Basic numerical assessments to perform a quasi-complete LES toward ic-engine applications

In industrial field, the numerical representation of a fully 3D unsteady turbulent flow behaviour is one of the most common situation that has to be faced. As well known, to answer this need three simulation approaches can be used: Reynolds Averaged Navier Stokes (RANS), Large Eddy Simulation (LES), and Direct Numerical Simulation (DNS). Surely, the RANS method is, today, the most used tool to perform turbulent flow simulation because it allows a good numerical reproduction of the mean flow conditions at an acceptable computational cost. More accurate than RANS is the LES simulation. LES approach allows the direct solution of the largest turbulent scales (anisotropy turbulence) while the smallest scales (isotropy turbulence) are numerically modelled by a sub-grid scale model. LES simulation could be considered the right simulation approach when the physical behaviour of the considered domain is dominated by the large scales of motion. DNS is surely the most complete approach because, in this case, all the turbulent scales are directly solved. However, today the DNS approach remains inapplicable in industrial field because of the prohibitive computational power required. During the last few years, significant advances have been performed in Large Eddy Simulation (LES) applications. Improvements about accuracy of numerical methods, sub-grid scale models, and computational performance support the application of LES on industrial field with a positive spin-off on the fluid dynamic comprehension of complex fluid dynamic systems. Today, one of the main questions that involves LES application is about the evaluation of the LES simulation quality. To clearly understand which kind of LES simulation is performing, it is necessary to calculate the resolved energy level and to use this information to define quality parameters. This paper presents different methods that could be applied to evaluate the LES simulation quality. For each method the main correlated advantages and disadvantages were evaluated. Simulations were made on a backward facing step test case to show the application of the defined LES quality parameters, and on a real IC engine geometry to show the quality parameter application on a real engine industrial case