Development of a K-k-∊ Phenomenological Model to Predict In-Cylinder Turbulence 2017-01-0542

The turbulent flow field inside the cylinder plays a major role in spark ignition (SI) engines. Multiple phenomena that occur during the high pressure part of the engine cycle, such as early flame kernel development, flame propagation and gas-to-wall heat transfer, are influenced by in-cylinder turbulence. Turbulence inside the cylinder is primarily generated via high shear flows that occur during the intake process, via high velocity injection sprays and by the destruction of macro-scale motions produced by tumbling and/or swirling structures close to top dead center (TDC) . Understanding such complex flow phenomena typically requires detailed 3D-CFD simulations. Such calculations are computationally very expensive and are typically carried out for a limited number of operating conditions. On the other hand, quasi-dimensional simulations, which provide a limited description of the in-cylinder processes, are computationally inexpensive. They require only a small fraction of the computational resources needed for CFD calculations and can be carried out for a large number of cases. Such simulations typically use zero dimensional (0D) phenomenological sub-models to simulate the various in-cylinder phenomena such as heat transfer, combustion, flow variations etc.

The current study presents a newly developed sub-model, available as a part of the software GT-SUITE, which governs the evolution of the mean and turbulent flow inside the cylinder. Within a 0D context, the accurate knowledge of in-cylinder turbulence levels close to ignition is essential to reliably model turbulent flame propagation. However, engines running with stratified charge and/or early spark timings also require the accurate estimation of turbulence levels over the intake and compression strokes. The new model aims to accurately predict in-cylinder turbulence levels over the entire engine cycle. It utilizes a K-k-∊ approach, where K and k are the mean and turbulent kinetic energy and ∊ is the turbulent dissipation rate. The study shows that the model, once calibrated for a particular engine using 3D-CFD results, has the capability to predict the temporal evolution of in-cylinder flow quantities and also responds well to changes in the operating conditions of the engine such as variations in speed, valve lift and timing.