Predicting and Minimizing Virtual Vehicle Cold Start Driveline Model with a Real-Time 1-D Gas Engine Code and Chemical Kinetics Aftertreatment 2018-01-1425

The upcoming World-harmonized Light-duty Vehicles Test Cycle (WLTC) together with the Real Driving Emissions (RDE) legislation used for the assessment of fuel economy and emissions, demand a start from a cold engine state. The process of warming up the engine from a cold start has a significant contribution to the emissions and fuel economy of the entire drive cycle. The process involves a multitude of interdependent components which means that modelling the phenomena has so far only been achieved using highly simplified approaches or accepting a very large penalty on calculation time.

This paper presents a modelling of the real-time running virtual vehicle whose parts are built in different domains connected with the Functional Mock-up Interface (FMI) co-simulation standard. A real- time 1-D gas thermodynamics code ‘WAVE-RT’ is used as a virtual gasoline engine providing detailed information about any chosen parameters at every engine crank angle. Real-time predictive spark ignition combustion is enhanced by a knocking model ensuring the correct combustion response within the entire engine operating range, keeping the engine from knocking during both cold and hot states.

The vehicle driveline equipped by an engine cooling circuit is modelled in ‘IGNITE’ physics-based system simulation package driving the virtual vehicle through a chosen emission cycle. Finally, the exhaust aftertreatment uses ‘R-CAT’ code for modelling and solving relevant chemical kinetic reactions within the catalyst brick. DoE optimization is used for minimizing CO₂ as well as other undesired emissions allowing the aftertreatment to be properly controlled by a suitable engine control strategy. This unique solution opens new possibilities due to its rapid simulation speed. The whole virtual vehicle model runs real-time on a common computer, providing fast turnaround times and allowing multiple optimization runs in parallel or the performance of the simulation on a laptop on board a tested vehicle.