WP7b: Industrial Guidelines
Group lead: Professor Mark Savill
A key UKTC theme is promotion of the cascade of information, understanding and guidance from fundamental DNS and LES research to the CFD modelling needed by industry. Such work has been aided by the European Research Community on Flow Turbulence & Combustion CFD best practice initiatives, and in particular the establishment of the QNET Knowledge Base data repository hosted by the Centre for Flow Modelling & Simulation ( http://cfms.org.uk ), which has led to the identification of a series of key industrial Application Challenge flows, each defined by a set of Underlying Flow Regimes that need to be accurately captured. Building partly on earlier UKTC work, Professors J.C.R. Hunt & M. Savill have established a simple four-parameter description, for mapping turbulent and transitional flows to the correct level of CFD modelling, that can provide a theoretical framework for application of the Knowledge Base. More recent UKTC activities have successfully quantified initial empirical estimates for two of the key parameters (the degree of non- localness and non-equilibrium) for a first set of simple shear flows, and so placed the original ‘Hunt-Savill’ Map on a firmer footing. An aim now is to use simulation data from all the current UKTC Work Packages to quantify representative time, length-scale, straining, and Gaussianity parameters for a much larger set of Underlying Flow Regimes. This work can thus contribute directly to the reinforcement of the widely-used ERCOFTAC Best Practice Guidelines for CFD modelling via the QNET CFD wiki ( http://qnet-ercoftac.cfms.org.uk ). In addition to the focus on RANS modelling guidance from DNS and LES across UKTC, this Work Package is also addressing how improved LES Sub-Grid Scale Models can be extracted from DNS, and how Simulations can be directly coupled to RANS within embedded-DNS and hybrid RANS-LES schemes using newly developed filtering and averaging techniques to allow industry to deploy simulation methods efficiently and effectively where they have been proved necessary to address flow physics sufficiently for accurate solutions.