Influence of various mesh densities and topologies on combustion process characteristics computation in a GDI engine

Avtomobilska industrija se dandanes sooča z vedno večjimi izzivi, razvojni cikel bencinskega motorja pa mora biti zaključen v izredno kratkem časovnem obdobju. Ugoditi okoljskim predpisom je le eden izmed poglavitnih izzivov, s katerimi so soočeni inženirji. Vsem zahtevam trga je v izjemno kratkem času možno ugoditi le izključno z pomočjo uporabe numeričnimi simulacij (RDT). V ta namen se uporabljajo programska orodja, ki omogočajo izdelavo računskih mrež zgorevalne komore, pri čemer je uporabniku ponujena možnost odločitve med dvema različnima topologijama. Izvedena je bila primerjava rezultatov pridobljenih z računsko mrežo izdelano s pomočjo avtomatiziranega postopka znotraj programa Fire FAME Engine Plus ter strukturirano računsko mrežo izdelano s programom Fire ESE Engine. Glede na pridobljene rezultate različnih topologij in gostot mrež ter debeline in števila dodeljenih plasti za popis razmer tik ob steni, so bila podana priporočila za nadaljnjo delo pri izdelavi tovrstnih računskih mrež.Automotive industry is nowadays facing with even bigger challenges, development time of a gasoline engine is getting shorter with every new generation of engine. To accept enviromental regulations is only one of bigger challenges which engineers have to cope with. All requests, which are coming from the automotive market, can be resolved only with use of numerical analysis (CFD). For this purpose we are using pre-processing tools, which allow us to create computational mesh of combustion chamber, where user has an option to choose between two different mesh topologies. One type of topology was created using automated process with a help of Fire FAME Engine Plus and second one was a structured mesh, which was obtained with a program Fire ESE Engine. A analysis of results was made. Based on gathered results of using meshes with different discretization density, topologies and number and thickness of boundary layers, recommendations for the future work regarding creation of computational mesh have been given