[liliana]

Hello.

I am simulating a supersonic flow inside a converging-diverging nozzle with non-equilibrium condensation model. For validation purposes, I am trying to validate the model comparing the results with experimental data from Moses and Stein (1978) and numerical data Yang et al. (2017). The CD nozzle was built using a 2D simplification.

I am using the following models: Coupled Energy; Coupled Flow; DMP; RANS k-e; Steady Solution; Multiphase Interaction. For simulating water vapor, I selected IAPWS-IF97.

In DMP phases, I have created a discrete phase following the configurations: Liquid, Constant Density, Two-way Coupling, Flow Model, Segregated Fluid Temperature and Discrete Quadrature S-Gamma.

In Multiphase Interaction, I have selected the following models: Drag Force (Schiller Naumann), Interaction Length Scale, Multiphase Material and Non-Equilibrium Condensation.

At inlet and outlet, the volume fraction of the discrete phase is zero. That is because I am trying to simulate an homogeneous condensation, where the particles are formed inside the nozzle when the properly conditions are achieved.

At first, the simulation was not converging at all. After a while, I changed the Sauter Mean Diameter in Initial Conditions inside DMP model to a smaller value, and the simulation converged. Also, in Interaction Length Scale I choose the Sauter Mean Diameter. Although I got convergence, the results are not what I expected to be. The nucleation rates are higher than expected and it happens later inside the nozzle.

By studying different materials and reading both physics simulations and theory manuals of Simcenter Star-CCM+, I undestood that the particles are formed when specific conditions of the flow are achieved. When the flow reaches these conditions, the critical radius r* is achieved and the growth process begins.

So, if the particle radius is given by this r* (which is calculated by using flow properties) and after that by the growth model, why do I have to choose a Sauter Mean Diameter? What is the order of the value I should use there? And why it has so much impact in the simulation?

Thanks in advance,

Liliana.