The Internal Combustion Engines (ICE) are inherently sources of the flow’s unsteadiness in the intake and exhaust ducts. Unsteady flow has a direct impact on the engine’s behavior and performance by influencing the filling and emptying of the cylinder. Air intake boxes as well as muffler geometries, which are very commonly used on the two-wheeled vehicles, have an impact on pressure levels and so, on air filling and performances levels. Thus, the purpose of this paper is to identify and analyze different typical geometries of these elements (air box and muffler) by comparing the test bench results with those obtained by 3D and 1D calculations. In this way, it is possible to establish a methodology for modeling the air box and muffler based on experimental tests and the development of 3D and then 1D model.
In a beginning, studies consist in describing the geometry of the air box and muffler using a combination of tubes and simple volumes. During one-dimensional simulations, the gases properties in a volume must be calculated taking into account a method of filling and emptying. Under transient conditions, the pipe element is considered essentially as one-dimensional. The gas dynamic is described by a system of equations: the equations of continuity, momentum and energy. In the three-dimensional case, all tubes and volumes are meshed and solved using various physical models, equations and hypotheses that will be detailed subsequently.
The study is performed on a shock tube bench. One of the main points is that this type of experimental test allows to test easily different pressure ratios, different geometries and to measure direct and inverse flow. In this way, the propagation of a shock wave is studied in our different geometries and is compared to the pressure signals obtained with 1D and 3D simulations. Once the 1D modeling is obtained, it must be validated in order to be applied in a simulation for Internal Combustion Engine. Validation will be done by direct comparison of results at each stage to ensure that the models and assumptions used in the calculations are correct.