Falfari, S., Forte, C., Brusiani, F., Bianchi, G. et al., "Development of a 0D Model Starting from Different RANS CFD Tumble Flow Fields in Order to Predict the Turbulence Evolution at Ignition Timing," SAE Technical Paper 2014-32-0048, 2014, doi:10.4271/2014-32-0048.
Faster combustion and lower cycle-to-cycle variability are mandatory tasks for naturally aspirated engines to reduce emission levels and to increase engine efficiency. The promotion of a stable and coherent tumble structure is considered as one of the best way to promote the in-cylinder turbulence and therefore the combustion velocity. During the compression stroke the tumble vortex is deformed, accelerated and its breakdown in smaller eddies leads to the turbulence enhancement process.The prediction of the final level of turbulence for a particular engine operating point is crucial during the engine design process because it represents a practical comparative means for different engine solutions. The tumble ratio parameter value represents a first step toward the evaluation of the turbulence level at ignition time, but it has an intrinsic limit. The tumble ratio parameter represents the value of the angular velocity of a single macro vortex, while the flow-field is often characterized by multiple vortexes, sometimes some rotating and some counter-rotating.The idea at the basis of the paper is: To develop a quasi-predictive 0D model for defining the final mean level of the turbulence at the ignition time. The model is fed by the curtain intake valve mass flow rate and the intake valve lift trend. In order to validate the 0D model the results were compared versus 3D CFD results.To extract from a 3D CFD flow field at IVC the type and the number of the vortexes. The 3D CFD RANS simulations were performed by AVL Fire code v. 2010.To demonstrate through some 3D CFD results that the flow field structure was a function only of the engine type and the load condition.Finally the flow field structure could be used in the 0D model on varying the engine speeds as a means of improvement of the model prediction capability.