Current and future legislation demands ever decreasing levels of pollution from gas
turbine engines, and with combustor performance playing a critical role in resultant
emissions, a need exists to develop a greater appreciation of the fundamental causes of
unsteadiness. Particle Image Velocimetry (PIV) provides a platform to enable such
investigations. This thesis presents the development of PIV measurement methodologies
for highly turbulent flows. An appraisal of these techniques applied to gas turbine
combustors is then given, finally allowing a description of the increased understanding of
the underlying fluid dynamic processes within combustors to be provided.
Through the development of best practice optimisation procedures and correction
techniques for the effects of sub-grid filtering, high quality PN data has been obtained.
Time average statistical data at high spatial resolution has been collected and presented
for generic and actual combustor geometry providing detailed validation of the turbulence
correction methods developed, validation data for computational studies, and increased
understanding of flow mechanisms. These data include information not previously
available such as turbulent length scales.
Methodologies developed for the analysis of instantaneous PIV data have also allowed
the identification of transient flow structures not seen previously because they are
invisible in the time average. Application of a new `PDF conditioning' technique has
aided the explanation of calculated correlation functions: for example, bimodal primary
zone recirculation behaviour and jet misalignments were explained using these
techniques. Decomposition of the velocity fields has also identified structures present
such as jet shear layer vortices, and through-port swirling motion. All of these
phenomena are potentially degrading to combustor performance and may result in flame
instability, incomplete combustion, increased noise and increased emissions.
A Doctoral Thesis. Submitted in partial fulfillment of the requirements for the award of Doctor of Philosophy of Loughborough University.