The effects of fluid properties and surface geometry on near nozzle fuel spray formation

Student thesis: Doctoral Thesis

Abstract

Combustion engines face stringent emissions targets due to global warming. Achieving these targets requires a thorough understanding of combustion and airfuel mixing processes, which are critical to reducing pre-catalyst emissions. The development of sustainable and synthetic liquid fuels, designed as drop-in replacements for fossil fuels, presents a promising solution. However, variations in the chemical composition of these fuels may affect their physical properties, altering fuel-air mixing and fluid dynamics, and potentially impacting emissions. Fuel injection events involve transient processes that lead to the expulsion of large, uncontrolled liquid structures from orifices, contributing to fuel-rich regions and soot (PM) production. A phenomenon known as fuel injector tip wetting, where fuel films form on the injector tip, can lead to carbonous deposits and degrade spray production. The impact of fuel properties on the processes which lead to these films, especially in Gasoline Direct Injection (GDI) systems, remains underexplored.

Experiments were conducted in various spray chambers and an optical diesel engine under conditions relevant to light passenger vehicles and gas-turbine aircraft. High-speed microscopy and machine learning (ML) analysis revealed that fluid properties influence fuel-air mixing and near-nozzle spray characteristics. High-speed infrared thermography was used to measure nozzle tip temperatures, providing critical boundary conditions for spray experiments.

The novel contributions of this work are a new methodology for measuring the injector nozzle tip temperature in-engine during normal operation. The influence of fluid properties remain subtle in the near-nozzle region, however do appear to
show minor trends when characterising sprays. Finally a novel application of ML object detection algorithms has been demonstrated for the application of classifying droplets under going transcritical mixing. This approach unveiled that two mixing regimes were possible at one operating condition whereas only one was previously
assumed.

Date of AwardOct 2024
Original languageEnglish
Awarding Institution
  • University of Brighton
SupervisorCyril Crua (Supervisor), Guillaume De Sercey (Supervisor) & Konstantina Vogiatzaki (Supervisor)

Cite this

'