Control-Oriented Model of the Mean and Dispersion of Diesel Combustion Phasing With Ignition Assist
Abstract Diesel engines equipped with ignition assist devices such as glow plugs may improve combustion behavior at low temperatures and with low cetane fuels found in remote fields. The coordination of injection timing and the energy input of the ignition assist needs to continuously adjust to maintain the best combustion phasing at all conditions. However, most diesel engines do not use closed-loop combustion control and operate in a sub-optimal manner because the dispersion of combustion phasing, also known as cycle-to-cycle variability, requires careful feedback controller design. This work presents an initial investigation of a control-oriented model that captures the average and statistical influence of commercial glow plugs used for ignition assist beyond the start-up phase. Experiments were conducted at a single speed and load operating point as a proof of concept to obtain a model that quantifies the combustion phasing statistics and thus can guide feedback control design. The developed phenomenological model includes the engine’s thermal state because it impacts combustion behavior over the course of repeated experiments. The 3-term mean phasing model and 2-term standard deviation model estimate start of combustion within 0.6 and 0.2 crank angle degrees, respectively, and can be readily expanded to more operating conditions.