Discretization Uncertainty Estimation based on a Least Squares version of the Grid Convergence Index

Abstract

This paper presents the application of a method for the estimation of the discretization uncertainty based on a Least Squares version of the Grid Convergence Index. Grid renement studies have been performed for two 2-D steady test cases: a Manufactured Solution that mimics a near-wall incompressible turbulent o w and the o w over a backward facing step. The computations were performed with the nite-dif ference and nite-v olume versions of PARNASSOS using the eddy-viscosity oneequation turbulence model of Spalart & Allmaras model and the baseline k ω two-equation turbulence model proposed by Menter.