Investigation of piston bowl geometry and speed effects in a motored HSDI diesel engine using a CFD against a quasi-dimensional model

The present work investigates the effect of varying the combustion chamber geometry and engine rotational speed on the gas flow and temperature field, using a new quasi-dimensional engine simulation model in conjunction with an in-house developed computational fluid dynamics (CFD) code served to validate the predicted in-cylinder flow field and gas temperature distribution calculated by the quasi-dimensional model, for three alternative piston bowl geometries and three rotational speeds. This CFD code can simulate three-dimensional curvilinear domains using the finite volume method in a collocated grid; it solves the generalized transport equation for the conservation of mass, momentum and energy, and incorporates the standard k-ε turbulence model with some slight modifications to introduce the compressibility of a fluid in generalized coordinates. On the other hand, the quasi-dimensional model solves the general transport equation for the conservation of mass and energy by a finite volume method throughout the entire in-cylinder volume, while for the estimation of the flow field a new simplified three dimensional air motion model is used. To compare these two models the in-cylinder spatial and temporal temperature distribution, the mean cylinder pressure diagram, as well as the mean in-cylinder radial and axial velocity are examined, for the three piston bowl geometries and the three speeds, for a high speed direct injection (HSDI) diesel engine operating under motoring conditions. From the comparison of calculated results, it becomes apparent that the two models predict similar in-cylinder temperature distributions and mean air velocity fields at each crank angle, for all cases examined. Thus, it is shown that the quasi-dimensional model with the proposed simplified air motion model is capable of capturing the physical effect of combustion chamber geometry and speed on the in-cylinder velocity and temperature field, while needing significantly lower computing time compared to the more detailed and accurate CFD model. On the other hand, the CFD model is more suitable when detailed simulation of the in-cylinder geometry is required and the way the corresponding transport phenomena are affected.