[virothi]

Hi all,,
You may find this question funny or easy, but I have to ask.
Imagine a cube (fluid: air) and there is a small inlet on the left side where a specific gas (say methane) with temperature 400k with pressure 2 bar. It expands inside the fluid domain. The rest are walls and we have an outlet on the right side.
My question: how to define the cube as air and how to provide the inlet with methane?

[aerosayan]

What you're defining is basically a flow from a pipe to a large domain.


You may find your problem statement to be very similar to that of a rocket nozzle simulation.


https://youtu.be/qPYso99wMyw




Edit : Found a better solution for you. Since I don't do gas injection simulations, I didn't know of examples.


Search "CFD gas injection" in youtube.

https://youtu.be/wTwd1xYhgKI


Here's an ANSYS tutorial on methane combustion : https://youtu.be/Miib-owkq-Q https://youtu.be/1ncUfEVcfbc

[sbaffini]

Your problem requires a multi-species formulation, or any other mean to track, for any given cell/volume, how much air and how much methane you have.

Without going into the details of specific models and staying on the math side, this means you need an additional pde to specify this information, including in bc, and the rest of your solver needs to be able to manage that in computing properties, etc.

[virothi]

Sayan, thank you for your quick reply.
But I don't model combustion. I just want to simulate heat transfer.
I just want to make it clear.
When I define the cube as a fluid domain, I can only choose either air or methane as a fluid.
But I need the cube with air and there should be a methane inlet. Is this is possible with fluent?

[virothi]

Dear Paolo,

I strongly believe that you gave good information, but honestly, to my knowledge, I understood a little.

If I understand you correctly, you recommend me to use Mutispecies flow. I thought about this, but as I said earlier,
I have a cube with quiescent air (there is no inlet for air, I want to define the cube is containing air), and a methane jet comes into it through a small inlet (on the right side of the cube with a small imprinted face). I simply didn't understand or I couldn't find a way to define these conditions.

Any suggestions?

Originally Posted by sbaffini

Your problem requires a multi-species formulation, or any other mean to track, for any given cell/volume, how much air and how much methane you have.

Without going into the details of specific models and staying on the math side, this means you need an additional pde to specify this information, including in bc, and the rest of your solver needs to be able to manage that in computing properties, etc.

[sbaffini]

Provided that Fluent has the multi-species feature and you can use it straightforwardly for your problem (just look the manual and/or some tutorial), I want to give you a grasp of how this would work in general, so that you understand better how to move.

The idea in multi-species formuation is that, besides the classical NS equations you also have N-1 additional equations (with N total number of species) that will track, for each cell, the fraction of a given specie with respect to the total mass. While the model is typically used for combustion, you can just avoid defining any reaction between the species and you will just track them in their transport and mixing.

The fact that you have additional N-1 equations also means that you have to specify boundary and initial conditions for them.

In your case of 2 species, air and methane, this will imply an additional equation with its own initial and boundary condition. In multi-species approach you are free to assign the additional equations to whatever specie you want but, it is common practice to assign them to the less abundant ones and leave the most abundant specie tracking to the global mass equation (as a complement to one).

For you this translates to assigning the additional equation to the methane mass fraction. Which, in turn, means that the additional equation (methane mass fraction) will have a 0 initial condition in the domain (as it is all air) and a 1 as inlet condition (it is only methane that enters the domain). It is just that simple.

The rest of the matter is how to tell Fluent to use this model and how to specify the material properties of the two species and the mixture. But that's what manuals are for.

[virothi]

Dear Paolo,

Thanks for your quick reply.

Now, I understood better. I'll get back to you once I finish or after I achieve some progress or if any new questions arise.

Once again, thank you

Originally Posted by sbaffini

Your problem requires a multi-species formulation, or any other mean to track, for any given cell/volume, how much air and how much methane you have.

Without going into the details of specific models and staying on the math side, this means you need an additional pde to specify this information, including in bc, and the rest of your solver needs to be able to manage that in computing properties, etc.

Originally Posted by sbaffini

Provided that Fluent has the multi-species feature and you can use it straightforwardly for your problem (just look the manual and/or some tutorial), I want to give you a grasp of how this would work in general, so that you understand better how to move.

The idea in multi-species formuation is that, besides the classical NS equations you also have N-1 additional equations (with N total number of species) that will track, for each cell, the fraction of a given specie with respect to the total mass. While the model is typically used for combustion, you can just avoid defining any reaction between the species and you will just track them in their transport and mixing.

The fact that you have additional N-1 equations also means that you have to specify boundary and initial conditions for them.

In your case of 2 species, air and methane, this will imply an additional equation with its own initial and boundary condition. In multi-species approach you are free to assign the additional equations to whatever specie you want but, it is common practice to assign them to the less abundant ones and leave the most abundant specie tracking to the global mass equation (as a complement to one).

For you this translates to assigning the additional equation to the methane mass fraction. Which, in turn, means that the additional equation (methane mass fraction) will have a 0 initial condition in the domain (as it is all air) and a 1 as inlet condition (it is only methane that enters the domain). It is just that simple.

The rest of the matter is how to tell Fluent to use this model and how to specify the material properties of the two species and the mixture. But that's what manuals are for.

[virothi]

Dear Paolo,
I've done multi species flow and initialized the solution to run. But I had to stop it in-between. Could you please comment on the below problem?

I modeled a conjugate heat transfer but I'm not certain about the results. Your opinion and suggestions really matter.

I have an 18mm dia cylinder(solid at 293[K]) and it is placed inside a cube (fluid at 293[K]). I have a pressure inlet (CO2 gas) with a high temperature (1273[K]) and Pressure 2[bar].

1. I've used the k-omega SST turbulence model for fluid flow.

2. I've used the 'coupled' condition at the interface created by the fluent. Since I used the 'coupled' option, I gave the heat flux as 'zero' at the cylinder wall. I've patched all the domains with the temperature '293.15[K]' during initialization.

I stopped the simulation in-between.

Reason: I've seen the rise in temperature of the solid cylinder by 1 degC for 0.17 sec. Is this is possible? Please comment with your experiences.

Originally Posted by sbaffini

Provided that Fluent has the multi-species feature and you can use it straightforwardly for your problem (just look the manual and/or some tutorial), I want to give you a grasp of how this would work in general, so that you understand better how to move.

The idea in multi-species formuation is that, besides the classical NS equations you also have N-1 additional equations (with N total number of species) that will track, for each cell, the fraction of a given specie with respect to the total mass. While the model is typically used for combustion, you can just avoid defining any reaction between the species and you will just track them in their transport and mixing.

The fact that you have additional N-1 equations also means that you have to specify boundary and initial conditions for them.

In your case of 2 species, air and methane, this will imply an additional equation with its own initial and boundary condition. In multi-species approach you are free to assign the additional equations to whatever specie you want but, it is common practice to assign them to the less abundant ones and leave the most abundant specie tracking to the global mass equation (as a complement to one).

For you this translates to assigning the additional equation to the methane mass fraction. Which, in turn, means that the additional equation (methane mass fraction) will have a 0 initial condition in the domain (as it is all air) and a 1 as inlet condition (it is only methane that enters the domain). It is just that simple.

The rest of the matter is how to tell Fluent to use this model and how to specify the material properties of the two species and the mixture. But that's what manuals are for.

[sbaffini]

Not really possible to give any advice on such specific results. However, no bc has to be set on the coupled interface between the coupled zones