Overset grids for fluid dynamics analysis of internal combustion engines

Abstract

Overset grids is emerging as an increasingly important and powerful tool in fluid dynamics simulations. A model built with the overset mesh technique consists of multiple spatially superimposed grids, one for each moving element (valves and piston in case of internal combustion engines) plus one background grid (representing the fixed portions, i.e. cylinder, ports, manifolds, etc.). Boundary information is exchanged between these grids via interpolation of the flow variables.Several reasons make this technique essential for different purposes. In the field of CFD simulations of internal combustion engines, the ease of replacing moving components makes overset mesh an interesting tool in order to optimize the design of single components without generating new grids for the whole engine assembly.Valves and piston motions are generally characterized by close gaps and self-intersecting trajectories. Overset grids allow to easily control valve and piston motion preserving a high mesh quality in the cylinder, at the valve seats and thereabouts: these areas are proven to be critical for the CFD analysis because of the high spatial gradients in the velocity flow field.The major drawbacks of the overset grid technique applied to internal combustion engines are the mass conservation error and the additional computational load introduced by the interpolation process.The paper shows the application of overset grids for the analysis of in-cylinder flow and turbulence in a research optical engine called TCC-III (Transparent Combustion Chamber). The TCC engine work has been funded by General Motors through the General Motors University of Michigan Automotive Cooperative Research Laboratory, Engine Systems Division.The applicability of overset meshing technique is assessed in terms of computational efficiency (stability, scalability, robustness) and accuracy. In particular, the effects of grid topology, grid density and turbulence modelling are evaluated by means of the comparison between CFD results and high-speed particle image velocimetry (PIV) experimental data. The comparison between RANS (Reynold-averaged Navier-Stokes) CFD simulations and the experimental data show a very high level of accuracy in the representation of the most relevant turbulent structures inside the cylinder.