[Boyuan]

I'm running a steady state simulation of a supercritical CO2 centrifugal compressor with CFX. During the iteration, the pressure goes out of the upper bound and persists. I tried increasing the pressure range but didn't work. I doubt something wrong with the RGP table.
1. The temperature and pressure range covers the zones of gas, liquid and supercritical. There may be two approaches to generate table. One is treating the gas and liquid as one component and making only one set of data. Then creating 'pure material' material. This is the approach I take. The other is treating the gas and liquid as two components and making two sets of data. Then creating 'Homogeneous Binary Mixture' material. Which is viable?
2. For the second approach, taking the gas table as example, since the temperature and pressure range covers a part of liquid zone, should I fill in the liquid parameter at the corresponding position? If not, the parameters of which temperature and pressure should I use. The same goes for liquid table.
Looking forward anyone's help.

[ghorrocks]

You are proposing a simulation covering the liquid, gaseous and supercritical sections of the material properties - this will always be a very difficult simulation, both physically (ie: what are the physics at work here?) and numerically (ie: how do you make it numerically stable and get a converged solution?).

Neither of your options in (1) mention non-equilibrium conditions from the phase change. Are there any significant non-equilibrium effects? If there are then this already difficult simulation will become much, much harder.

Assuming no significant non-equilibrium effects are present then then the choice between your two options in (1) will depend on exactly what you are modelling and the range of properties it covers.

But the golden rule before starting any difficult simulation like this is: start simple and obtain a converged solution to a simplified case and add the complexity one step at a time. If you turn everything on at once you have no hope of getting this working. So start with a super-simple case - maybe an ideal gas - and make sure that works before adding the complex physics one bit at a time.

[Martin_Sz]

The choice between treating CO2 as a single component or a homogeneous binary mixture in your RGP table for a supercritical CO2 centrifugal compressor simulation depends on several factors:
* Accuracy Requirements: If high accuracy is critical, especially near the critical point, using a homogeneous binary mixture approach might be preferable. This accounts for the distinct liquid and gas phases and their property variations.
* Computational Efficiency: The single component approach is simpler to implement and computationally less expensive. If the pressure excursions are moderate and accuracy near the critical point is less critical, this approach can be sufficient.
Here's a breakdown of both approaches:
Single Component Approach:
* Treat CO2 as a single fluid with properties encompassing the entire range (gas, liquid, supercritical).
* Create a "pure material" in CFX.
Homogeneous Binary Mixture Approach:
* Treat CO2 as a mixture of liquid and gas phases.
* Generate two separate RGP tables, one for each phase, covering their respective temperature and pressure ranges.
* Create a "homogeneous binary mixture" material in CFX.
Liquid Property Handling in Binary Mixture Approach:
* For the gas table, if the pressure-temperature range extends into the liquid region, it's not recommended to fill those entries with liquid properties. Instead, extrapolate gas properties or limit the table range to exclude the liquid zone.
* Similarly, handle the liquid table by either extrapolating gas properties into the gas zone or limiting the table range.
Recommendation:
* Start with the single component approach for simplicity.
* If the pressure excursions persist or you require higher accuracy near the critical point, consider the binary mixture approach.
For detailed guidance on RGP table generation and CFX material creation, consult the CFX documentation or seek help from experienced CFX users on forums like CFD Online.