Parallel methodology to capture cyclic variability in motored engines

Abstract

Numerical prediction of cycle-to-cycle variability in spark ignition engines is extremely challenging for two key reasons: (1) high-fidelity methods such as large eddy simulation are required to accurately capture the in-cylinder turbulent flow field and (2) cycle-to-cycle variability is experienced over long time scales, and hence, the simulations need to be performed for hundreds of consecutive cycles. In this study, a methodology is proposed to dissociate this long time-scale problem into several shorter time-scale problems, which can considerably reduce the computational time without sacrificing the fidelity of the simulations. The strategy is to perform multiple parallel simulations, each of which encompasses two to three cycles, by effectively perturbing the simulation parameters such as the initial and boundary conditions. The proposed methodology is validated for the prediction of cycle-to-cycle variability due to gas exchange in a motored transparent combustion chamber engine by comparing with particle image velocimetry measurements. It is shown that by perturbing the initial velocity field effectively based on the intensity of the in-cylinder turbulence, the mean and variance of the in-cylinder flow field are captured reasonably well. Adding perturbations in the initial pressure field and the boundary pressure improves the predictions. It is shown that this new approach is able to give accurate predictions of the flow field statistics in considerably less time than that required for the conventional approach of simulating consecutive engine cycles.