Date of Award


Document Type

Master's Thesis

Degree Name

Master of Science in Mechanical Engineering (MS)

College, School or Department Name

Department of Mechanical Engineering-Engineering Mechanics


Mahdi Shahbakhti


Homogeneous charged compression ignition (HCCI) is a promising combustion mode for internal combustion (IC) engines. HCCI engines have very low NOx and soot emission and low fuel consumption compared to traditional engines. The aim of this thesis is divided into two main parts: (1) engine instrumentation with a step towards converting a gasoline turbocharged direct injection (GTDI) engine to an HCCI engine; and (2) developing controller for adjusting the crank angle at 50% mass fuel burn (CA50), exhaust gas temperature Texh, and indicated mean effective pressure (IMEP) of a single cylinder Ricardo HCCI engine.

The base GTDI engine is modified by adding an air heater, inter-cooler, and exhaust gas recirculation (EGR) in the intake and exhaust loops. dSPACE control units are programmed for adding monitoring sensors and implementing actuators in the engine. Control logics for actuating electronic throttle control (ETC) valve, EGR valve, and port fuel injector (PFI) are developed using the rapid control prototyping (RCP) feature of dSPACE. A control logic for crank/cam synchronization to determine engine crank angle with respect to firing top dead center (TDC) is implemented and validated using in-cylinder pressure sensor data.

A control oriented model (COM) is developed for estimating engine parameters including CA50, Texh, and IMEP for a single cylinder Ricardo engine. The COM is validated using experimental data for steady state and transient engine operating conditions. A novel three-input three-output controller is developed and tested on a detailed physical HCCI engine plant model. Two type of controller design approaches are used for designing HCCI controllers: (1) empirical, and (2) model-based. A discrete sub-optimal sliding mode controller (DSSMC) is designed as a model-based controller to control CA50 and Texh, and a PI controller is designed to control IMEP. The results show that the designed controllers can successfully track the reference trajectories and can reject the external disturbances within the given operating region.