Someone can explain how can i define solution of a specie from gas phase into the liquid phase and how can i implement a chemical reaction in CFX?

hello Hossein,

There are two ways to deal with a chemical species in CFX. The choice will depend upon if the physical properties of the solution is affected by chance in the concentration of the components - dynamic viscosity and mixture density. If the change due to mass transfer and/or chemical reaction will change these properties, the Navier-Stokes equation solution is coupled with each chemical specie transport equation. If NOT, the flow is uncoupled with each chemical specie transport.

1st Case - chemical species do change the dynamic viscosity and density:

You need define each material, O2, B and C components, and the solvent (water and air is already present in the CFX library) .

Next, you need to define a Mixture formed by each material that you have previously defined (in this definition you can define the material such a composition that can vary, or a reacting mixture - this later is common to combustion simulations. This is needed due to mass transfer and chemical reaction process. Observation: The solvent material is a constrain, there is no transport equation to it, its mass fraction is calculated from other ones that must be sum 1.0.

The molar weight is need only to calculate the molar fraction in the post process ( CFD-Post) , or to some algebraic calculation such pH or other thing such that.

Repeat each step to the 2nd Mixture (gas or liquid), and if any specie can be transfer through the interface, this specie also must be part of the 2nd material. It doesn't matter if in the begin it is not present, each mass fraction is given in the initial condition, and if the specie can be transfer, its mass fraction will change from zero to some value between 0 and 1 according to the physics.

Do the two-phase flow setup selecting the two material mixtures

In the Domain setup, for each phase (fluid mixture) you can define the mass transfer process ( a user defined by an math expression), or ( if one fluid is dispersed ( a CFX defined correlation to calculate the mass transfer coefficient, e.g. Ranz-Marshall model. ) The CFX default model is such like that: Area x Km x ( mf_fluid1-mf_fluid2) and the Km is given by the CFX model or by a user defined. The Area ´s given by the interfacial area density - t here is differences among freee surface and particle models, but it is in in the manual.

Whithout the mass transfer checked, no mass will be flow from on phase to another!!

Now, you need define the way to calculate each specie: Transport Equation, Algebraic Equation, Poisson, Diffusive Transport. Chose Transport Equation for each species except the solvent - it is the constrain component. Give the kinematic diffusivity ( is not the molecular one, it is the effective or the dispersion coefficient in m2/s2

Define in B.C and the I.C. the mass fractions remebering that it must be sum equal one!

These steps define a multi-component two-phase flow with mass transfer through the interface.

I strongly advice to try fist do it without chemical reactions to avoid numerical issues.

2st Case - chemical species do not change the dynamic viscosity and density:

All steps can be used ans before, but there is another faster way: In this case the fluid flow is blinded to the chemical transport and you can run each simulation alone. Run the flow without consider the species, and in the expert parameters-Model Overrides, change the flow and turbulence to FALSE and in the solver manager give the flow solution as Initial condition. The solution will be done using the velocity field to each chemical transport equation. Or you can run all together - try the best for the case. Observation: the mixture materials and all geometric names must be the same to CFX consider as the same case

The protocol is almost the same such before else but no material or mixture material must be provided.

For each chemical specie, create a New Variable. Each one is a passive scalar, and you can define any unity such as mol/m³ kg/m³ kg/kg (non-dimensional).

Follow the steps 5-8 from the Case 1.

Define the values of each component in the boundary and initial conditions.

Try to one component and after, add others.

Again, These steps define a multi-component two-phase flow with mass transfer through the interface. But with a huge difference: there is no mass fraction because the components does not form a material. It is valid when there is low values of concentration and the presence of them does not change the density and viscosity.

Again, I strongly advice to try fist do it without chemical reactions to avoid numerical issues.

Chemical Reactions

You need define the a subdomain. In the subdomain you can define any source to any governing equation. Choose the proper equation and define its expression. Be aware with the units, if the units do not match, CFX will not accept and a error message will be shown.

in the 1st Case you will work with MASS FRACTIONS, that is kinetics such ra"' =kCaCb in molar basis is not sense. It must be given in mass basis ( kg/m³) . Often an algebraic work is needed to re-write the kinetic in the mass basis.
You can work with some kind of the Eddy Break-up (EBU) or Eddy Diffusivity Concept (EDC). It works very well when good calibrated - seek for "combustion models"...
A good practice is give a Source Coefficient. It is due numeric reasons. It will improve the conditioning of the matrix, and consequently, the convergence.
Be S(phi) the source , hence dS/dPhi is its coefficient
You can do it by means a Taylor Series
It need be NEGATIVE, always. Try some algebraic form to reach that.
If dS/dPhi is hard to calculate, tray use S/max(Phi, 10E-15). The max function is to avoid negative or zeros in the division. It is a numeric good practice. The same is valid to the whole source term. Max and Min functions to avoid nonphysical Source behavior. The use is according to case.
If the negative is not possible, let this field with zero or void.

I hope be useful to you,

Bests!

Md. Washim Akram

https://www.sciencedirect.com/search?qs=solution%20of%20a%20species%20from%20gas%20phase%20in%20the%20liquid%20phase&show=25&sortBy=relevance

https://www.sciencedirect.com/search?qs=implementation%20of%20a%20chemical%20reaction%20in%20CFX&show=25&sortBy=relevance

Hossein Jangara

Thanks Ricardo Vicente

your guidance has been so useful. solution of O2 into the liquid mixture was done successfully. now i want to implement a chemical reaction between dissolved O2 and a component (B component which exist in the liquid) within the liquid. for this purpose, i used Reactions option in the Outline Tree. is it true?

Thanks Washim

no, no.

The "reactions" are a library with some reaction regarding to COMBUSTION models. It is a predefined reaction library.to commons fuels.

You need define a source - and sources are defined into a sub-domain. The sub-domain can be all domain or part of it. The reaction it will be a math expression that you will write in the EXPRESSIONS tab. Take care with the unities, otherwise, CFX will not accept no matches in the unities of the transport equation and the source.

Once the expression of reaction has written, and the sub-domain defined. Indicate to each equation its source. Choose the square root of alpha button and write the name of the function you have written in the Expression tab. If all OK no error will be shown. If some unity are not consistent, fix that in the expression and repeat all.

I do not know anything about Eddy Break-up model. how can i get required knowledge about it?

https://pt.slideshare.net/MohammadJadidi/04-reactive-flows-eddy-disipation-model?qid=e8d8b5f5-b627-4bbb-aa46-5a000911fc02&v=&b=&from_search=2

How can i write the kinetic reaction rate (-ra=kCaCb) in the expression tab?

Or how can i write that in mass basis?

The expression is in that form -kCaCb, using the variable names. If you are working with a mixture with variable composition, you need to change to mass basis. And define the subdomain to associate the reaction source.

If A and B are passive scalars, you can work in the molar basis, and the procedure is the same-> subdomain->reaction source

What is the by-product of ozone and fumaric acid reaction?

Do anyone know the chemical reaction between ozone and fumaric acid (kinetic and products of this reaction)?

Do anyone know the diffusion coefficient of fumaric acid in water?

How can i define liquid side Reynolds number in the CFX for bubble column?

what is the simplest definition of ion exchange capacity

what is the simplest definition of ion exchange capacity

Can we classify HV and LV lines based on line length?

Irradiation data for JAPAN?

Resistor definition in Piezoelectric bimorph with ANSYS workbench

How do I increase the yield of my PCR product?

How the parameters (such as Mw, Chemical bonds, polymer bland and so on) affect the mechanic property of polymer (such as storage, loss modulus) ?

Waters 2707 HPLC autosampler problem?

Why are the velocities so high in my Boussinesq approximation Fluent simulation?

What is the characteristic length to determine the Reynolds number for the scattered turbulent fluid flow?