Turbulent Jet Ignition (TJI) Combustion Modelling in GT-Suite

The purpose of this thesis work is to develop a methodology that allows to model a Turbulent Jet Ignition Predictive Combustion through the usage of the latest version of the software GT-ISE (v2020). The work is developed in collaboration with Maserati S.p.A. (located in Modena - MO) and Powertech Engineering S.r.l. (Torino - TO). Turbulent Jet Ignition (TJI) is an advanced pre-chamber initiated combustion system for an otherwise standard spark ignition engine found in current on-road vehicles. Turbulent Jet Ignition combustion systems produce multiple and distributed ignition sites, enabling very fast burn rates and allowing for increased levels of dilution when compared to conventional spark ignition combustion. The high speed of combustion can be exploited to get significant benefits in terms of fuel consumption and pollutant emissions, making this kind of technology very promising for the development of internal combustion engines. Due to the importance of these systems, a methodological study aimed to model the Jet Ignition combustion is necessary. A fully predictive combustion model would allow to evaluate the TJI combustion with respect to some engine design parameters variation, like the prechamber geometry or the compression ratio, or operating parameters as the spark timing or VVA strategies. This study was carried out on a turbocharged SI-engine, which was developed by Maserati company, using the GT-SUITE environment. GT-SUITE is the leading engine and vehicle simulation tool used by engine makers and suppliers. It is suitable for analyses of a wide range of issues related to vehicle and engine performance. A methodology to be followed has been built. After an analysis of the experimental data, a single-cylinder simplified model was set up, in order to reduce the computational time. Later, the prechamber system was modeled. The v2020 of GT-ISE showed to be unable to properly model a predictive combustion in the prechamber, while the Jet Ignition calibration turned out to be more robust if conducted with the aim of minimizing a user-defined error calculated on the main chamber pressure. Despite this, the degree of correlation achieved on the full load curve does not meet the expectations, but there is room for improvement expected for the v2021.