Study of thermal boundary layer in pulsatile flows

Fully integrated reactive simulations in internal combustion (IC) engines have become a critical target for the automotive industry, where CFD has an increasing impact in the decision process for the design of new prototypes. Although a substantial level of maturity has been reached for simulations, there are some fields where the potential of CFD can still be leveraged to improve engine energetic and environmental efficiencies. One of these fields concerns thermal losses due to heat transfer at the wall between a hot combustion chamber and the cool surroundings. Given thermodynamic conditions, it is extremely difficult, to monitor the heat loss experimentally. CFD represents a possible way to get a reliable prediction of heat loss. Besides, at the same bulk and surrounding conditions, the experimental study conducted in Kearney et al. (2001) concluded that the heat loss magnitude could vary by a factor of two from steady to pulsatile bulk flows. A pulsatile flow is characterized by steep variations of the intensity of its velocity, strong enough to possibly inverse its sign. This kind of flow is typically found in cylinders of internal-combustion engines, where successive sequences of compressions and expansions occur. The identification of key parameters that control the heat loss intensity can substantially improve the engine efficiency. However, although thermal boundary layers are well understood in steady state flows, their characterization in transient flows is only at its early phase Costamagna et al. (2003) have identified some quasi-coherent structures in these types of flows. The objective of this study is to set up an efficient numerical framework for the study of the thermal boundary layer in a model cylinder of an internal combustion engine. The remainder of this report is organized as follows. A compressible formulation is first presented, as well as the ALE formalism used to treat the moving geometry due to the compression/expansion phase. Since the Mach number is typically small in these flows, Section 3 presents a first attempt to build a low-Mach solver accounting for variations in density due to temperature variation. Section 4 presents verification results for dilatational flows. Finally, a summary of the accomplishments and associated perspectives is given in the concluding section.