Create the stator and rotor components as follows:

In this section, you will set properties of the fluid domain and some solver parameters. Note that initially you will choose the fluid to be Water Ideal Gas, but later you will create a new fluid based on the IAPWS database for water and override this initial setting with it.

CFX-Pre will try to create appropriate interfaces using the region names viewed previously in the Region Information section (in the Component Definition setup screen). In this case, you should see that a periodic interface has been generated for both the rotor and the stator. The generated periodic interface can be edited or deleted. Interfaces are required when modeling a small section of the true geometry. An interface is also needed to connect the two components together across the frame change.

CFX-Pre will try to create appropriate boundary conditions using the region names presented previously in the Region Information section. In this case, you should see a list of generated boundary conditions. They can be edited or deleted in the same way as the interface connections that were set up earlier.

Earlier in the physics definition portion of the Turbomachinery wizard, you specified Water Ideal Gas as the fluid in the domain. Here, you will specify a homogeneous binary mixture to replace it. To create the mixture, you will take two pure fluids from the IAPWS database for water and combine them. The pure fluids that will be combined are H2Og, representing water vapor, and H2Ol, representing liquid water. The mixture will be named H2Olg.

The present simulations use the published IAPWS-IF97 (International Association for the Properties of Water and Steam - Industrial Formulation 1997) water tables for properties. The published IAPWS-IF97 equations have been implemented in Ansys CFX, enabling you to directly select them for use in your simulations. The present example uses these properties in a tabular format requiring you to specify the range of the properties (such as min/max pressure and temperature bounds) and the number of data points in each table. Note that the IAPWS-IF97 properties have been tested for extrapolation into metastable regions, a fact that will be used for the non-equilibrium calculations that require this kind of state information.

You now need to update the initial setting for the domain fluid (initially set while in the Turbomachinery wizard) with the new homogeneous binary mixture fluid that you have just created. This mixture (H2Olg) acts as a container fluid identifying two child materials, H2Ol and H2Og, each representing the liquid and vapor properties in the pure fluid system. The equilibrium solution uses the binary mixture fluid, H2Olg, and assumes that equilibrium conditions relate H2Ol and H2Og at all times. In the non-equilibrium solution (in the second part of this tutorial), H2Ol and H2Og are each used separately to define the fluids that are active in the domain; this is a requirement since the equilibrium constraint is no longer applicable in that case.

Make a 2D surface at 50% span to be used as a locator for plots:

In the equilibrium solution, phase transition occurs the moment saturation conditions are reached in the flow. The amount of moisture ultimately created from phase transition is determined in the following manner.

The static pressure solution field yields the saturation enthalpy through the function , available from the IAPWS database. During the solution, if the predicted mixture static enthalpy, , falls below at any point, the mass fraction of the condensed phase (also called wetness) can be directly calculated also based on the IAPWS properties. The degree to which wetness is generated depends on the amount by which is less than .

Note that the mass fraction is predicted at each iteration in the solution, and is used to update other two-phase mixture properties required at each step in the solution. The final results therefore include all of the two-phase influences, but assuming equilibrium conditions. The equilibrium solution mass fraction contour plot will be compared to the non-equilibrium solution one in Viewing the Results Using CFD-Post.

In the next step, you are asked to create the pressure and mass fraction contour plots.

In the equilibrium solution, the condensed and gas phases share the same temperature and, as a result, predictions of thermodynamic losses are not possible.

In the following step you are asked to view the temperature field in the solution, and note that, due to the equilibrium constraint, it represents conditions for the mixture.

The non-equilibrium case is different from the equilibrium case in that the creation of the second phase (that is, the liquid water) is based on vapor supercooling in conjunction with fluid expansion rate and a nucleation model. The location where the phase transition happens is not specified, but evolves as part of the solution.

In this section, you will specify a nucleation model for the phase that is considered condensable.

You will also set the droplet number and volume fraction of the condensable phase to zero at the inlet. The condensable phase will appear within the domain by homogeneous nucleation. If this case were to involve wetness at the inlet, you would have a choice of specifying either droplet number or droplet diameter as a boundary along with the volume fraction.

Make a 2D surface at 50% span to be used as a locator for plots:

The non-equilibrium solution provides considerable detail on conditions related to phase transition. In particular, the solution tracks the evolution of metastable conditions in the flow through a supercooling variable obtained on the basis of local pressure and gas phase temperature. The supercooling level is the primary variable influencing the nucleation model. The nucleation model provides an estimate of the rate of production of critically sized nuclei that are stable enough to grow in a supercooled vapor flow.

In the next step, you will plot the degree of supercooling, where the supercooling represents the difference between the saturation temperature, set by the local pressure in the flow, and the associated local gas phase temperature. At the inlet, supercooling is often shown as negative, indicating superheated conditions. At equilibrium conditions, no supercooling is allowed since liquid and vapor phases always share the same temperature. Also, critical homogeneous nucleation at pressures below one atmosphere generally involves supercooling levels of up to 35 to 40 K, depending on rate of expansion in the flow.

In the next step, you will plot the nucleation rate to show the level of nucleation attained, along with the droplet number concentration that is present in the flow following nucleation. Note that the nucleation rates reach very high levels, with peak values remaining in the flow for only a short time (that is, as long as supercooled conditions remain). The supercooled droplets released at nucleation grow rapidly, taking mass from the vapor phase and releasing thermal energy, which acts to rapidly reduce vapor supercooling to near zero. This fact removes further significant nucleation and is why the phase transition process is limited to a narrow region in the flow in most cases.

In this section, you will plot the mass fraction (or wetness) of the condensed phase. From the mass fraction and droplet number, it is possible to derive a particle diameter, which you will also plot in this section. Notice that the particle diameters appear in the flow at very small sizes, in the range of 5.0E-9 m, but grow rapidly so that, when leaving the nucleation zone, they are approximately an order of magnitude larger. Since the pressure is dropping through the turbine, droplet sizes continue to increase along with the wetness. To emphasize the particle diameters proceeding out of the nucleation zone, you will set the contour range for viewing. In flow regions near walls, where expansion rate is reduced, small amounts of liquid may have droplet diameters that grow to large sizes relative to those coming from the nucleation zone. Without setting the range, these particle diameters near the walls will be emphasized.

In a non-equilibrium prediction, the gas phase temperature is different from the condensed phase temperature. The former is determined from a transport equation and the latter from an algebraic relationship relating droplet temperature to its diameter (through a small droplet model already described).

In this section, you will plot the different temperature fields. It should be noted that, for the case of the condensed phase temperature field, before droplets are actually formed, there is no meaningful droplet temperature. At the point droplets are formed by nucleation, their temperature is at the gas phase. Once the droplets have grown in size, their temperature is very close to the saturation temperature. Because the non-equilibrium solution considers the condensed phase temperatures separate from the gas phase, the influence of thermodynamic losses are included in the predictions. This is because it becomes possible to account for heat flow between the vapor and condensed phases as they pass through the domain. Due to this, non-equilibrium efficiency predictions are more accurate than ones obtained using an equilibrium model.