A method for diesel combustion simulation in a “real-time” engine model for control applications.

It is well known that in automotive applications problems related to control and management are nowadays of paramount importance to improve engine performance and to reduce fuel consumption and pollutant emissions. In the design of control and diagnostics systems the use of theoretical models proved to be very promising, also to reduce development time and costs, as widely documented in the open literature. From this point of view, the complexity of present engines due both to the continuous enhancement of existing subsystems (e.g., turbochargers, exhaust gas recirculation systems, aftertreatment components, etc.) and the introduction of specific devices (e.g., Variable Valve Actuation systems) give rise to challenging issues for modelling development and application. The paper describes a theoretical model of a turbocharged automotive engine built up starting from the original library developed in Simulink® by the authors for the simulation of last generation automotive engines. The tool was used in former works to build up Mean Value Models (MVMs) of automotive Diesel engines suitable for “real-time” simulations and therefore used in several HiL applications [6]. The model proposed in this work is an enhancement of the mentioned ones and it has been developed in order to allow for a “crank-angle” simulation of the engine. To this extent several block were built up for the simulation of intake and exhaust valves (with user-defined lift curves and variable actuation) and of in-cylinder processes. Combustion process has been described following a classic single-zone approach based on a proper Heat Release Rate (HRR). The library blocks were used to model other components of the intake and exhaust system. Through a specific calibration procedure, the model was fitted on a typical layout of an automotive Diesel engine allowing for transient simulation of the engine behaviour. Calculated results are compared in the paper with experimental data measured on a test bench, showing a good agreement both in steady and transient operating conditions