Computations were carried out to simulate in-cylinder flow field and mixture preparation of a small port scavenged direct-injection two-stroke spark-ignition engine using a modified version of KIVA-3 code. Simulations of the interaction between air flow and fuel were performed on a commercial Piaggio (125 cc) motorcycle engine modified to operate with a hollow-cone injector located in different positions of the dome-shaped combustion chamber. The engine has a large exhaust port and five smaller transfer ports connecting the cylinder to the crankcase. The numerical grid of this complex geometry was obtained using an IBM grid generator based on the output of engine design by CATIA solution. To take into account the rapid distortion of flow, the standard k-ε turbulence model in KIVA-3 was replaced by the RNG k-ε model. Three cases of hollow-cone injector locations were explored: one, located on the head, injected the fuel along the cylinder axis; another one, also located on the head, injected the fuel in counter flow; the third one, located on cylinder wall, injected the fuel towards the combustion chamber. It was found that the most important parameters that strongly influence in-cylinder droplet vaporization process and spatial vapor distribution are: fluid flow pattern, injector location, injection pressure and injection timing.
Modelling the Mixture Formation in a Small Direct-Injected Two-Stroke Spark-Ignition Engine
GENTILI, ROBERTO;
1997-01-01
Abstract
Computations were carried out to simulate in-cylinder flow field and mixture preparation of a small port scavenged direct-injection two-stroke spark-ignition engine using a modified version of KIVA-3 code. Simulations of the interaction between air flow and fuel were performed on a commercial Piaggio (125 cc) motorcycle engine modified to operate with a hollow-cone injector located in different positions of the dome-shaped combustion chamber. The engine has a large exhaust port and five smaller transfer ports connecting the cylinder to the crankcase. The numerical grid of this complex geometry was obtained using an IBM grid generator based on the output of engine design by CATIA solution. To take into account the rapid distortion of flow, the standard k-ε turbulence model in KIVA-3 was replaced by the RNG k-ε model. Three cases of hollow-cone injector locations were explored: one, located on the head, injected the fuel along the cylinder axis; another one, also located on the head, injected the fuel in counter flow; the third one, located on cylinder wall, injected the fuel towards the combustion chamber. It was found that the most important parameters that strongly influence in-cylinder droplet vaporization process and spatial vapor distribution are: fluid flow pattern, injector location, injection pressure and injection timing.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
