Model Based Charge Control for 3-Cylinder TGDI Miller Engine Containing Variable Geometry Turbocharger 2024-26-0043
For ensuring environmental safety, strong emphasis on CO2 pollution reduction is mandated which led to evolution of miller cycle engines.
However, the inherent Miller engine characteristic is the lower volumetric efficiency when compared to otto engines because of which small turbo chargers with variable geometry turbines are used to induct air into the engine. With miller engine and VGT turbo charger combination arises the challenges of charge controllability because of lower inertia and reduced vane control area. With conventional turbo charger control methods, the response time is slow thereby leading to turbo lag or severe over boosting, this is overcome by accurate engine modelling and using the same as input for charger control.
In this study, model-based calibration approach was performed on a 3-cylinder Miller GDI 1.2L engine to model the charge exchange of the engine and use the same for determination required turbine vane positions to achieve the desired airflow induction into the engine. The charge exchange model consists of two components namely compressor model and the turbine model. Compressor model predicts the air induction into the engine by virtue of the power provided by the turbine and trajectory of compressor speed. The compressor speed trajectory is determined as a function of compressor inertia and the pressure upstream and downstream of compressor. The Turbine modelling involves accurately modelling the exhaust pressure and temperature across the turbine and deriving the turbine power and Vane position required to achieve the desired pressure at compressor. In addition to this, better surge, and compressor overspeed prediction is possible.
Results show that with model-based turbo control mechanism, efficient control of airflow induction is possible with reduced turbo lag. Additionally Model based calibration approach reduces Calibration Effort for Variants using Same Turbo charger.