Structural Uncertainty Estimation of Turbulence Models in Organic Rankine Cycle Applications

orc2019 Tracking Number 131

The investigation of the complex non-ideal fluid flows of interest for Organic Rankine Cycle (ORC) turbomachines largely relies on numerical tools. In Computational Fluid Dynamics (CFD) simulations, a popular strategy to predict the flow behavior is to rely on the Reynolds- Averaged Navier-Stokes equations (RANS). In RANS computations, turbulence models must be employed, to reconstruct the Reynolds stress term arising from the time-averaged decomposition of the Navier-Stokes equations. Generally, the accuracy of RANS simulations is questionable for flow configurations involving adverse pressure gradients, inhomogeneous flow directions, flow separation or strong stream lines curvature (as in the case of turbomachinery applications). For such flows, the inherent model-form assumptions in the RANS approach introduce potential accuracy limitations which affect the credibility of CFD predictions [1]. Moreover, turbulence models often consist in empirical or semi-empirical closures that depend on a set of coefficients optimized for a specific fluid flow. Tough literature is teemed with works reporting on the estimation of turbulence coefficients for flows of fluids of common interest (air, water and many other), little if nothing can be found regarding complex non-ideal fluid flows of interest for ORC applications. Indeed, the scarce amount of experimental data regarding non-ideal flows prevents the empirical estimation of turbulence coefficients. Moreover, due to the complexity of the task, the direct quantification of the errors introduced by RANS closure models is intractable in general. Recently, formal uncertainty quantification techniques have been developed to provide a probabilistic characterization of the corresponding confidence levels. Here, we apply the Eigenspace Perturbation Method (EPM) [2] to a set of exemplary flow configurations of interest for ORC applications. Namely, a non-ideal flow expanding through a converging-diverging nozzle, a non-ideal supersonic stream over a backward facing step and the flow around a typical ORC turbine stator blade. Numerical results show that a systematic and comprehensive treatment of the RANS inherent uncertainties is fundamental for the further improvement and optimization of ORC power production systems. [1] Karthik Duraisamy, Gianluca Iaccarino and Heng Xiao, Turbulence Modeling in the Age of Data, Annual Review of Fluid Mechanics, Vol 51:357:377, 2018 [2] Michael Emory, Johan Larsson and Gianluca Iaccarino, Modeling of Structural Uncertainties in Reynolds- Averaged Navier-Stokes closures, Physics of Fluids 25 (110822) 2013