[R4z0rX]

Hey guys,

I am having a problem modeling a phase change from liquid methane to supercritical methane vapor.
The liquid methane enters a channel at 100K and given mass flow rate as inlet conditions and gets heated up by the heated walls of the channel thus changing to supercritical vapor with 50 bar static pressure at the outlet.

I set up a homogenous binary mixture using an RGP-file containing the tables for the liquid and vapor phases and saturation tables.
The vapor phase is set to equilibrium constraint and the liquid phase to equlibrium fraction in the domain definition. I set the equilibrium fraction to 1 at the inlet and the domain initialization. Morphology set to continuous fluid.

After a couple iterations the solver crashes due to a fatal overflow in the linear solver and until that, I am getting a lot of clipping and extrapolation warnings for exceeding the upper bound of the RGP-tables for the liquid phase.

I think this is caused by problems of the phase changing from liquid state to vapor phase.

Do you have any suggestions how to model this phase change properly?

Thank you and best regards!

[evcelica]

You do not have "liquid" methane in your case, and no phase change occurs, as you never cross the saturation curve. So material properties will have a smooth transition, defined completely by the temperature and pressure.
Luckily, you are above the critical pressure (46 bar) in your entire simulation, therefore you have single phase supercritical fluid (gas) the entire time, so model it as such.

Supercritical fluid is when the Temperature OR Pressure is above the critical values.
I had a similar case once, but unfortunately I was below the critical pressure to start (Liquid Helium in closed piping system). Adding heat energy would cause the pressure to rise and the liquid to transition from "liquid" (one RGP file) to "supercritical fluid" (a different RGP file) without ever crossing the saturation curve. So I had a terminology change, and table change, without ever having an actual phase change, and I could not figure out how to model that. Pressure would just increase until it reached the critical pressure and reach the end of the liquids rgp table, but never transition to gas as it didn't cross the saturation curve. Luckily it was for a transient relief case, where I could just change my initial conditions by adding heat energy until my pressure was above the critical pressure, and start my simulation with all supercritical fluid. This made it a much simpler single phase case, just like your case.

If anyone knows how my case would possibly be performed let me know.
A custom .rgp file with the liquid and gas tables combined to represent both fluids and the smooth transition is all I can think of.

[R4z0rX]

Thank you for your help but as far as I'm concerned I do have liquid methane at the beginning. I'm expecting the methane to have a pressure of about 95 bar at the inlet and a temperature of 100K. At the outlet, I am expecting the methane to have a pressure of about 55 bar and a temperature of 615K.

Wikipedia states: "A supercritical fluid (SCF[1]) is any substance at a temperature and pressure above its critical point [...]"

So I found a phase diagram for methane and extrapolated it so it shows the temperature range I need. The blue line represents the line between the two states I am expecting. It is above the critical point so you're right, I don't have a real phase change but a smooth transition as you said is occuring.

Because I am starting in the liquid region I thought it would be best practise to model it as homogenous binary mixture.

Maybe I am missing something.

[evcelica]

What Wikipedia states about the definition of supercritical fluid is not representative of how CFX organizes fluid property tables. Understand "supercritical fluid" is simply terminology, and this is different between CFX and the Wikipedia definition. You need to use the property table that includes data at the (T,P) range you are simulating. Look through your rgp file(s) for data at your inlet conditions, tell me which phase it is, liquid or Vap? Obviously not liquid, the program even told you that itself with all the clipping/out-of-range messages for your liquid phase.

I suggest you read through the CFX documentation. Research "Table Interpolation and Saturation Curve Clipping" in the CFX-Solver_Modeling_Guide: Section 12.2, specifically see Figure 12.5. In the future, please listen when a CFX expert tries to help you, and research/test their advice, especially when they have experience with your exact case! You should explore the avenues they give you before stating they are wrong, because it isn't their job to have to convince you properly model your case.

[R4z0rX]

I am sorry if you thought I wanted to lecture you, I was confused by the different terminology and was trying to have an objective discussion about it. I did not say you were wrong, I just stated the definition of a supercritical fluid as I know it and if you thought that was the case, I am sorry, that wasn't my intent. Besides that, it's hard to tell on an online forum who is an expert and who is not.

Anyways, you proved me wrong and I really appreciate your help! I will try to model it as you suggested. Thank you very much!

[blacksatellite]

Originally Posted by evcelica

What Wikipedia states about the definition of supercritical fluid is not representative of how CFX organizes fluid property tables. Understand "supercritical fluid" is simply terminology, and this is different between CFX and the Wikipedia definition. You need to use the property table that includes data at the (T,P) range you are simulating. Look through your rgp file(s) for data at your inlet conditions, tell me which phase it is, liquid or Vap? Obviously not liquid, the program even told you that itself with all the clipping/out-of-range messages for your liquid phase.

I suggest you read through the CFX documentation. Research "Table Interpolation and Saturation Curve Clipping" in the CFX-Solver_Modeling_Guide: Section 12.2, specifically see Figure 12.5. In the future, please listen when a CFX expert tries to help you, and research/test their advice, especially when they have experience with your exact case! You should explore the avenues they give you before stating they are wrong, because it isn't their job to have to convince you properly model your case.

Hi Sir,

İ try to modelling in CFX the same problem, but i couldn't solve,
İnlet temp 110K
Pressure outlet 50 bar