DNS and approximate deconvolution as a tool to analyse one-dimensional filtered flame sub-grid scale modelling

Abstract

A procedure using approximate deconvolution and explicit filtering is discussed to evaluate topology-based sub-grid scale (SGS) combustion models. A direct numerical simulation (DNS) database is first filtered, then a deconvolution operator constructed from the topology-based SGS model is applied, to compare the approximate three-dimensional fields against the exact ones. The DNS is obtained from an already well-resolved large eddy simulation (LES) of a bunsen flame, by refining the mesh up to full resolution of the reaction zones and the turbulent flow scales. The SGS model evaluation via approximate deconvolution is applied to a flamelet-like closure based on the tabulation of filtered one-dimensional flames. The various sources of errors are analysed in a statistical manner in terms of flame topology. Aside from the a priori analysis, results from LES are also reported with the one-dimensional flame deconvolution and compared against those resulting from an approximate three-dimensional deconvolution, confirming the need for accounting for the full 3D flame dynamics in SGS modelling. All the study is performed with tabulated detailed chemistry.