San-Mou Jeng, manager of the Combustion Diagnostic Laboratory (CDL) for his guidance and constant encouragement throughout the course of this thesis and also for providing financial aid. I would also like to thank Jun Cai for his help in the implementation of my setup and for helping me with my experiments. I also wish to thank Nick Flohre for guiding me through the initial stages of my work and for providing many helpful insights on the way. Thanks are also due to Mr. Curtis Fox, associate researcher at the CDL, for providing invaluable technical insights and help with the design and manufacture of the setup. Thanks to Omar Elshamy for the several technical discussions we had which really helped me gain an insight into the subject. And finally, I thank my family and friends for their constant support and encouragement without which I could not have persevered.

No abstract

No abstract

Summary The NexGen (Sonic) burner is the new burner developed by the Federal Aviation Administration, FAA, to replace old oil burners used for the required fire certification tests on power plant‐related materials, as it provides the capability to control both air and fuel flow rates. During a fire test, the burner is supposed to simulate a certain fire condition, so the flame properties should be robust and repeatable. The NexGen burner can achieve this due to the precise fuel and air controls. However, the current calibration criterion (ISO2685:1998 and AC20‐135) may not be good enough to ensure consistent flame properties. In the presented work, the sensitivity of the burner performance to air and fuel flow rate, as measured by the temperature and heat flux for calibration purposes, was studied. Additionally, the influence of the turbulator and the thermocouple size used for flame calibration was also studied. The impact of varying fuel/air ratio and thermocouple sizes was studied by conducting fire tests on aluminum samples, to show the inadequacies in the current calibration standards.

In this experimental work, a series of tests have been conducted to further study the aerodynamics of linearly-arranged 5-swirler arrays, using Laser Doppler Velocimetry (LDV). Two major parameters have been investigated for the 5-swirler arrays in this work, including the inter-swirler spacing, and the end wall distance. An additional effect of dome recession was studied for a single swirler in order to provide some insight into the results for the 5-swirler arrays. The 5-swirler arrays with an inter-swirler spacing of 1.75D, 2D, 2.5D, and 2.75D were studied, where D is the diameter of swirler exit. For the inter-swirler spacing of 1.75D or 2D, the center swirler is shown to have a weak, large center toroidal recirculation zone (CTRZ). The swirlers next to the center swirlers have strong, compact CTRZs whereas the outer swirlers have weak, large CTRZs. Thus, starting from the center swirler, the CTRZs exhibit a large – small – large pattern in either direction. For the swirler spacing of 2.5D or 2.75D, the trend is reversed, with a strong CTRZ for the center swirler and a small – large – small CTRZ pattern. The end wall distances of 0.75D, 1D, 1.25D, and 2D cases were studied. The end wall distances are seen to affect the strength of each CTRZ and the corner swirling flow pattern, though the bulk flow structure did not change significantly. The dome recession shows a clear impact on the expansion of swirling flow generated by a single swirler. As the dome recession distance increases, the expansion angle of the swirling jet reduces significantly for the unconfined cases. A phenomenological description is discussed to understand the reason for the periodically alternating CTRZ pattern observed in the experimental results of the 5-swirler arrays.