1D model and rule-based calibration strategy to improve the performance of a turbocharged spark ignition engine over the whole engine map

This paper deals with the performance improvement of a turbocharged spark ignition (SI) engine redesigned by adopting a refined 1D model and a rule-based (RB) calibration strategy. The new SI engine operates without the throttle valve and combines an early intake valve strategy with a two-stage compression ratio device. The resulting SI unit represents a promising short-term technical solution and it is also suitable as a flexible fuel engine.

In a first phase, the engine model is validated with the experiments in the original configuration, and then it is virtually modified to obtain the redesigned solution. In this redesign process, the effects of the variations of engine geometry and valve strategy on combustion and performance are considered by phenomenological in-cylinder sub-models. Afterwards, the RB strategy is implemented into the re-designed engine model and assessed by comparing the results with the ones of an advanced calibration approach, based on an optimization with genetic algorithm performed by coupling the 1D model with an optimizer.

The RB method replicates with acceptable accuracy the numerical trends of performance and control parameters of the SI engine coming from the optimizer. Once the capability is verified, the RB strategy is adopted to compute the steady operating map of the redesigned engine, with significantly less effort than an optimization. The map provides noticeable benefits in terms of full load torque, fuel consumption at medium-to-low loads and a slight extension of the minimum fuel consumption region. The combustion stability is maintained at acceptable levels, although it is improvable at very low loads and speeds. The presented methodology has a general validity for conventional SI engines and can be efficiently exploited to support the redesign stage of SI units for improved performance, with reduced computational effort. It also offers a method to rapidly compute the operating map of SI engines for subsequent on-vehicle analyses.