Efficient Uncertainty Quantification of Turbulent Flows through Supersonic ORC Nozzle Blades

This work aims at assessing different Uncertainty Quantification (UQ) methodologies for the stochastic analysis and robust design of Organic Rankine Cycle (ORC) turbines under multiple uncertainties. Precisely, we investigate the capability of several state-of-the art UQ methods to efficiently and accurately compute the average and standard deviation of the aerodynamic performance of supersonic ORC turbine expanders, whose geometry is preliminarily designed by means of a generalized Method Of Characteristics (MOC). Stochastic solutions provided by the adaptive Simplex Stochastic Collocation method, a Kriging-based response surface method, and a second-order accurate Method of Moments are compared to a reference solution obtained by running a full-factorial Probabilistic Collocation Method (PCM). The computational cost required to estimate the average adiabatic efficiency, Mach number and pressure coefficient, as well as their standard deviations, to within a given tolerance level is compared, and conclusions are drawn about the more suitable method for the robust design of ORC turbines