Development and validation of an ICE + ATS co-simulation model for a CNG heavy duty SI engine

Given the need to achieve the “EU Green Deal” targets, with net-zero Global Greenhouse Gas and neutral CO2 emissions in 2050, always more restrictive regulations have increased the interest of the automotive industries in sustainable and efficient Internal Combustion Engines (ICE) design. Especially in Heavy-Duty engines, Natural Gas (NG) became more and more attractive as alternative fuel, reducing adverse health effects of air pollution and, at the same time, ensuring high performance. As a result, the use of predictive numerical models has become necessary to support the experimental campaign, exploring multiple engine operating conditions and offering an important contribution in time-to-market reduction. The thesis work has been carried out thanks to the cooperation between Politecnico di Torino and CNR-STEMS research institute in Naples. Goal of the present activity is the implementation of a co-simulation strategy between a 1D Compressed Natural Gas (CNG) HD SI engine model and a “quasi-steady” Three-Way Catalyst (TWC) model equipped with comprehensive oxygen storage and release submodels, already validated by CNR-STEMS in Steady-State and dynamic conditions in GT-Suite Platform. The coupling of the two models was developed in Matlab/Simulink environment, in order to create a single model able to simulate the whole ICE+TWC system. This task has been accomplished in two steps: in the first stage, the only 1D engine model has been linked to Simulink in order to control and run the engine model simulation directly in this platform both in steady-state and dynamic conditions. Afterwards, the TWC has been added to the system, with a dedicated link appropriately designed to connect the two GT-Suite models, allowing the outputs of the engine model to become the inputs of the TWC one. The strength of the coupling has also been verified for the complete ICE+TWC model not only in stoichiometric λ conditions, but also during a complete SS λ sweep from extreme rich to lean conditions and fast dynamic λ perturbations. As a matter of fact, in CNG engines, the correct reference Air-to-Fuel ratio is a critical aspect to consider due to sudden changes in behaviour around the stoichiometric in terms of performance and TWC conversion efficiency. The calibrated co-simulation model has been validated both in the aforementioned Steady-State and dynamic conditions and during a WHTC homologation cycle, showing a good reliability in the reproduction of the experimental results. Due to a semi-predictive combustion model, calibrated exclusively in stoichiometric conditions and extended in this analysis to the other lambda conditions, greater deviations are present in the extremely rich or lean phases. Nevertheless the proposed co-simulation model logic has shown a wide flexibility and high potential of use that can be spread also to further powertrain applications, representing a good starting point for a necessary virtual vehicle model design.