EFFICIENT ROBUST OPTIMIZATON TECHNIQUES FOR UNCERTAIN DENSE GAS FLOWS

Dense Gases (DG) are single-phase vapors of molecularly complex uids operating close to saturation conditions which can be used to improve the eciency of Organic Rankine Cycles (ORC) turbines. This work is devoted to the design of ecient procedures for optimizing a 2D dense gas turbine with multiple sources of uncertainty (thermodynamic model, operating conditions, geometry). Uncertainty Quantication (UQ) stochastic tools, dense gas CFD solvers and a multi-objective optimizer are coupled to produce a set of robust optimal shapes using mean eciency and its standard deviation as objectives.