Learning Forum

[F123rooq]

Can you please walk me through setting up my 2D geometry for supercritical co2. I have a box, with 2 pipes sticking out of it. The two pipes will have ScCO2 coming to the box which has air in it, and mixing.

I had typed in a code that I found online into ansys, and it gave me the option to use the NIST, so I selected co2.fld file, then it said that calculations will be performed for the vapour phase. I have selected multiphase, and then Volume of Fluid. I have read someone did a similar study, and they used VOF to simulate the stream of jet that I want too.

I have chosen my inlets, my outlets

I have two phases, one is air, and other real gas co2 from the refprop etc. The problem is that my box, and the two pipes have 2 phases in the cell zone conditions. But I only need phase 1 (air) in the box, and phase 1 and 2 in the pipes together, because they open to the atmosphere, so air would already be mixed in the pipes I presume, I don't expect it to be 100% ScCO2 in the pipes. But if having one phase (ScCO2) in the pipes makes the setup easier, then I am happy with that for the time being.

[DrAmine]

Why using multi-phase model if you do have two gases? Use Species model and define a multicomponent mixture of air and Co2.

[F123rooq]

Species model is not highlighted, it won't let me select it

[DrAmine]

Start from scratch enable energy equation than enable specie model. Afterwards you enable multicomponent nist as in the docu.

[F123rooq]

The mixture material does not have co2 and air, so I closed species, enabled the nist mode real gas co2 etc. and the species disappeared. I wanted to put the mixture material as co2 and air

[DrAmine]

Then define a nist multispecie material

[F123rooq]

If there is not equation of state for Supercritical CO2, how are people simulating Supercritical CO2 in ANSYS. I have an error message No mixture data are available for one or more binary pairs in the specified mixture. The mixture is outside the range of the model and calculations will not be made

[F123rooq]

Changing method of "Density" to "real-gas-mixing-law" -- data required.

Changing method of "Cp (Specific Heat)" to "real-gas-mixing-law" -- data required.

Changing method of "Thermal Conductivity" to "real-gas-mixing-law" -- data required.

Changing method of "Viscosity" to "real-gas-mixing-law" -- data required.

 

Changing method of "Mass Diffusivity" to "multicomponent" -- data required.

[DrAmine]

Material properties are your Input and your responsibility. People are doing super-critical CO2 runs with cubic EOS and NIST.

So now what is your issue: perhaps it has some constraints when dealing with a mixture of two components. What is the pressure range / temperature range you are look into? I guess it is getting some problems with the Air component?

[F123rooq]

well, supercritcal co2 pressure is around 74 bar and upwards, and temperature is 33 C and upwards. For the inlet pressure I have put 80 bar, and velocity as 10m/s. The outlet conditions are atmospheric pressure and velocity is zero, because I want the flow into a big tank coming from a tiny pipe, so my outlet is the other end of the tank, which I don't care what it leaves as, so I put zero.

In terms of the error is says that there is no interaction parameter for the CO2 and air.

 

Thank You

Regards,

[DrAmine]

Try with three components: Co2, Oxygen and Nitrogen.

 

[F123rooq]

I entered the three components, and the error went. However, in my cell zone conditions, air was there, so I selected the air for the tank, and co2 for the pipes. Then I had the problem of floating point exception. Which I searched up and it said that it was dividing by zero somewhere. I tried to play about with the mesh, smaller, bigger, triangles, quads, both etc, and in the end I think the error was still there, but it allowed me to proceed with the run, so I initialised and run the calculations. 

In the tank that the co2 is coming into, to mix with the air, so I have the tank domain as air, or should it be a mixture in the cell zone conditions.

 

Thank You  

[F123rooq]

In the residuals in the iterations, it does not have air, it just has continuity, x velocity, y velocity, energy, k, epsilon, co2, oxygen. Is that correct? It should have air right?

 

Thank You

[F123rooq]

I have got an error message:

Error: WorkBench Error: Could not register files of type:Real-gas-mixture

Error Object: #f

Registering Real-gas-mixture, ("hmx.bnc")

 

Error: WorkBench Error: Could not register files of type:registeredTypes

Error Object: #f

Writing data to (my folder location)

x-coord

y-coord

pressure

x-velocity

y-velocity

temperature

k

epsilon

Done.

 

 

Calculation complete.

 

Also, my calculations did not converge in 300 iterations.

[DrAmine]

Not sure how to respond...

Air residual is not plotted because air is ballast component so it us not solved via transport equation.

Regarding workbench error: I would for testing always use standalone Fluent outside the workbench.

Now regarding the materials: what are the ranges and how does the model look like. I do not expect convergence so quick when real gas is used.

[F123rooq]

Also, I used hybrid initialisation, that shouldn't matter too much should it. 

[F123rooq]

That is the model, my inlet velocity is 10m/s. Inlet pressure is 80bar. Outlet pressure 101325 pa.

[F123rooq]

Element size 0.0001m. All triangles method

[DrAmine]

If steady state should not matter how to start: it is only an initial guess.

[F123rooq]

Yeah its steady not transient time

[F123rooq]

Whats the minimum iterations would you say for real gas

[F123rooq]

I tried to load the pathlines, and the model took a long time so I shut it, now started it again, I have to run the simulations again, but there were no movements in the pipe and the the exit of the pipes, was all blue, and near the outlet of the tank, there was read, yellow, green. Sorry couldn't give the numbers, just the colours, but blue low, red high. I will give the numbers once I run again

[DrAmine]

Number of iterations case dependent better to create monitors and scruitinze the trend.

[DrAmine]

Regarding the pathlines: I don't understand.

[F123rooq]

in the solution I tried to get contour of pathlines of the domain, and it crashed,

As far as supercritical co2 simulations is concerned, is this all you have to do to simulate scCO2.

 

Thank You

[F123rooq]

I have floating point exception error

[F123rooq]

But I changed it to transient time and initialised from all zones. and it allowed me.

[DrAmine]

Stick to polite, proper and detailed discussion.

[F123rooq]

Alright, apologies if my words were offensive. I did not mean to be offensive.

 

Regarding the detailed discussion, I have only been doing ANSYS CFD for like 1 month or maybe a little longer, EVERYDAY. But, I am still extremely new, so for me to hold detailed discussions would be a miracle. Every time you have said something complex, I have gone to research on it and that has lead me to further things. I am learning with every of your message. I have come across Homogeneous Equilibrium Method. I don't think its what I am doing, but was nice to read.

But apologies for using impolite words.

 

Kind Regards,

 

[F123rooq]

Hi Amine,

I ran the simulations for 500 iterations, it converged in 348. And this is the result of the contour of velocity. I am super confused as to why this has come to basically zero velocity. Can you please help.

 

Thank you for your help.

 

Kind Regards,

[DrAmine]

As beginner you need to do tutorials. This is mandatory as you want to deal with something complex like real gas.

The velocity contour is dominated by high velocity in the inflow pipes that us why us dominated by blue in almost all region. Probably you need to run longer.

[F123rooq]

Hello,

I ran it for 1500 iterations, and this is what I got for velocity contour. I did get quite high reverse flow. 485 faces, 74.7% area. That is way to high, is it the reason why it's all zero velocity. At the inlet when I zoomed in, its really high velocity. Any suggestions?

 

Kind Regards,

[DrAmine]

I recommend to start first with ideal gas EOS and check boundary conditions. Also when showing contour plots clip the velocity values to see the plenum region.

[F123rooq]

Hi Amine,

Yeah I am going to scrap this supercritical co2. I am going to do liquid co2. Can I ask one question please. You know in ansys, there is CO2 and there is CO2-, the one with the minus is liquid isn't it. I want to do just a very basic simulation of liquid co2 and air in the tank. Is that something that can be easily done, compared to scCO2.

 

Thank you for your help so far.

 

Kind Regards,

[DrAmine]

All want to do is feasible with Fluent. Just start with some basics add complexities step by step so that you have success stories.

Generally default materials names will tell if its liquid. You can define your materials with your input.

[F123rooq]

Yep that's correct. I need to start somewhere basic and almost break my problem down into small bits, and try to solve those small bits and then put it together, and yeah I do need to have some success stories, because I am getting extremely frustrated with the software. Maybe if I understand the problems and errors, I will be less frustrated.

Thank you for your help with everything, I have definitely learnt how to set up supercritical fluids, be it, the solution came out with errors, but at least I know where to start now.

 

Kind Regards,

[DrAmine]

Welcome!