A comparative study of intake and exhaust port modeling strategies for scale-resolving engine simulations

Due to their capability to capture cycle-to-cycle variations and sporadically occurring phenomena such as misfire and knock, scale-resolving simulations are becoming more and more important for internal combustion engine simulations. Compared to the frequently used unsteady Reynolds-averaged Navier-Stokes approaches, scale-resolving simulations require significantly greater computational costs due to their high spatial and temporal resolution as well as the need to compute several cycles to obtain sufficient statistics. It is well established that the appropriate treatment of boundary conditions is crucial in scale-resolving simulations and both temporally and spatially resolved fluctuations must be prescribed. However, different port modeling strategies can be found in the literature, especially with respect to the extent of the computational domain (boundary close to the flange vs. the entire system up to the plenum) and the numerical treatment of the intake/exhaust when the valves are closed (enabled vs. disabled). This study compares three different port modeling strategies, namely a long ports version, a short ports version and a version with short and temporarily disabled ports based on the well-established Darmstadt benchmark engine. The aim is to identify the requirements for scale-resolving simulations in terms of the treatment of the intake and the exhaust ports to obtain accurate statistics (mean and variance) and cycle-to-cycle variations of the in-cylinder flow field.