On the applications of low-reynolds cubic k-ε turbulence Models in 3D simulations of ICE intake flows

The evaluation of the steady-flow dischargecoefficient of ICE port assemble is known to be verysensitive to the capability of the turbulence sub-modelsin capturing the boundary layer dynamics. Despite thefact that the intrinsically unsteady phenomena related toflow separation claim for LES approach, the presentpaper aims to demonstrate that RANS simulation canprovide reliable design-oriented results by using low-Reynolds cubic k-ε turbulence models.Different engine intake port assemblies andpressure drops have been simulated by using the CFDSTAR-CD code and numerical results have beencompared versus experiments in terms of both globalparameters, i.e. the discharge coefficient, and localparameters, by means of static pressure measurementsalong the intake port just upstream of the valve seat.Computations have been performed by comparing twoturbulence models: Low-Reynolds cubic k-ε and High-Reynolds cubic k-ε.The analysis leaded to remarkable assessmentsin the definition of a correct and reliable methodology forthe evaluation of engine port breathing capabilities.Comparison between numerical results and experimentsshowed that the low-Reynolds cubic k-ε model isunavoidable to correctly capture the influence of portfeature variations on engine permeability. In particular,the deficiencies demonstrated by High-Reynolds cubick-ε turbulence model in resolving the influence of nearwallshear and adverse pressure gradient effect onboundary layer dynamics are completely overcome bythe use of the Low-Reynolds formulation.