A comparative test for assessing the performances of large-eddy simulation codes Authors: F. Denaro 1 , A. Abbà 2 , M. Germano 3 , M. Icardi 4 , D. Marchisio 4 , S. Rolfo 5 , P. Lampitella 6 , E. Colombo 6 , F. Inzoli 6 , ,A. Aprovitola 7 , F.S. Marra 7 , M. Ioveno 3 , D. Tordella 3 . 1 Department of Aerospace and Mechanical Engineering, Second University of Naples, Italy. 2 Department of Mathematics, Politecnico di Milano, Italy 3 Department of Aerospace Engineering, Politecnico di Torino, Italy 4 Department of Materials Science and Chemical Engineering, Politecnico di Torino, Italy 5 School of Mechanical, Aerospace and Civil Engineering, University of Manchester, UK 6 Department of Energy, Politecnico di Milano, Italy 7 Istituto di Ricerche sulla Combustione, CNR Naples, Italy LESinItaly Group E-mail:lesinitaly@caspur.it ; Web: http://cfd.caspur.it/ Keywords: LES, commercial and open-source codes, channel flow. SUMMARY. This paper summarizes the activity of the LESinItaly group in a project supported by the CASPUR centre during the year 2010. Commercial, open-source and own-made LES codes were tested by several units in the simulation of the turbulent flow in a plane channel. The codes produced solutions on two different grids, providing also the corresponding no-model solutions. 1 INTRODUCTION: THE LESINITALY PROJECT The LES is a theoretical-numerical methodology born more than 40-years ago in the environmental flows field and devoted to reduce the degree of freedom in the numerical solution of turbulence at such high Reynolds numbers for which the direct numerical simulation (DNS) appears prohibitive [1-3]. LES is based on a localized filtering operation that can be performed either explicitly or implicitly, allowing for a separation in large and small characteristic scales of flow. Sub-grid time-space scales (SGS) of the turbulent motion are effectively removed from LES solution but, differently from approaches based on statistical averaging (RANS/URANS), large energetic, 3D, time-dependent flow structures are directly resolved, providing details of the flow field that are extremely useful in some practical engineering applications. Indeed, the enormous development of both theories and numerics during the ’90, as well as the increasing computational power, makes LES a candidate to substitute efficiently, at least for problems of small/mid scale, the traditional statistical methodologies. Nevertheless, the quality in LES solutions is greatly influenced by several parameters [4-8], such as type and accuracy of the used numerical method as well as discretization of the computational domain. The type of SGS model and its discretization, the boundary conditions are further elements influencing the quality of the LES solution. A conscious use of the LES method in engineering applications appears crucial as well as the definition of clear guide-lines in the numerical procedure and reliable evaluation criteria. Actually, depending on the scientific background of the researchers, many interpretations of the