Mass Transfer
 

I´m simulating a 900W energy source in a domain with liquid water and vapour water, that will heat the fluids. I know that the fluids will mix, be turbulent and evaporate next to the source and condensate near the walls (phase change).

I choose the Particle model (if you think I should select a beter one pls tell me)

This are my setup options:

Transient Simulation with 0,1s of iteration step time.

Basic Settings:

Liquid Water - Continuous Fluid ; Vapour Water - Dispersed Fluid with 0,05 mm (mean diameter)


FLUID MODELS:

Multiphase: Inhomogeneous

Heat Transfer: Total Energy (Including Viscous Work Term)



FLUID SPECIFIC MODELS:

Fluid Buoyancy Model: Density Difference

Turbulence: Liquid Water - k-epsilon; Vapour Water - Dispersed Phase Zero Equation



FLUID PAIR MODLES:

Interphase Transfer: Particle Model

Momentum Transfer: Drag Force - Ishii Zuber

Turbulence Transfer: Sato Enhanced Eddy Viscosity

Phase Change Model - Thermal Phase Change

Heat Transfer : Two Resistance :::Liquid Water - Ranz Marshall; Vapour Water - Zero Resistance


If I solve with this parameters all goes right but then, when I introduce mass transfer:

Mass Transfer - Phase Change

The momentum components of the liquid water diverges allot. it only works if the mean diameter of the dispersed fluid is 1mm, which I think it´s too big comparing with the reality. Does anybody have any idea how to solve this problem with mass transfer?


Is the mass transfer really needed to completely simulate the phase change between the waters?

Yes, mass transfer really is needed to simulate the mass transfer between the phases. Kinda obvious I would think.

The Laplacian perssure of a 50um bubble will be quite large, and a 1mm bubble quite small. This will result in a large superheat - are you sure this right? Most bubbles I see in my kettle are closer to 1mm than 50um. Are you sure your bubbles are 50um? Are you just guessing or do you have evidence to show it should be 50um?

My recipient is about 35 mm long so I thought 1mm diameter will be big, but yes, I´m just guessing.

When I work with 1mm diameter, my vapour water volume fraction goes to 1*10^-15, which is too small, and the pressure get´s higher that comparing with reality. But it is the only way that the solver doesn´t diverge.

And I really think that is the nearest model to my situation, cause mixture is for two continuos fluids, the particle trasnport doesn´t apply, and homogeneous neither.

The only think that is missing is a portion of air, that in reality it exist in the recipient ( that could explain the pressure higher than usual)but when I add air the software diverges too.

What can I do?

I can work with 1mm, but what is the best way to add air? With variyng composition mixture of vapour and air as dispersed fluid? or just 2 dispersed fluids, water vapour and air separated.

Have you done the phase change CFX tutorials? Have you done any validation or verification of your boiling model? In other words, have you got some quality benchmark results to check your modelling technique is good before you commit to your actual model?

There's where I need help, how can I test if this model fits my simulation?

I cannot say whether your model is good or not. But part of the skill of a CFD practitioner is to be able to work that out for yourself. That's what my previous post tried to explain.

You do the tutorials so you understand how to set the models up in CFX. Also read the documentation on the relevant models so you know what they can do and their limitations.

You find some quality benchmark results in the literature which is similar to your case and you model their results. If you can accurately reproduce their results you know you are in a good position to successfully complete your actual model. I guarantee that this will not be easy - so the time spent getting this right will be worth it because then you will not waste your time producing results which you do not know if they are correct or not.

I have done the tutorials that envolve multiphase. Can you name which one you are talking about?

The ones using boiling models.

Have you found any benchmark results for boiling similar to your case?

I have not found any tutorial that simulates boiling, the only one that is close my simulation is steamjet.


I have a 3 cm recipient with liquid water and air. Then I have a 900W energy heat source in the middle, and water vapour forms near the source and condenses near the cooled walls. I expect that all the fluids mix and phase and mass transfer between the gas/liquid water occurs. Notice that I expect a average pressure near to 30 bar, so there should be like 85% or more of liquid water and little gas.



For what I saw in my research I can´t decide if the best option is (between this 5):

- Thermal Phase Change Model with Particle Model: (OPTION1)

Variable Composition Mixture (Air and Water Vapour) Dispersed
Fluid)

Fixed Composition Mixture ( Liquid Water) (Continuous Fluid)

Mass Transfer between the two mixtures activated (even though air
is in it).



- Thermal Phase Change Model with Particle Model: (OPTION2)

3 Fluids: Air (Dispersed Fluid), Water Vapour (Dispersed Fluid) and
Liquid Water (Continuous Fluid)

Mass Transfer between only, Liquid Water and Vapour Water
activated)



- Thermal Phase Change Model with Particle Model: (OPTION3)

2 Materials: H2Ovl (Homgeneous Binary Mixture) (Continuous Fluid)
and Air ( Dispersed Fluid)


Mass Transfer between the two fluids deactivated because it is
already happening in the HBM)


- Droplet Condensation Model with Particle Model: (OPTION4)

3 Fluids: Air (Dispersed Fluid), Water Vapour (Droplets (Phase
Change) ) and Liquid Water (Continuous Fluid)

Mass Transfer activated between Liquid and Vapour Water
Phase Change --------------- Small Droplets


- Droplet Condensation Model with Particle Model: (OPTION5)

3 Fluids: Air (Dispersed Fluid), Water Vapour (Continuous Fluid) and
Liquid Water (Droplets (Phase Change))

Mass Transfer activated between Liquid and Vapour Water
Phase Change --------------- Small Droplets

Here is one which looks useful: https://support.ansys.com/AnsysCusto...ANSYS+CFX+16.0

Also look in the CFX-Pre documentation (Ch 37, library objects) on the boiling water template.

These documents describe the models available in CFX in detail and should help you choose which option to go ahead with.

I am a student so I don´t have acess to the Customer Portal. I tried to register but it only lets me register if I have a private Company email. Can you send it to my email? p.de.oliveira@hotmail.com

About the boling water template, for what I read it doesn´t add nothing new to what I know already.

Thanks anyway Ghorrocks for always trying to help. I´m going to simulate the different cases and compare them with reality to know which fits better.

If your school/university has a research license then you should be able to access the customer portal through that. If you are a student you will be able to access the ANSYS academic website, but that does not include advanced modelling such as boiling models.