CHARACTERIZATION OF THE COMBUSTION PROCESS OF LIQUIFIED PETROLEUM GAS AND DIMETHYL ETHER BLENDS FOR USE IN SPARK IGNITED INTERNAL COMBUSTION ENGINES

Abstract

To meet the increasing demand for carbon intensity reduction in the heavy-duty transportation sector, this dissertation investigates the development of high-efficiency combustion strategies utilizing both liquefied petroleum gas (LPG) and renewable dimethyl ether (rDME). LPG serves as a viable alternative fuel primarily due to its ability to reduce emissions and its favorable physical and chemical properties, which allow for efficient transport and storage in a liquid state at moderate pressures. The research begins by establishing a numerical foundation for baseline LPG combustion, utilizing an extreme flash KH-RT spray model coupled with the ALPINE 153 chemical kinetic mechanism. This framework was validated against experimental High Pressure Spray Chamber (HPSC) data using Schlieren and Mie-scattering imaging techniques to ensure the phase change and flash-boiling plume expansion of LPG were resolved before moving to engine scale simulations. With the spray physics validated, a numerical investigation into mixture formation for pure LPG was conducted to characterize the fluid dynamic drivers of stable Direct Injection (DI) operation compared to Port Fuel Injection (PFI) benchmarks. These 3D CFD studies identified that Direct Injection LPG operation is exclusively sensitive to Start of Injection (SOI) timing due to in-cylinder charge cooling and turbulent decay. Specifically, the work identified that a late-cycle recovery of Turbulent Kinetic Energy (TKE) at SOI 120° bTDC provides the necessary aerodynamic shear to overcome reduced residence times, identifying the piston-top velocity profile as a potential universal design parameter for stable DI operation across varying compression ratios. To further reduce the carbon intensity of LPG operation, this work explores the potential of blending LPG with more reactive, renewably produced fuels such as DME. The inherently high MON and RON of LPG allow it to be utilized in higher compression ratio engines due to its strong resistance to autoignition compared to traditional petroleum fuels. This characteristic places LPG in a unique position where it can be blended with highly reactive fuels while maintaining favorable and controllable combustion. Utilizing rDME blends leverages LPG’s favorable combustion properties to reduce carbon intensity on two fronts, through tailpipe emissions and cleaner fuel production. To understand the properties of these novel LPG/DME blends, octane characterization was experimentally determined using a modified Cooperative Fuels Research (CFR) engine. This testing identified a non-linear promoting effect of rDME on autoignition and established a 30% DME blend as a viable 89-octane gasoline surrogate. Additionally, high-speed Schlieren spray imaging confirmed that rDME addition up to 30% does not fundamentally alter macroscopic spray morphology, ensuring compatibility with existing LPG hardware. The final phase of the work involved experimental testing on a heavy-duty single-cylinder Cummins X15 engine and the development of the ALPINE-DME 158 chemical kinetic mechanism. The mechanism was rigorously validated across 0D ignition delay times and 1D laminar flame speeds before being implemented in a 3D CONVERGE environment to predict engine performance of LPG/DME blends. Experimental results demonstrated a significant departure from natural gas performance limits, achieving a peak brake thermal efficiency (BTE) of 41.1% at low loads and 42.9% BTE at high load conditions by leveraging an increase of end gas autoignition (EGAI). The ALPINE-DME 158 mechanism successfully captured the experimental heat release rates and autoignition events by resolving radical bottlenecks specifically premature HO2 and H2O2 accumulation observed in legacy chemical kinetic models. Collectively, this work provides a validated, predictive roadmap for leveraging high-reactivity oxygenated fuels to narrow the efficiency gap between spark-ignited and diesel heavy-duty architectures.