I have alread combined the lagragian heat transfer solver with interfoam. The melting works quite nice however I have no idea how to implement the evaporation. Did you have success in implementing the evaporation in the interfoam solver?
If yes, where do I have to add the source terms?
Best reagards,
Karl-Heinz

shu May 29, 2007 11:29

Hallo Karl-Heinz, I model t
 
Hallo Karl-Heinz,

I model the evaporation with some terms in all the conservation equations, i.e. mass, Navier-Stokes, energy, and VOF-equation. The terms is derived from the rate of evaporation which depends on the heat transfer at the phase interface and the latent heat.

However I have problem with the smeared phase interface described with VOF. Because in this zone the heat conductivity can not be determined correctly. And then false evaporation rate turns out. For this reason I am working on Level-Set-Method. Maybe this problem will not be so serious in case of melting, because the ratio of heat conductivities is smaller then that in case of evaporation.

Regards,
Bitan

rajeevrkris January 30, 2008 23:55

I am new to OpenFoam. I am sol
 
I am new to OpenFoam. I am solving for hot air flow (1000K) at a velocity of 5m/s on top of water (300K)flowing at 2 m/s in a converging duct. How shall I model evaporation of water into air stream using volume of fluid in interFoam? Can anybody help me out?
Thank you,
Rajeev Krishnan

kar February 1, 2008 05:54

> Has anyone set up two or mor
 
> Has anyone set up two or more domains with the different properties, e.g. the thermal conductivities, or densities?

Looks like post from somebody who's done that would be appreciated! I'm also interested..

shu February 1, 2008 10:23

I have submitted a paper to th
 
I have submitted a paper to the conference CHT-08, in which a model for the phase change is described. To model the phase change some additional terms are introduced into the equations. Film boiling can be simulated successfully. It is based on interFoam.

Bitan

arghaz July 21, 2008 12:37

Hello Bitan Shu, I'm workin
 
Hello Bitan Shu,

I'm working on modelling Film Boiling and I'm interested by your CHT-8 paper. Thus, can you send me an example.

Arghaz

shu July 23, 2008 10:18

Hi Arghaz, it would be a pl
 
Hi Arghaz,

it would be a pleasure for me to send you the paper. Please drop me an email: shubitan <at> gmx dot net

Bitan Shu

mahaputra February 10, 2009 16:06

Hi Bitan Shu, what about co
 
Hi Bitan Shu,

what about condensation? phase change from steam (water vapour) into liquid phase ?

does your solver work for this case?

i have a problem within phase change with the multiphase flow.

could you please guide me to solve the problem?

and could you make a tutorial on internet for your case of film boiling?

kindly please help me

regards

Nugroho Adi

haghajani March 5, 2009 15:22

Dear all, I want to simulate
 
Dear all,
I want to simulate Liquid Hydrogen release under high pressures, and am investigating if the multiphase solvers can helpful;

Liquid Hydrogen evaporates shortly (flash evaporation/boiling Temperature 20K) after releasing in to atmosphere and I am thinking to a solver based on "ras/interfoam" capable of handling with two phase flow + evaporation.

Would you please let me know, where/How I should modify, if the idea feasible?

Any other suggestion, I will be thankful;
Hamed Aghajani
hamed (dot) aghajani (at) gmail (dot) com

karthik April 17, 2009 14:17

help plz
 
actually i plan to simulate phase change in centrifugal casting.
so basicall 3 phase
1>air that wil fil the rest of the domain and interact with the molten metal and involve in heat transfer
2>molten metal that will spin and develop its own flows
3>molten metal that wil interact with the mould and get transformed to solid

4>the solid rotating mould with the sourrounding atmosphere(dunno bout tis

an my self very confuse regarding the mesh and the models, relaxation etc)

so if any of u guys can enlighten me if this problem CAN ACTUALLY BE

SIMULATED
.i hav already obtained the fluid flow patterns alon in Fluidyn

unfortunatly fluidyn doesnot support heat transfer with phase change hence i

plan to use openFoam(i know verry little of foam)so plz help me out guys

isabel June 16, 2009 03:12

nucleate boiling
 
I want to simulate nucleate boiling with OpenFOAM for my doctoral studies. Can anybody send me papers or information to lamasgaldo@yahoo.es

karthik June 17, 2009 13:51

it can be done but am not sure of wat solver to use :( sorry

isabel June 22, 2009 11:03

I am studying interFoam solver and I have two doubts:

- What is pEqn.H ?
- Where is mass conservation equation defined? It would be gradient(u)=0

sandy July 8, 2009 00:25

Quote:

Originally Posted by isabel (Post 220071)
I am studying interFoam solver and I have two doubts:

- What is pEqn.H ?
- Where is mass conservation equation defined? It would be gradient(u)=0

==================

To a phase change VOF problems, div(u) =? 0, I don't know too.

However, in the InterPhaseChangeFoam solver, div(u) = (1/rho1 - 1/rho2) * (mDotP[0] -mDotP[1]) * (p - pSat) = (vDotcP-vDotvP)*(p - pSat), so in pEqn.H of this solver, the source term of pdEqn become into (vDotcP-vDotvP) * (pd - pv[=rho*gh - pSat]), and the (vDotcP-vDotvP) * pd is implicitly solved as the Sp term of pdEqn and (vDotcP-vDotvP) * pv is a explcit term as the Su term of pdEqn. Am I right?

But, I don't know why this equation could not be solved convergently, after I activated the cavitation model to it and gammaEqn?

Who know the reason? Please help me out. Thanks a lot.

isabel July 16, 2009 06:23

Thank you very much, sandy. In the interPhaseChangeFoam, the terms vDotcP and vDotvP are divided by the pressure, so when you solve the continuity equation, you multiply for the pressure again to have the units 1/s.

Then, my equation which is

div(U) = source

in pEqn.H of the interPhaseChangeFoam solver would be typped as:

fvScalarMatrix pEqn
(
fvc::div(phi) - fvm::laplacian(rUAf,pd) - source
);


Am I wright?

sandy July 16, 2009 08:42

Quote:

Originally Posted by isabel (Post 222894)
Thank you very much, sandy. In the interPhaseChangeFoam, the terms vDotcP and vDotvP are divided by the pressure, so when you solve the continuity equation, you multiply for the pressure again to have the units 1/s.

Then, my equation which is

div(U) = source

in pEqn.H of the interPhaseChangeFoam solver would be typped as:

fvScalarMatrix pEqn
(
fvc::div(phi) - fvm::laplacian(rUAf,pd) - source
);


Am I wright?

I think, in interphasechangefoam, the source term is that the terms vDotcP and vDotvP are divided by the rho (1/rho1 - 1/rho2), so

div(U? or phi)= source = (p - pSat) * ( vDotcP - vDotvP).

If it is "phi" but "U", whether or not we also need to interpolate RHS to the face? Why not?


All times are GMT -4. The time now is 22:46.