The accurate modeling of the intricate droplet spectrum arising from non-equilibrium condensation is essential for predicting droplet behavior, estimating exergy loss, and assessing erosion rates. This chapter developed and validated a polydisperse model utilizing the Quadrature Method of Moments (QMOM). The model incorporated a transition SST model, moments characterization, and entropy generation analysis. Employing a spline-based algorithm, the probability density function (PDF) of droplet radius was reconstructed, highlighting superior predictive capabilities of the polydisperse model compared to monodispersed model, particularly in estimating Sauter radius. Investigations into nozzle dynamics under the influence of asymmetric lambda shock and droplets evaporation revealed distinct distributions of moments and droplet spectra, being similar to the gamma distribution in nucleation zones and log–normal distribution in growth zones during supersonic outflow. For the supersonic flow in turbine, where oblique shocks induce complex evaporation and secondary condensation, the droplet spectrum shape closely approximated a gamma distribution. Notably, maximum exergy destruction reached 25.293 kJ/kg, with exergy destruction rates ranging from 1.04 to 4.45% and Baumann factors between 0.574 and 1.312. Additionally, erosion rates in the polydisperse model were found to be 58.4–64.3% lower than those in the monodispersed model. This study underscores the polydisperse model enhanced accuracy in predicting droplet spectra and energy losses in supersonic flows.