Performance of High Order Method on Simulation of In-Cylinder Tumble Flow

Accurate and efficient simulation of unsteady turbulent in-cylinder flow process of the piston engine is essential in better understanding the physical process and better designing the cylinder scavenging system. The engineering RANS method is dominated by second order. This paper studies the performance of the high order method in the numerical simulation of the cylinder scavenging under the RANS framework. The current study is with a 2-D setup to simulate tumble flow during the intake process of an SI engine. We adopt the monotonicity preserving(MP) scheme from fifth order (MP5) to up to ninth order (MP9) with SST turbulence model equipped. The simulation results are compared between conventional second order and high order ones with a same resolution of mesh grid. It is shown that the vorticity resolving ability of RANS method can be better improved by using the high order scheme. Different reconstruction variables such as primary variable, conservative variable and characteristic variable can also influence the capturing of these details to a certain extent. For flow processes that do not have strong discontinuities, the use of primary variable reconstruction provides the best simulation precision. In terms of computational efficiency, the high order scheme is comparable to the second order scheme. Therefore, the high order scheme is very applicable in RANS method for the simulation and design purpose of the cylinder.