Pressure waves and cavitation in diesel fuel injection rate characterisation
File(s)
Author(s)
Pearce, Daniel
Type
Thesis
Abstract
To meet stringent control and emissions requirements, diesel fuel injectors need to be characterised accurately in terms of timing and rate of change of mass flow rate. Such characterisation is carried out with various rate metering devices which overwhelmingly utilise a liquid into liquid injection process. These devices have historically been hampered by unwanted 'noise' in the measurement signal whose source was poorly understood and mitigation relied on post processing filtering techniques. A model of a constant volume metering device with optical access was constructed and a hybrid schlieren laser imaging technique applied to the flow field external to the nozzle with simultaneous chamber pressure measurement. This technique is sensitive to the second derivative of density and so directly able to visualise pressure waves within the domain. LES simulations were also performed to extract relationships not available through experimental data. The experimental results show cloud cavitation in the shear layer of the jet whose vapour bubbles begin collapsing shortly after injection begins and persist more than 500 microseconds after the end of injection. Compression waves due to the collapse of cavitation bubbles are visualised directly and parameters such as their spatial origin and time are calculated. Compression wave diffraction, reflections and interaction between individual jets are demonstrated. The 'noise' in a constant volume chamber is therefore shown to actually be an accurate representation of the pressure field arising from the superposition of complex flow phenomena in the near field of the injector nozzle due to an interaction of pressure waves and cavitation. The novel use of a hybrid schlieren technique demonstrates the utility of this arrangement to fully three dimensional problems. Cavitation and its associated processes are shown to be the dominating force in liquid to liquid injection processes for the flows encountered in fuel injection metering.
Version
Open Access
Date Issued
2017-06
Date Awarded
2017-12
Copyright Statement
Attribution NoDerivatives 4.0 International Licence (CC BY-ND)
Advisor
Taylor, Alex
Hardalupas, Yannis
Jones, William
Sponsor
Delphi (Firm)
Publisher Department
Mechanical Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)